Oxygen regulation of vascular smooth muscle cell. proliferation and survival

Transcription

1 Oxygen regulation of vascular smooth muscle cell proliferation and survival by Julie Basu Ray A thesis submitted in conformity with the requirements for the degree of doctor of philosophy Institute of Medical Sciences University of Toronto Copyright by Julie Basu Ray, 2009

2 Oxygen regulation of vascular smooth muscle cell proliferation and survival Julie Basu Ray Doctor of Philosophy Institute of Medical Sciences University of Toronto 2009 ABSTRACT Arterial smooth muscle cells (SMCs) from the systemic and pulmonary circulations experience a broad range of oxygen concentrations under physiological conditions. The hypoxic response, however, has been inconsistent, with both enhanced proliferation and growth arrest being reported. This variability precludes a definitive conclusion regarding the role of oxygen tension in arterial disease. In the first part of this study, we determined if hypoxia elicits different proliferative and apoptotic responses in human aortic SMCs (HASMCs) incubated under conditions which do or do not result in cellular ATP depletion and whether these effects are relevant to vascular remodeling in vivo. Gene expression profiling was used to identify potential regulatory pathways. In HASMCs incubated at 3% O 2, proliferation and progression through G 1 /S interphase are enhanced. Incubation at 1% O 2 reduced proliferation, delayed G 1 /S transition, increased apoptosis and cellular ATP levels were reduced. In aorta and mesenteric artery from hypoxia exposed rats, both proliferation and apoptosis are increased after 48hrs. p53 and p21expression is differentially affected in HASMCs incubated at 1% and 3% O 2. Hypoxia ii

3 induces a state of enhanced cell turnover, conferring the ability to remodel the vasculature in response to changing tissue metabolic needs while avoiding the accumulation of mutations that may lead to malignant transformation or abnormal vascular structure formation. A unifying hypothesis in which events at the G 1 /S transition and apoptosis activation are coordinated by effects on p53, p21, their downstream effector genes and regulatory factors is proposed. Differences in the contractile responses of systemic and pulmonary arterial smooth muscle cells to hypoxia are well studied. Differences in proliferation and survival are anticipated because of differences in embryonal cell origin, oxygen concentrations within their respective microenvironments and in cellular energetics but these responses have not been directly compared. In the second part of the study, human pulmonary arterial SMCs (HPASMCs) proliferated at oxygen concentrations which inhibited cell growth in HASMCs. HPASMCs survived and maintained their intracellular ATP levels at levels of hypoxia sufficient to deplete ATP and induce apoptosis in HASMCs. In vivo studies in rats show proliferation and apoptosis in main or branch PASMCs only after 7 days of hypoxia. VSMCs are able to proliferate under hypoxic conditions as long as cellular ATP levels are maintained. HPASMCs have an enhanced capacity to maintain cellular energy status compared to HASMCs and hence their viability is preserved and the proliferative response predominates at lower oxygen concentrations. iii

4 I have become my own version of an optimist. If I can't make it through one door, I'll go through another door - or I'll make a door. Something terrific will come no matter how dark the present. Rabindranath Tagore Take up one idea. Make that one idea your life - think of it, dream of it, live on that idea. Let the brain, muscles, nerves, every part of your body, be full of that idea, and just leave every other idea alone. This is the way to success. Swami Vivekananda iv

5 ACKNOWLEDGMENTS The path towards this work spans several years of research and it is a pleasure to thank the many people who made this thesis possible. I acknowledge my debt to all those who have helped along the way, who have been involved and contributed to the presented ideas and understanding gained. It is difficult to overstate my gratitude to my Ph.D. supervisor, Dr. Michael E Ward. With his enthusiasm, his inspiration, and his great efforts to explain things clearly and simply, he helped to make research fun for me. I am indebted to his continued encouragement and invaluable suggestions especially during my thesis-writing period. He has been a wonderful guide and a great teacher. I would also like to include my gratitude to Dr. Linda Penn, my cosupervisor, and Dr. Philip Marsden who have provided support for this research all along the way. I am deeply indebted to my student colleagues at the Terrence Donnelley Research Laboratories at the St. Michael s Hospital, for providing a stimulating and fun environment in which to learn and grow. I would specially like to thank Jeff He, Lakshmi Kugathasan, Massey Rezai and Yupu Deng for their extremely valuable support, and insights. Many others who have been involved also deserve recognition. It is, however, not possible to list them all here. Their support in this effort is, however, greatly appreciated. I would like to thank the many people who have initiated me into the rites of science - my high school teachers at Calcutta Girls High School, my undergraduate teachers at Presidency College, and my graduate teachers at the University of Calcutta, India. v

6 I wish to thank my friends in high school, college and university, and my friends in Toronto, Buffalo and Boston, whose continued support helped me get through the difficult times, and for all their emotional support, camaraderie, entertainment, and caring they provided. Finally I want to thank my family. A special thought is devoted to my parents, Anil and Jayanti Ghosh for a never-ending support. They bore me, raised me, supported me, taught me, and loved me. I am indebted to my entire family for providing a loving environment for me. The encouragement and support from my husband, Indranill, have always been a powerful source of inspiration and energy. Lastly, and most importantly, I thank from the bottom of my heart my son, Ishan, whose adjustment and sacrifice of many a childhood demand and wish have helped me sail through to this goal. To him I dedicate this thesis. vi

7 CONTRIBUTIONS The work presented in Chapter 2 has been published in Am J Physiol Heart Circ Physiol Feb 2008; 294: H839 - H852. Basu Ray J, Arab S, Deng Y, Liu P, Penn L, Courtman DW, Ward ME. Oxygen Regulation of Arterial Smooth Muscle Cell Proliferation and Survival. Permission has been obtained from the American Physiological Society and all of the authors for inclusion of the paper in the thesis. As the first author of the publication, I contributed to study design, figure making and manuscript writing. I performed all of the experiments and data analysis except Table 2.1: Experiment and data analysis was done by Dr. Sara Arab from Dr. Peter Liu s lab at the Toronto Genomic Core Centre at the Hospital for Sick Children. Dr. Yupu Deng has helped with the animal sacrifices for Figures 2.10 and 2.11 and confocal microscopic analysis of sections. The work presented in Chapter 3 has been written into a manuscript and is expected to be submitted for publication before December As the first author of this manuscript, I contributed to study design, figure making and manuscript writing. I performed all of the experiments and data analysis. Dr. Yupu Deng has helped with the animal sacrifices for Figures 3.7 and 3.8. In the Supplement chapter, experiment design and data analysis for Figure S1 has been done by Karen Ho from Dr. Philip A Marsden s lab. The research work has been funded by Canadian Institute of Health Research Grant and Keenan Collaborative Research Award by Keenan Research Foundation, Toronto. vii

8 TABLE OF CONTENTS ABBREVIATIONS xi LIST OF FIGURES xiii LIST OF TABLES xvi CHAPTER 1 Review of literature 1.1 Introduction Systemic and pulmonary circulations Vascular smooth muscle cells Oxygen delivery Physiological responses to hypoxia Systemic responses Regulation of cellular metabolism Regulation of gene expression Hypoxia-inducible factors Regulation of HIF activity HIFs as transcriptional regulators HIF independent transcriptional activation Hypoxic regulation of mrna stability Hypoxic repression of transcription Hypoxic control of protein translation Cell cycle and hypoxia The mammalian cell cycle Effects of hypoxia on cell cycle Apoptosis and hypoxia Apoptotic pathways Regulation of apoptosis during hypoxia Role of p53 in hypoxia-induced apoptosis 38 viii

9 Role of Bcl-2 family proteins Role of PI3-kinase pathway Role of electron transport chain inhibition Oxygen sensing mechanisms Evidence of heme as oxygen sensor NAD(P)H oxidases Mitochondria Thesis objective Aims and hypotheses 52 CHAPTER 2 Oxygen regulation of systemic arterial smooth muscle cell proliferation and survival 2.1 Introduction Materials and Methods Antibodies and reagents Cell culture studies Cell counting [ 3 H]-thymidine incorporation Ki67 protein levels Annexin V- Propidium iodide labeling Caspase activation TUNEL Cell cycle analysis Mitochondrial membrane potential Intracellular ATP concentration Western blotting Microarray analysis In vivo apoptosis In vivo proliferation 64 ix

10 2.3 Results Discussion 89 CHAPTER 3 Oxygen regulation of pulmonary arterial smooth muscle cell proliferation and survival 3.1 Introduction Materials and Methods Antibodies and reagents Cell Culture Studies Cell counting Cell cycle analysis / BrdU incorporation Annexin V- Propidium iodide labeling Caspase activation Mitochondrial membrane depolarization Measurement of Intracellular ATP concentration Western blotting In vivo apoptosis In vivo proliferation Results Discussion 128 CHAPTER 4 General discussion and conclusions 135 CHAPTER 5 Future directions 143 Supplement 150 REFERENCES 158 x

11 ABBREVIATIONS ANOVA analysis of variance APAF apoptotic protease activating factor AP-1 activator protein 1 ARD1 arrest defective 1 ARNT aryl hydrocarbon receptor nuclear translocator ATM ataxia telangiectasia mutated ATP adenosine triphosphate ATR ATM and rad3 related Bcl B-cell leukemia/lymphoma bhlh basic helix-loop-helix BNIP BCL2/adenovirus E1B 19kD interacting protein like BrdU Bromodeoxyuridine CBP CREB binding protein CCN Cyclin CDK Cyclin dependent kinase CDKI Cyclin dependent kinase inhibitor CO carbon monoxide Dec1 deleted in esophageal cancer 1 DPG 2,3-disphosphoglycerate eif eukaryotic initiation factor ets-1 erythroblastosis virus E26 oncogene homolog-1 ET-1 endothelin-1 ETC electron transfer chain FIH factor-inhibiting HIF H 2 O 2 hydrogen peroxide HIF hypoxic inducible factor HO heme oxygenase HPV hypoxic pulmonary vasoconstriction HRE hypoxia responsive element HUVEC human umbilical vein endothelial cell HVR hypoxic ventilatory response IAP inhibitor of apoptosis IRES internal ribosomal entry site JC-1 5,5,6,6 -tetrachloro-1,1,3,3 -tetraethylbenzimidazolcarbocyanine iodide K v voltage activated potassium current MAPK mitogen-activated protein kinase MMP-2 matrix metalloproteinase-2 mtor mammalian target of rapamycin NLS nuclear localization signal NO nitric oxide ODD oxygen dependent degradation domain ORP150 oxygen-regulated protein 150 xi

12 PERK PKR-like endoplasmic reticulum kinase PHD prolyl hydroxylase PI propidium iodide po 2 partial pressure of oxygen PPAR peroxisome proliferator-activated receptor gamma ROS reactive oxygen species SRC sarcoma (Schmidt-Ruppin A-2) viral oncogene homolog TAD transactivation domain TIF transcriptional intermediary factor 1 TIMP tissue inhibitors of matrix metalloproteinase TNF- tumor necrosis factor UTR untranslated region VEGF vascular endothelial growth factor VHL von Hippel-Lindau VSMC vascular smooth muscle cell xii

13 LIST OF FIGURES CHAPTER 1 Figure 1.1 Regional distribution of po 2 from the airways to the cytosol 7 Figure 1.2 Oxyhaemoglobin dissociation curve 8 Figure 1.3 HIF-1, HIF-2 and HIF-3 subunit structure 13 Figure 1.4 Regulation of HIF activity 16 Figure 1.5 Cell cycle phases and G1/S transition 30 Figure 1.6 Pathways of apoptosis 37 Figure 1.7 Heme sensor model 43 Figure 1.8 Structure of NAD(P)H oxidase 45 Figure 1.9 NAD(P)H oxidase as oxygen sensor 46 Figure 1.10 Mitochondrial electron transport chain 47 CHAPTER 2 Figure 2.1 Figure 2.2 (A) Effects of hypoxia on HASMC cell numbers. Effects of HASMC cell numbers to PDGF-BB at (B) 1% O 2 and (C) and 3% O [ 3 H]-Thymidine incorporation in HASMCs after incubation at (A) 1% O 2 and (B) 3% O 2 compared with the normoxic cells. 70 Figure 2.3 The percentage of cells positive for the Ki67 antigen after incubation at (A) 1% O 2 and (B) at 3% O Figure 2.4 CDC6 (A and B) and MCM2 (C and D) protein levels after normoxic and hypoxic (1% O 2 and 3% O 2 ) incubation. 72 xiii

14 Figure 2.5 Flow cytometric analysis of propidium iodide stained cells at (A) 1% O 2 and (B) 3% O Figure 2.6 Apoptosis assays in HASMCs. (A) Annexin V/PI (B) Caspase activity and (C) TUNEL. 75 Figure 2.7 Mitochondrial membrane depolarization after incubation at 1% O 2 (A) and 3% O 2 (B). 78 Figure 2.8 Figure 2.9 Cellular ATP concentration after incubation at 1% O 2 (A) and 3% O 2 (B). 79 Nuclear levels of (A) HIF-1, (B) p21 and (C) p53 after incubation of HASMCs at normoxic and hypoxic (1% or 3% O 2 ) conditions. 82 Figure 2.10 (A) PI staining of en face sections and (B) TUNEL in paraffin embedded sections of normoxic and hypoxic rat aorta and mesenteric artery. (C) Quantitative analysis 84 Figure 2.11 (A) Immunohistochemical staining of incorporated BrdU (B) Double staining with TO-PRO-3 and (C) -Smooth Muscle Actin in paraffin embedded sections of normoxic and hypoxic rat aorta and mesenteric artery (D) Quantitative analysis of incorporated BrdU and TO-PRO-3 staining. 87 CHAPTER 3 Figure 3.1 Effects of hypoxia on human pulmonary artery smooth muscle (HPASMC) (A) cell numbers and (B) viability 109 Figure 3.2 % BrdU incorporated cells in HPASMCs 111 Figure 3.3 (A,B) Annexin V/PI (C,D) Caspase activity and (E) JC-1 monomer formation after incubation at 3%, 1% and 0% O xiv

15 Figure 3.4 Cell cycle analysis of propidium iodide stained normoxic and hypoxic (3%, 1% or 0% O 2 ) HPASMC cells. 117 Figure 3.5 Nuclear levels of (A) p21, (B) p53 and (C) HIF-1 in HPASMCs at normoxic and hypoxic (1% or 3% O 2 ) conditions. 119 Figure 3.6 Cellular ATP concentrations in normoxic and hypoxic (3%, 1% or 0% O 2 ) HPASMC cells. 122 Figure 3.7 Figure 3.8 (A) PI staining of en face sections of normoxic and hypoxic rat pulmonary artery and pulmonary artery branch, (B) Quantitative analysis of PI stained cells, (C) TUNEL in paraffin-embedded sections of normoxic and hypoxic rat pulmonary artery and pulmonary artery branch (D) Quantitative analysis of TUNEL positive cells. 124 (A) Immunohistochemical staining of incorporated BrdU and -smooth muscle actin in paraffin-embedded sections of pulmonary artery and pulmonary artery branch from normoxic and hypoxic rats. (B) Quantitative analysis of BrdU positive cells. 126 Supplement Figure S1 Average fold change of mitochondrial DNA levels 152 Figure S2 Cytoplasmic levels of Phosphoglycerate kinase and Enolase protein in HPASMCs (A, C) and HASMCs (B, D), lactate concentration after incubation of HPASMCs (E) and HASMCs (F) under normoxic and hypoxic (3, 1 or 0% O 2 ) conditions. 153 xv

16 LIST OF TABLES Table 1.1 Hypoxia-inducible genes harboring HRE sequences 19 Table 2.1 Table 3.1 Table 3.2 Table 5.1 Normalized expression of pro- and antiproliferative genes and pro- and antiapoptotic genes under hypoxia. 80 Influence of hypoxia on pulmonary artery smooth muscle cell proliferation. 100 Medial wall thickness of pulmonary artery and pulmonary artery branch from normoxic and hypoxia exposed rats. 127 Phenotypic heterogeneity in pulmonary artery smooth muscle cells. 144 xvi

17 Chapter 1 CHAPTER 1 Review of literature 1

18 Chapter Introduction The efficient delivery of oxygen to the tissues of the body is required for aerobic ATP production to support their metabolic activities and as a substrate in the synthesis of a number of signaling molecules such as carbon monoxide and nitric oxide [1-3]. Inadequate oxygen supply will impair the capacity to meet these needs and result in the failure of vital functions. Hypoxia refers to conditions under which oxygen concentration becomes limiting for normal cellular processes [4]. The oxygen concentration in the atmosphere is 20.9% (partial pressure ~ 160 mmhg at sea level). The cells that comprise the vascular wall, however, experience much lower oxygen tensions (25 mm Hg at the preterminal arterioles) [5, 6] with even lower levels (bordering on anoxia) reported in vessels affected by disease [7-9]. At levels of 3-5% oxygen vascular cells are close to the hypoxic range, although oxygen availability is not yet limiting to cellular viability or function. Any further decrease in oxygen levels, however, will trigger hypoxia-induced responses, which includes regulation of both cell proliferation and/or cell survival to alter the structure of the vessels [10, 11]. These are aimed at both enhancing the capacity to utilize the available oxygen supply and, in the event that hypoxia is prolonged to protect cell viability and function. The molecular mechanisms underlying these responses are complex and remain poorly understood. Their elucidation will aid in the development of therapeutic approaches to ameliorate the effects of hypoxia in diseases associated with reduced systemic oxygen delivery. 1.2 Systemic and pulmonary circulations Structural and functional differences between the systemic and pulmonary circulations support their respective physiological functions. The pulmonary circulation is a low pressure 2

19 Chapter 1 system with a mean pressure of mm Hg, compared to mm Hg in the systemic circulation. The pulmonary vasculature is thin walled compared to the systemic circulation, and contains much less vascular smooth muscle. In the systemic circulation, 75 80% of vascular resistance is maintained by small muscular arterioles whilst resistance is relatively evenly distributed throughout the normal pulmonary circulation. Pulmonary vascular reactivity to endogenous and exogenous vasoconstrictors and to hypoxia is influenced by the level of basal tone. Baseline vascular tone is low in the normal lung but is enhanced during hypoxia due to mechanisms intrinsic to the smooth muscle and because of the effects of locally released and circulating vasoactive mediators such as Endothelin-1, vasoconstrictor prostaglandins, histamine and serotonin [12, 13]. In the systemic circulation basal tone is maintained by tonic activity of the sympathetic nervous system. Sympathetic innervation of the pulmonary circulation does exist and its activation has similar effects as in the systemic circulation but contributes little to the maintenance of basal vasomotor tone. Pulmonary arteries exhibit a vasoconstrictor response to hypoxia in contrast to the vasodilator response to hypoxia exhibited by the systemic circulation [14, 15]. In the foetus, this hypoxic pulmonary vasoconstriction (HPV) serves to increase pulmonary vascular resistance and divert the circulation through the ductus arteriosus. As a result the foetal pulmonary circulation only receives ~10% of the cardiac output unlike the situation after birth where exposure to atmospheric oxygen fully dilates the pulmonary circulation which henceforth receives 100% of the cardiac output. After birth, HPV is required for ventilation-perfusion matching. Despite its clinical and physiological relevance the mechanism of HPV remains largely unresolved. Recent 3

20 Chapter 1 hypothesis for HPV proposes hypoxia-induced inactivation of voltage activated potassium (K v ) channels in the pulmonary circulation [16]. The larger arteries in the vasculature provide little resistance to blood flow and therefore serve as a rapid conduit for blood to travel. The walls of these vessels contain large amounts of elastic and fibrous tissue. As the arteries branch into smaller arteries, the amount of elastic tissue in the walls decreases while the amount of smooth muscle increases. Arteries less than 0.1mm in diameter lose most of their elastic properties and are sometimes called muscular arteries. The combination of stiffness and flexibility enables arteries to act as pressure reservoirs to ensure a continual smooth flow of blood through the vasculature even when the heart is not pumping blood. The arterioles are the blood vessels that provide the greatest resistance to blood flow. In the systemic circuit, blood enters arterioles at an average pressure of about 90 mmhg and leaves them at a pressure of about 40 mmhg. The walls of arterioles contain little elastic material but have an abundance of circular smooth muscle that forms rings around the arterioles. Resistance is regulated by the contraction and relaxation of the circular smooth muscle. The arteries have two functions. One is to deliver an adequate supply of blood to peripheral tissues and to smooth out pressure oscillations due to intermittent ventricular ejection. The efficiency of conduit function is related to the width of the arteries and the almost constancy of mean blood pressure along the arterial tree. Resistance arteries with an internal diameter of 150 m contribute significantly to total peripheral resistance and basal vascular tone. Resistance arteries are continuously subjected to changes in mechanical forces (flow and pressure) that regulate active vasomotion, fitting blood flow continuously to local demands. The fundamental function of resistance-sized arteries is control of blood flow to the capillary beds, partly achieved 4

21 Chapter 1 by a putative pressure-sensing mechanism. Vascular remodeling is an adaptive process occurring in response to long-lasting changes in arterial pressure or flow, and whose ultimate effect tends to maintain the constancy of tensile and/or shear stresses. In response to blood pressure increase, the luminal diameter in large conduit arteries is usually unchanged while width of wall increases. In distal resistive arteries and arterioles, luminal diameter is reduced but medial layer is not hypertrophied. 1.3 Vascular smooth muscle cells The vascular smooth muscle cell (VSMC) in mature animals is highly specialized whose principal function is contraction and regulation of vessel tone and diameter, blood pressure, and distribution of blood flow. SMCs within adult blood vessels proliferate at a low rate and express a unique repertoire of contractile proteins, ion channels, and signaling molecules required for the cell's contractile function [17, 18]. Three independent embryonic origins for VSMCs have been identified: (a) Vessels that recruit SMCs from progenitors that originate in cardiac neural crest; (b) Coronary SMCs arise from mesothelial cells that line villus-like projections of the proepicardial organ and (c) Vessels that recruit SMCs from either lateral or splanchnic mesoderm depending on the position of a particular vessel within the embryo [19-23]. SMCs are also recruited from endothelial cells and from circulating multipotential stem cells at later stages of development and in adults [24-27]. The majority of VSMCs exhibit common properties regardless of their origins, however, certain lineage-specific differences in growth and transcriptional responses to various cytokines and other factors implicated in the progression of arterial diseases persist beyond the embryonic period. 5

22 Chapter 1 Unlike either skeletal or cardiac muscles, that are terminally differentiated, VSMCs within adult animals retain remarkable plasticity. The ability of VSMCs to be plastic in their growth responses is a key mechanism by which the vasculature responds to hemodynamic, developmental, and injurious stimuli. Biological processes during which VMSC growth is vital include vessel development, the vascular response to tissue injury, and vessel remodeling in response to changes in tissue demand [28-32]. Pathological examples include atherosclerosis, hypertension, restenosis post angioplasty, and vasculitis. In these situations, interactions between endothelial cells and VSMC, as well as between VSMC and other cells (e.g., fibroblasts, dendritic cells, and inflammatory cells) within the vessel wall, determine the nature of the growth response [33]. The role of SMCs is not a simple function of alterations in its growth state but rather is a function of very complex changes in the differentiated state of the SMC including increased matrix production [34], production of various proteases [35], participation in chronic inflammatory responses including production of inflammatory cytokines and expression of inflammatory cell markers [36, 37], altered contractility and expression of contractile proteins [38]. On one hand, the plasticity exhibited by VSMCs prevents accumulation of replication errors or mutations. On the other, the high degree of plasticity exhibited by the VSMCs predisposes the cells to abnormal environmental cues/signals that can lead to adverse phenotypic switching and the acquisition of characteristics that can contribute to development and/or progression of vascular disease. 1.4 Oxygen Delivery The primary function of the cardiovascular system is the delivery of oxygen that we breathe from the air to the cells that comprise the body. The partial pressure of oxygen (po 2 ) of 6

23 Chapter 1 dry air at sea level is ~160 mmhg (21/100 x 760=159.6) [39]. However, by the time the inspired air reaches the trachea it has been warmed and humidified by the upper respiratory tract and, taking water vapor pressure (47 mmhg) into account, the po 2 in the trachea while breathing air is ~150 mmhg (19.7%). By the time the inspired gas has reached the alveoli, the po 2 has fallen to about 100 mmhg (because of diffusion of oxygen and CO 2 from and into the alveolar gas, respectively. The median po 2 in systemic arteries is ~92 mmhg (12%), however, it falls to ~50 mmhg (6.6%) in arterioles and ~25 mmhg (3.3%) in precapillary arterioles and capillaries as a result of transarterial wall oxygen diffusion [6, 40, 41]. Oxygen gradients exist across the aortic wall where po 2 ranges from ~85 mmhg (11.2%) at the lumen to ~17mm Hg (2.2%) at a depth of 150 m [6]. Figure 1.1 Regional distribution of po 2 from the airways to the cytosol. Source: Ward J (2007) Oxygen sensors in context [39]. Oxygen diffuses from the alveolus to the pulmonary capillary until the po 2 in the capillary is equal to that in the alveolus. This process is normally complete by the time the blood has passed one third of the way along the pulmonary capillary. Oxygen dissociates from 7

24 Chapter 1 haemoglobin in red blood cells to the tissues according to the oxyhaemoglobin dissociation curve. Figure 1.2 Oxyhaemoglobin dissociation curve for normal adult haemoglobin. Source: Sequential branching of the arteriolar tree forms microvessels of decreasing diameter, which, in turn, increases the surface area per unit volume available for the diffusion of oxygen to the tissue [40]. In any oxygen-consuming tissue, the rate of intravascular oxygen loss is inversely related to arteriolar vessel diameter, thereby creating an intravascular longitudinal oxygen gradient. The affinity with which oxygen binds to hemoglobin is also influenced by ph, carbon monoxide (CO), temperature and erythrocyte 2, 3-disphosphoglycerate (DPG) concentration [42, 43]. Lastly, increasing capillary perfusion increases the capacity for oxygen extraction during exercise [44, 45]. 8

25 Chapter Physiological responses to hypoxia The tissue oxygen supply is regulated by the number and function of the blood vessels, whereas the demand is regulated by the number of cells in the tissue and their rate of metabolism. All nucleated cells in the body respond to reduced O 2 availability, through a series of coordinated responses in a time and oxygen concentration-dependent manner. Stimulusresponse pathways induced by hypoxia can be categorized as either acute or chronic. Acute responses are of rapid onset and short-term duration, whereas chronic responses are of delayed onset and long-term duration. This difference in kinetics reflect the underlying molecular mechanisms: acute responses involve post-translational modifications of existing proteins that alter their activity whereas chronic responses are comprised of transcriptional and posttranscriptional events involving changes in gene expression that result in the synthesis of novel proteins or increased synthesis of proteins already present in the cell Systemic responses During acute hypoxic exposure, oxygen supply to essential organs is maintained by the following: (i) the hypoxic ventilatory response increases the respiratory rate and tidal volume [46]. In humans, this is almost solely due to depolarization of glomus cells in the carotid body which leads to enhanced ventilation and increased alveolar oxygen concentrations [47], (ii) the pulmonary vasculature O 2 sensors initiate hypoxic pulmonary vasoconstriction (HPV) to increase efficiency of gas exchange. Pulmonary arterial vasoconstriction directs blood to better oxygenated regions of the lung while changes in bronchial and bronchiolar tone optimize the distribution of gas flow within the lung. An increase in pulmonary arterial blood pressure forces blood into greater numbers of alveolar capillaries than normal [22], (iii) activation of the 9

26 Chapter 1 sympathetic system increases oxygen extraction by increasing the heart rate and diverting unnecessary blood flow away from organs such as the kidneys and splanchnic viscera toward the essential organs like the heart and brain [43], (iv) Vessels in essential organs accommodate the increased blood flow through both a sympathetically-mediated increase in arteriolar tone and the release of vasodilators in areas of imbalance between metabolic demand and oxygen supply. The sympathetic excitation results partly through chemoreceptor reflexes and partly through altered baroreceptor function, (v) The O 2 sensors in the vasculature of other tissues activate expression of VEGF-1 to promote angiogenesis, (vi) O 2 sensors in the kidney and liver activate the expression of erythropoietin to up-regulate red blood cell mass to improve oxygen carrying capacity Regulation of cellular metabolism: Effect of hypoxia on mitochondria Mitochondria are the seat of oxidative phosphorylation and the main source of high energy phosphate bond molecules in normal cells. Studies on isolated mitochondria have shown that limited oxygen availability inhibits the electron transport chain and increases the proton leak, although phosphorylation is less affected [1]. The inhibition of the respiratory chain occurs at po 2 levels high above the K m of cytochrome c oxidase, indicating that a specific inhibitory mechanism, still unknown, is switched on well before oxygen concentration by itself would limit the activity of this enzyme [48] Cellular adaptation to hypoxia 10

27 Chapter 1 At the cellular level, adaptation to hypoxia is brought about on one hand by increased anaerobic glycolysis activity, and on the other hand by decreasing energy-consuming processes [49, 50]. Ion-motive ATPases and protein synthesis are the dominant energy-consuming processes of cells at normal metabolic rates, making up more than 90% of the ATP consumption in rat skeletal muscle and 66% in rat thymocytes [50]. As energy becomes limiting, protein synthesis and RNA/DNA synthesis are the first to be inhibited while Na + /K + pumping and Ca 2+ cycling are potentiated. This phenomenon, known as oxygen conformance, involves precise regulatory mechanisms mostly at the level of translation initiation [51]. The switch between aerobic and anaerobic pathways of ATP regeneration during hypoxia was first noted by Pasteur in the late 19th century, hence its name "Pasteur effect." Although glycolysis is less efficient than oxidative phosphorylation in the generation of ATP, in the presence of sufficient glucose, glycolysis can sustain ATP production due to increases in the activity of the glycolytic enzymes. 1.6 Regulation of Gene Expression Faced with a hypoxic challenge, the early physiological responses include increased ventilation and cardiac output, a switch from aerobic to anaerobic metabolism, improved vascularization, and enhancement of the O 2 carrying capacity of the blood. In the longer term, these responses are reinforced by up-regulation of genes encoding factors which facilitate these responses, such as (i) tyrosine hydroxylase, which is involved in dopamine synthesis in carotid body type I cells; (ii) glycolytic enzymes and glucose transporters Glut-1 and Glut-4; (iii) VEGF, PDGF which promote angiogenesis, and inducible NO synthase which increases vasodilation; 11

28 Chapter 1 and (iv) erythropoietin and transferrin receptors that favor erythrocyte production [52]. This transcriptional response is mediated in large part by the action of HIF Hypoxia-inducible factors Hypoxia increases nuclear translocation of a family of hypoxia inducible transcription factors (HIFs) activates expression of genes participating in the compensatory mechanisms that support cell survival in a potentially lethal microenvironment [53]. HIF transcription factors are composed of one of three alpha subunits (1, 2 or 3 ), and beta ( ) subunits. HIF-1 is also denoted as the aryl hydrocarbon receptor nuclear translocator (ARNT) [54]. In the subunit, the basic helix-loop-helix (bhlh) and the Per Arnt Sim (PAS) domains in the N-terminus are important for dimerization and DNA binding [55-57]. HIF-1 and HIF-2 proteins also contain two transactivation domains (TADs) in their C-terminal region. Within the N-terminal TAD there is an oxygen-dependent degradation (ODD) domain that is responsible for degradation of the subunit under normoxic conditions [58, 59]. The main function of the C-terminal TAD is to recruit transcriptional coactivators such as CBP, p300, SRC1 and TIF-2 [56, 60, 61]. Under hypoxic conditions, the HIF heterodimer (named HIF-1, -2 or -3) translocates to the nucleus where it binds to a core DNA sequence (5 -ACGTG-3 ) - the hypoxia-responsive element (HRE), located in the promoter/enhancer regions of many hypoxia-regulated genes [53, 62]. 12

29 Chapter 1 Of the three HIF- -subunits, HIF-1 is the best studied and characterized to date. Nevertheless, the understanding of HIF-2 (also known as endothelial PAS domain protein 1, EPAS-1) function has increased dramatically, whereas the most recently identified and consequently less-well studied subunit is HIF-3 [63, 64]. Although the HIF-1 and -2 subunits can bind to the same DNA motifs, they appear to control rather distinct biological functions. Mouse knockout studies have shown the vital importance of HIF-1 for development and survival, and that HIF-2 -/- mice have different phenotypes depending on their genetic 13

30 Chapter 1 background, thus illustrating the importance of HIF-2 [65, 66]. In addition, a recent study has demonstrated that HIF-2 cannot functionally substitute for HIF-1 in embryonic stem cells [67]. Whereas HIF-1 is expressed in virtually all cell types, HIF-2 exhibit a more restricted expression pattern in endothelial cells and in catecholamine producing cells in the organs of Zuckerkandl [68]. The biological function of HIF-3 is under investigation, but a HIF-3 splice variant, denoted inhibitory PAS (IPAS), appears to function as a negative regulator of hypoxiainducible responses [69, 70]. HIF- is not controlled by oxygen levels and is found constitutively expressed in all cell types [71]. By contrast, -subunit levels are under tight control. In response to changing oxygen levels, the control of HIF- subunit expression is achieved by regulating protein level, although other stimuli such as oncogene activation and cytokines can induce both transcription and protein synthesis increases of the subunits [72] Regulation of HIF activity Nuclear levels of HIF- proteins, which are extremely low under normoxic conditions, dramatically increase in response to hypoxia. The presence of HIF- subunits only in conditions of low oxygen tension or after treatment with iron chelators had puzzled researchers until a group of novel oxygen-dependent enzymes essential for regulating HIF- protein levels was discovered. These are the prolyl hydroxylases (PHDs), of which four isoforms (PHD1, PHD2, PHD3 and PHD4) have been identified at this time [73]. The PHD enzymes are members of the 2-oxoglutarate-dependent hydroxylase superfamily. Hydroxylation of two proline residues (402 and/or 564) in the HIF- ODD domain by PHDs serves as a recognition/binding site for the von Hippel-Lindau (pvhl) E3 ubiquitin ligase 14

31 Chapter 1 complex [73, 74]. Binding of pvhl targets HIF- for polyubiquitylation and subsequent degradation by the 26S proteasome [75]. To obtain full transcriptional activity, HIFs must bind to the HRE DNA sequence and recruit transcriptional co-factors. In the presence of oxygen, asparagine 803 in the C terminal TAD gets hydroxylated by a HIF asparaginyl hydroxylase called factor-inhibiting HIF (FIH-1) [76, 77]. This will silence the TAD domain by preventing the binding of transcriptional co-activators CBP/p300. The enzymatic modifications effected by the prolyl and the asparaginyl hydroxylases are dependent on oxygen and iron (Fe2+), explaining the fact that HIF- subunits escape degradation in the absence of oxygen or iron [78-81]. Additional pathways for silencing HIF activity under normoxic conditions include acetylation of lysine 532 in HIF-1 ODD domain by the arrest defective 1 (ARD1) acetylase [82]. This modification increases the interaction with pvhl, resulting in enhanced degradation of HIF-1. The stability and activity of HIF- subunits may also be influenced by reactive oxygen species, hydrogen peroxide as well as by growth factor- and cytokine-induced phosphorylation [56, 83-86]. 15

32 Chapter HIFs as transcriptional regulators 16

33 Chapter 1 When the protein levels of HIF- increase, e.g. in response to hypoxia, it translocates to the nucleus, dimerizes with the subunit and activates the transcription of a number of target genes displaying an HRE motif (Table 1.1). Nuclear localization signal (NLS) domains in the and subunits confer autonomous translocation into the nucleus [87]. One group of HIF-1 target genes is involved in the adaptive response facilitating oxygen delivery to oxygen-deprived tissues. These include the genes encoding erythropoietin, vascular endothelial growth factor-a (VEGF-A) and the inducible NOS (inos) [62]. The erythropoietin (Epo) gene, was discovered as the first true hypoxia-inducible gene in 1992 [53]. EPO stimulates red blood cell production (erythropoiesis), thereby increasing oxygen delivery. Hypoxia also promotes iron uptake and transport by increasing the expression of transferrin and the transferrin receptor [88, 89]. Another well-known hypoxia-regulated gene is Vegf-a, which plays a crucial role in development and growth of blood vessels [90, 91]. One of the VEGF receptors, encoded by the Vegfr-1 gene, is also a direct HIF target, harboring an HRE motif [92]. The Vegfr-2 gene, which at first was reported to lack HIF binding sites, has now been shown to be upregulated by HIF-2 [93]. Hypoxia also affects vascular tone and local blood flow by induction of vasoconstrictors, such as endothelin-1 (ET-1) [94], or by increased expression of genes regulating vasodilation, such as heme oxygenase-1 (HO-1) and inos [95, 96]. Another group of genes upregulated by HIF-1 acts to compensate for the loss of oxygendependent metabolism in hypoxia. The increased expression of various glucose transporters and glycolytic enzymes under hypoxic conditions, allows for oxygen-independent generation of ATP (glycolysis). When oxygen levels fall to a critical point, metabolic switches turn off oxidative phosphorylation and mitochondrial electron transport and instead oxygen-independent or anaerobic energy production (glycolysis) is induced. In the glycolytic pathway, four ATP 17

34 Chapter 1 molecules are produced when glucose is metabolized to two molecules of pyruvate. As two ATP molecules are consumed during this process, this leaves a net yield of two ATP. Compared to aerobic conditions where pyruvate is further oxidized in the Kreb s cycle and the net yield is 31 molecules of ATP, anaerobic glycolysis is much less efficient. In addition to the classic hypoxia-inducible genes, that are direct transcriptional targets of HIF, the response to low oxygen triggers expression of select micro RNAs (mirnas), which in turn down regulate specific genes. MicroRNAs are short non-coding transcripts. A wide set of hypoxia-regulated mirs (HRMs) have been identified. Among them HIF plays an important regulatory role for mir-210, 26 and 181. Studies have revealed a highly complex spectrum of candidate targets of HRMs. These include key genes of the apoptotic pathway such as BID (mir- 23), BIM (mir-24); CASP3 (mir-30), CASP 7 (mir-23), APAF1 (mir-27), BAK1 (mir-26), Bnip3L (mir-23). Conversely, antiapoptotic Bcl2 is a target of mir15 and 16. Another process known to be affected by hypoxia is proliferation, since many cell types undergo cell cycle slowdown or arrest during oxygen deprivation. A multitude of cell cycle genes are HRM targets, a few examples being cdc25a (mir-21, mir-103/107), cyclin D2 (mir-26, mir-103/107), cyclin E1 (mir-26), cyclin H (mir-23), cdk6 (mir-26, mir-103/107) [97-99]. HRMs mir-16, mir-20, let-7b, mir-17-5p, mir-27, mir-106, mir-107, mir-193, mir-210, mir-320 and mir- 361 have been shown to target VEGF [100]. Table 1.1 Hypoxia-inducible genes harboring HRE sequences [62]. 18

35 Chapter 1 Gene Oxygen supply 1B -adrenergic receptor Adrenomedullin Atrial natriuretic peptide (ANP) Breast cancer resistance protein (BCRP) Endothelial nitric oxide synthase (enos) Endothelin-1 Erythropoietin Ferrochelatase Heme oxygenase 1 Inducible nitric oxide synthase (inos) Leptin Transferrin Transferrin receptor Plasminogen activator inhibitor-1 (PAI-1) Vascular endothelial growth factor-a (VEGF-A) VEGF-D VEGF receptor-1 (VEGFR-1) VEGFR-2 Function Vessel diameter Vessel diameter Blood volume Heme binding Vessel diameter Vessel diameter Erythropoiesis Heme synthesis Vessel diameter Vessel diameter Metabolism/ Angiogenesis Iron transport Iron transport Blood flow Angiogenesis Angiogenesis Angiogenesis Angiogenesis Cellular metabolism Aldolase A Glycolysis Carbonic anhydrase-9 (CA-9) ph regulation Cytochrome P450 2C11 (CYP2C11) Metabolism CYP3A6 Metabolism CYP4B Eicosanoid synthesis Enolase 1 Glycolysis Glucose transporter 1 (Glut1) Glucose uptake Glucokinase Glycolysis Glutathione peroxidase-3 (GPx-3) Glutathione peroxidase Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) Glycolysis Lactate dehydrogenase A Glycolysis Multidrug resistance gene 1 (MDR1) Xenobiotic transporter Phosphoenolpyruvate carboxykinase (PEPCK) Gluconeogenesis Phosphofructokinase L (PFKL) Glycolysis 6-Phosphofructo-2-kinase/fructose-2,6-biphosphatase-3 Glycolysis 6-Phosphofructo-2-kinase/fructose-2,6-biphosphatase-4 Glycolysis Phosphoglycerate kinase 1 (PGK1) Glycolysis Table 1.1 Hypoxia-inducible genes harboring HRE sequences (contd) 19

36 Chapter 1 Gene Cell growth and metabolism Connective tissue growth factor (CTGF) Ecto-5 -nucleotidase (CD73) Endoglin Insulin growth factor binding protein-1 (IGFBP-1) Intestinal trefoil factor Intestinal barrier function Transforming growth factor- 3 (TGF- 3) Function Growth factor Intestinal barrier function TGF- coreceptor Growth factor Intestinal barrier function Placenta development Cell growth and apoptosis CXCR4 Bcl-2/E1B 19kDa interacting protein (BNip3) Met Myeloid cell factor-1 (Mcl-1) Nip3 Noxa Nucleophosmin Nur77 Serine/threonine protein phosphatase 5 (PP5) Stromal cell-derived factor-1 (SDF-1 or CXCL12) Telomerase reverse transcriptase (TERT) Wilms tumor suppressor (Wt1) Chemokine receptor Pro-apoptotic Proto-oncogene Anti-apoptotic Pro-apoptotic Pro-apoptotic p53 inhibition Orphan steroid receptor Anti-apoptotic Chemokine Telomere extension Tumor suppressor gene Others CD18 Leukocyte adhesion Cited2/p35srj Transcription cofactor Collagen prolyl 4-hydroxylase I Hydroxylase DEC1 and DEC2 Transcription factors Ets-1 Transcription factors Furin Pro-protein convertase Glucose-regulated protein 94 (GRP94) Chaperone Inhibitor of differentiation/dna binding protein 2 (ID2) Transcriptional repressor Membrane type-1 matrix metalloproteinase (MT-1) Matrix metalloproteinase Prolyl hydroxylase domain protein 2 and 3 (PHD2/PHD3) Oxygen sensing Retrotransposon VL30 Retrotransposon 20

37 Chapter 1 Besides prolyl hydoxylases, stabilization and/or synthesis of HIF-1 under hypoxia is also dependent on the activity of the PI-3 kinase/akt pathway [84]. PI-3K inhibitors inhibits the accumulation of HIF-1 in these conditions, while dominant negative mutants for PI-3K or for Akt decrease the hypoxia-induced overexpression of VEGF. Conversely, disruption of PTEN, a phosphatidylinositol triphosphate phosphatase that inactivates Akt, leads to increased VEGF expression in normoxic cells [101]. Finally, growth factor- or cytokine-induced activation of HIF-1 in normoxia results from an increased synthesis of HIF-1 which is also dependent on the PI3K/Akt pathway. It remains unclear, however, how the PI-3K/Akt pathway interacts with the prolyl hydroxylase-pvhl system to regulate HIF-1 protein level. Post stabilization, redox status, dissociation from the chaperone hsp90, association with co-activators like CBP/p300 or SRC-1 as well as phosphorylation are also required for full transcriptional activity (36, 37). Hypoxia directly regulates the association of HIF-1 with the coactivator CBP/p300. Similar to prolyl hydroxylase, an asparagyl hydroxylase, whose activity strictly depends on the presence of oxygen, hydroxylates HIF-1 carboxyl-terminal transactivation domain on Asn 803. This modification prevents its association with CBP/p300 under normoxic conditions [76, 77] HIF independent transcriptional activation Although HIF-1 is a pivotal regulator of transcription in hypoxia, other transcription factors induced in response to hypoxia include the early growth response protein (Egr-1), Nuclear factor- B (NF- B) and activator-protein 1 (AP-1). The early growth response protein (Egr-1) is a zinc finger nuclear phosphoprotein that is induced and activated within minutes of oxygen deprivation [50]. Induction of Egr-1 DNA binding activity leads to activation of tissue 21

38 Chapter 1 factor gene transcription. In Egr-1 null mice, expression of tissue factor and intravascular fibrin deposition were severely decreased after hypoxia. This general effect of Egr-1 activation extends to the tissue factor (TF), VEGF, plasminogen-activator inhibitor (PAI), intracellular adhesion molecule (ICAM), as well as several interleukins. The Egr-1 and HIF-1 pathways appear to be initiated independently of each other, indicative of the separate role they each play in inducing different facets of the adaptive response to hypoxia [102]. The p50-p60 heterodimer of Nuclear factor- B (NF- B) is induced both under hypoxia and following re-oxygenation. To date the role of NF- B in mediating induction of hypoxiaresponsive genes is poorly understood. Studies by Koong et al and Imbert et al have suggested hypoxia-induced activation of NF- B to occur via a mechanism involving tyrosine phosphorylation of the upstream inhibitory subunit I B. One of the genes thought to be regulated by NF- B in hypoxia is cyclooxygenase-2 (COX-2), which is induced in human vascular endothelial cells by the binding of p65 to the NF- B consensus element in the COX-2 promoter [ ]. The activator-protein 1 (AP-1) is a dimeric transcription factor comprising subunits from the jun and fos multigene families. The DNA binding and transcriptional activity of AP-1 has been demonstrated to be strongly induced by hypoxia [107]. A single cysteine residue in the DNA-binding domain of fos and jun is responsible for this redox sensing and signaling. Activation of VEGF [108], tyrosine hydroxylase [109], collagenase IV [110], endothelin-1 [111] and c-jun [112] has been shown to correlate with the activation of AP-1. The hypoxic regulation of AP-1 may be functionally distinct from other AP-1-inducing stresses, as it was shown that ectopically expressed c-jun functionally cooperates with HIF-1 to regulate HRE-dependent 22

39 Chapter 1 reporter expression without binding to AP-1. Furthermore, induction of c-jun mrna expression and phosphorylation by prolonged hypoxia was dependent on HIF-1 [113]. 1.7 Hypoxic regulation of mrna stability Both transcriptional activation and post-transcriptional mechanisms contribute to the hypoxia-mediated regulation of gene expression. Hypoxic regulation of cis-acting regulatory elements found at the 3 -UTR of the hypoxia-responsive mrnas are responsible for hypoxiaspecific message stabilization [114] and the most commonly described cis-acting sequences include the AU-rich element (ARE), stem-loop element and pyrimidine-rich element. Hypoxia has been known to specifically increase stability of the mrnas of vascular endothelial growth factor (VEGF), tyrosine hydroxylase (TH), glucose transporter (GLUT-1) and erythropoietin (EPO) [115]. Hypoxic induction of VEGF occurs in a biphasic manner: initial activation via transcriptional induction is followed by augmented mrna stability mediated by binding of specific proteins such as the heterogeneous nucleoprotein L (hnrp L) and the RNA binding protein, HuR [ ]. Likewise, as shown in pheochromocytoma-derived PC12 cells, the hypoxia-dependent stabilization of the TH gene mrna is due to a hypoxia-inducible protein binding site (HIPBS) in the 3 -UTR of the TH mrna [119]. 1.8 Hypoxic repression of transcription In contrast to inducing the expression of specific genes, hypoxia can also result in specific gene repression. Several proteins have been reported to contribute to transcriptional repression in hypoxic cells. These include negative cofactor 2 (NC2), differentiated embryo chondrocyte 1 (Dec 1), histone deacetylases (HDAC), msin3a, and p53. 23

40 Chapter 1 NC2, a transcriptional repressor, activated in extracts from hypoxia-treated hepatoma cells inhibits the formation of the pre-initiation complex through direct interaction with the TATA binding protein (TBP). This interaction prevents access to the promoter by TFIIB, thus preventing the formation of the RNA polymerase holoenzyme, and thereby blocks transcription [120, 121]. Because of the large number of TATA-containing genes in the human genome, the activation of NC2 may contribute to a global repression of transcription during hypoxia. A more specific hypoxia-dependent transcriptional repressor is Dec1 (Stra13, Sharp2). Dec1 is a member of the basic helix loop helix family of transcription factors and has been demonstrated to repress expression through binding to E-box elements [122]. Hypoxia can prevent differentiation or cause dedifferentiation in a number of cell types including adipocytes, breast carcinomas and neuroblastomas [123, 124]. Hypoxic induction of Dec1 has been shown to be able to block the expression of PPAR 2 in pre-adipocytes, blocking their differentiation [125]. In addition to DNA binding by direct acting repressors, gene expression can be down regulated by corepressors such as histone deacetylases (HDACs) that act to modify the local chromatin. HDACs function as co-repressors through their association with transcription factors, such as p53, which can recruit and target them to specific genes. Hypoxia has been shown to elevate HDAC activity [126] in addition to increased interactions with factors like p53. p53 protein has been shown to be stabilized under severe hypoxia, but the protein does not activate transcription of its typical target genes [127]. However, in addition to its transactivation property, p53 can also repress target genes, and this activity is retained under hypoxia [128]. To repress gene transcription, p53 selectively interacts with its known transcriptional co-repressors msin3a and HDAC1 in hypoxic cells [129]. In addition, studies show that hypoxic induction of this 24

41 Chapter 1 complex containing p53, HDAC1 and msin3a represses genes such as stathmin and Map4, proteins which play a role in microtubule organization and ultimately in G 2 /M phase growth arrest [ ]. 1.9 Hypoxic control of protein translation Regulation of gene expression by hypoxia may occur at a post-transcriptional level. Phosphorylation of eif2 by the endoplasmic reticulum associated kinase PERK, during severe hypoxia, results in a global reduction in protein synthesis. The activation of the PERK kinase is a recognized response to ER stress, and the blocking of new protein synthesis is a means of reducing that stress. Koumenis et al also shows that cells deficient in this response have also been shown to be more sensitive to hypoxia-induced toxicity [131, 132]. Mammalian cells respond to wide ranges of oxygen concentration through alterations in both metabolic states and growth rates. Hypoxia alters cellular proliferation by regulating the cell cycle as well as by programmed cell death or apoptosis. The following two sections (1.10 and 1.11) give an overview of the cell cycle and apoptosis and outlines the effects of hypoxia on each of these processes Cell cycle and hypoxia Overview of mammalian cell cycle The mammalian cell cycle consists of four distinct phases: G 1, S (synthesis), G 2 (collectively known as interphase) and M (mitosis). In the interphase, the cell grows, accumulating nutrients needed for mitosis and duplicating its DNA and in the M phase, the cell 25

42 Chapter 1 splits itself into two daughter cells. M phase is itself composed of two tightly coupled processes: mitosis, in which the cell's chromosomes are divided between the two daughter cells, and cytokinesis, in which the cell's cytoplasm divides forming distinct cells. Activation of each phase is dependent on the proper progression and completion of the previous one. Cells that have temporarily or reversibly stopped dividing are said to have entered a state of quiescence called G 0 phase. The first phase within interphase, from the end of the previous M phase until the beginning of DNA synthesis is called G 1 (G indicating gap). This phase is marked by synthesis of various enzymes that are required in S phase, mainly those needed for DNA replication. In the S phase the chromosomes are replicated with each chromosome having two (sister) chromatids. Rates of RNA transcription and protein synthesis are low during this phase, except for histone production, most of which occurs during the S phase [133]. The cells then enter the G 2 phase, which lasts until the cell enters mitosis. Significant protein synthesis occurs during this phase, mainly involving the production of microtubules, required during mitosis. Inhibition of protein synthesis during G 2 phase prevents the cell from undergoing mitosis. The M phase has been broken down into several distinct phases, sequentially known as prophase, prometaphase, metaphase, anaphase and telophase leading to cytokinesis. Control mechanisms ensuring the fidelity of cell division are called the checkpoints. These verify whether the processes at each phase of the cell cycle have been accurately completed before progression into the next phase. DNA damage checkpoints sense DNA damage both before the cell enters S phase (a G 1 checkpoint) as well as after S phase (a G 2 checkpoint). Damage to DNA before the cell enters S phase inhibits the action of CDK2 thus stopping the progression of the cell cycle until the damage can be repaired. In case of irreparable damage, the 26

43 Chapter 1 cell self-destructs by apoptosis. Damage to DNA after S phase (the G 2 checkpoint), inhibits the action of Cdk1 thus preventing the cell from proceeding from G 2 to mitosis. The first checkpoint is located before entry into S phase, making the key decision of whether the cell should divide, delay division, or enter a resting stage. The G 1 checkpoint (restriction point) is where eukaryotes typically arrest the cell cycle if environmental conditions make cell division impossible [134]. The restriction point is mainly controlled by action of the CKI- p16 (CDK inhibitor p16). This protein inhibits CDK4/6 and ensures that it can no longer interact with cyclin D1 to cause cell cycle progression. The second checkpoint is located at the end of G 2 phase, triggering the start of the M phase (mitosis). The CDKs associated with this checkpoint are phosphorylated by the "Maturation promoting factor" (or Mitosis Promoting Factor, MPF). The MPF activates the CDK in response to environmental conditions being right for the cell and allows the cell to begin DNA replication. An activating phosphatase, Cdc25, under favourable conditions removes the inhibitory phosphates present within the MPF complex. However, DNA is frequently damaged prior to mitosis, and to prevent transmission of this damage to daughter cells, the cell cycle is arrested via inactivation of the Cdc25 phosphatase (via phosphorylation with other protein kinases). There are also spindle checkpoints that detect any failure of spindle fibers to attach to kinetochores and arrest the cell in metaphase until all the kinetochores are attached correctly (M checkpoint). In differentiated mammalian cells, G1 to S progression is regulated by the hypophosphorylated Rb gene or its related proteins, p107 and p130, which inhibit the expression of genes required for entry into S phase by sequestering the E2F family of transcription factors. During G1 phase the Rb/HDAC repressor complex binds to the E2F-DP1 transcription factors inhibiting downstream transcription. Eukaryotic cell cycle progression is dependent, in part, on 27

44 Chapter 1 the tightly regulated activity of CDKs. CDK4/CDK6 and Cdk2 whose regulatory partners are the D-type cyclins (D1, D2 and D3) and cyclin E, respectively, represent two different classes of G1- specific CDKs whose activation is required for entry into S phase. Cyclin D/CDK4 CDK6 activity occurs in mid-late G1 phase, upstream of CDK2/cyclin E activity. The mitogenic activity of CDKs are inhibited by cell cycle inhibitory proteins, including p15 (INK4B), p16 (INK4A), p18 (INK4C), p19 (INK4D), p21 (CIP), p27 (Kip1) and p57 (Kip2). Studies have suggested that cyclin D/CDK complexes also play a second non-catalytic role in G1 progression by sequestering proteins of the Cip/Kip family, including p27 (Kip1) and p21(cip1), two potent inhibitors of CDK2 [135]. Binding of Cip/Kip proteins to cyclin D1/CDK4 stabilizes the complex and facilitates its nuclear import [136]. Mitogen withdrawal results in the disassembly of the cyclin D/CDKs and in addition mobilizes the latent pool of p27kip1, which blocks the activity of cyclin E/CDK2 and facilitates cell cycle exit. Murine embryonic fibroblasts (MEFs) lacking p27 and p21 do not express D-type cyclins and have a significant reduction in CDK activity, but continue to proliferate normally, suggesting that D- type cyclins might not be essential for cell cycle progression, at least in a setting where Cip/Kip proteins are absent [136]. Further studies have shown that activation of the cyclin D1/CDK4 complex occurs when quiescent p21/p27-null MEFs are stimulated to re-enter the cell cycle. In addition, the ectopic expression of p34 SEI-1, a mitogen-induced CDK4 activator, increased the levels of active cyclin D1/CDK4 complex in the absence of p21 and p27, suggesting that there are several independent pathways to stimulate the assembly of the cyclin D1/CDK4 complex [137]. More recent studies have highlighted the role of an additional cell cycle regulatory mechanism at the G1 to S transition that is able to govern the initiation of histone gene expression needed for packaging of newly replicated DNA [138]. This is commonly referred as 28

45 Chapter 1 the S point and is initiated by cyclin E/Cdk2-dependent phosphorylation of p220 NPAT and the formation of a functional HiNF-p220 NPAT complex that controls H4 gene transcription. The expression of cell cycle inhibitory protein, p21, is tightly controlled by the tumor suppressor protein p53, through which this protein mediates the p53-dependent cell cycle G1 phase arrest in response to a variety of stress stimuli. p21 can interact with proliferating cell nuclear antigen, a DNA polymerase accessory factor, and plays a regulatory role in S phase DNA replication and DNA damage repair [139]. p21 has been reported to be specifically cleaved by Casp-3 like caspases, leading to apoptosis [140]. p53 is a transcription factor which in humans is encoded by the TP53 gene. The three main functions of p53 include (a) activation of DNA repair proteins when DNA has sustained damage; (b) induction of growth arrest at G1/S by activation of p21 expression, to allow DNA repair proteins time to fix the damage; and (c) intiation of apoptosis if the DNA damage proves to be irreparable. p53 becomes activated in response to a variety of stress signals including hypoxia. The half-life of p53 is increased causing p53 accumulation in stressed cells. Also a conformational change causes p53 to function as a transcriptional regulator in these cells. Phosphorylation of the p53 N-terminal domain makes it a traget of two groups of protein kinases, namely the MAPK family (JNK1-3, ERK1-2, p38 MAPK), and the ATR, ATM, CHK1 and 2 kinases which are implicated in the genome integrity checkpoint. Oncogenes also stimulate p53 activation, mediated by p14arf. In unstressed cells, p53 levels are kept low through its continuous degradation. Mdm2 binds to p53, preventing its action and transports it from the nucleus to the cytosol. Also Mdm2 acts as ubiquitin ligase and covalently attaches ubiquitin to p53 leading to its proteasomal degradation. This is reversible and a ubiquitin specific protease, USP7, can cleave ubiquitin off p53, preventing its proteasome-dependent degradation. 29

46 Chapter 1 Phosphorylation of the N-terminal end of p53 disrupts Mdm2-binding. Other proteins, such as Pin1, are then recruited to p53 and induce a conformational change in p53 which prevents Mdm2-binding even more. Phosphorylation also allows for binding of transcriptional coactivators, like p300, which then acetylate the carboxy terminal end of p53, exposing the DNA binding domain of p53, allowing it to activate or repress specific genes. Deacetylase enzymes, such as Sirt1 and Sirt7, can deacetylate p53, leading to inhibition of apoptosis [141]. Figure 1.5 Cell cycle phases and G1/S transition. Source: Herrup K (2007) Cell cycle regulation Effects of hypoxia on cell cycle When cells are exposed to severe hypoxia cell cycle progression and DNA synthesis rapidly cease. The induction of HIF-1 activation prevents G 1 /S transition through the action of CKIs and the regulation of cyclin E expression [142, 143]. Expression of p21 and p27 is increased transcriptionally in a HIF-1-dependent manner [142, 144]. Sustained expression of 30

47 Chapter 1 these CKIs is observed in wild-type cells, but not in HIF-1 null cells. These CKIs suppress cyclin/cdk2 activity, and thus reduce the ratio of phosphorylated to dephosphorylated Rb protein, resulting in cell cycle arrest at the G 1 /S interface [145, 146]. HIF-1 may also regulate cyclin E protein levels; CCNE binds to CDK2 and modulates its kinase activity dependent upon cell cycle phase [146]. It has been reported that hypoxic cells lacking HIF-1 displayed enhanced and sustained accumulation of cyclin E, without any effect on CDK2 protein expression, relative to wild-type cells. In accordance with changes in cyclin E expression, cyclin E/CDK2 kinase activity in HIF-1 -deficient cells was also increased, resulting in somewhat retarded, but still substantial, cell growth, even under hypoxia. Hypoxia causes an increase in the CDKN1A mrna in a p53-independent manner [127]. The number of cells in G 1 phase in p53 null cultures is increased relative to wild-type cultures. This change may be attributable to enhanced HIF-1 activity by inactivation of p53, rather than the direct action of p53, because expression and transcriptional activity of p53 change little under hypoxia. HIF-1 null, p53 wild-type cells do not show any hypoxia-induced G 1 arrest. Rather, S-phase entry is accelerated, indicating that HIF-1, but not p53, plays an essential role in the regulation of cell cycle progression under hypoxia [147, 148]. Cells lacking functional copies of both p53 and HIF-1 have been shown to display no change in the proportion of cells entering S- phase, as was seen in HIF-1 null cells. These cells appear to lose the ability to sense and respond to hypoxia. Collectively, these data strongly suggest that both transcription factors, HIF- 1 and p53, cooperate to regulate the cell cycle progression through distinct mechanisms, but HIF- 1 serves as the primary determinant for cell cycle regulation under hypoxia. Previous studies have indicated that hypoxia-induced cell cycle arrest is accompanied by a decreased activity of CDKs and Rb protein, leading to inhibition of cell cycle progression. Also cyclin G2, a negative 31

48 Chapter 1 regulator of cell cycle progression via binding with protein phosphatase 2A in certain cell types, is induced by hypoxia through HIF-1 activation [149, 150]. Hypoxia-induced S phase-dependent arrest is mediated by a rapid shutdown of DNA synthesis through a block to replicon initiation [151]. This block persists as long as the cells are held hypoxic, and is signaled through the ATR kinase [152, 153] Apoptosis and hypoxia Apoptotic pathways Cells can activate an intracellular death program and commit suicide in a controlled way, a process known as apoptosis. Programmed cell death (apoptosis) was first described in 1972 by Currie and colleagues [154]. Apoptotic cell death is important for the maintenance of tissue homeostasis under physiological conditions as well as for pathogenesis during disease states including myocardial infarct, neurodegenerative disorders, autoimmune diseases, and cancer [155, 156]. Alternately, cells can die by an uncontrolled process known as necrosis. Apoptosis can be induced by a variety of factors, including ligand activation of death receptors, growth factor deprivation and hypoxia. Characteristics of apoptosis include chromatin condensation, membrane blebbing, phosphatidylserine exposure, cytoplasmic shrinkage, formation of apoptotic bodies, and DNA fragmentation. The apoptosis pathway is dependent upon caspase activation. Caspases comprise an expanding family of cysteine proteases that exist as inactive pro-enzymes in viable cells [157]. Activated caspases acquire the ability to cleave key intracellular substrates as well as activate other caspases, resulting in the induction of a protease cascade that can kill the cell. Caspase activation is an ATP dependent process and is sufficient to induce all of the morphological features of apoptosis. In contrast, necrosis does not involve the 32

49 Chapter 1 activation of caspases and is not an energy dependent process [158]. Characteristics of necrosis include organelle swelling and cell bursting, leading to an inflammatory response. This inflammatory response does not occur under apoptotic conditions since apoptotic cells display phagocytosis markers and are engulfed by neighboring cells [159]. There are two possible mechanisms of apoptosis - intrinsic and extrinsic [160]. The critical regulators of the intrinsic pathway are the Bcl-2 family members [161]. The family can be divided into three different groups based on Bcl-2 homology (BH) domains and function. The anti-apoptotic members, such as Bcl-2 and Bcl-XL, typically have BH1 through BH4 domains. The pro-apoptotic members can be divided into two groups. The first group consists of proteins such as Bax and Bak that contain BH1, BH2 and BH3 domains. The second group consists of proteins such as Bad and Bim that contain only BH3 domains. The BH domains have functional and structural significance. Many members of this family, such as Bcl-2 and Bcl-XL, are predominantly localized to the outer membrane of mitochondria, while others interact with mitochondria indirectly. In response to a variety of apoptotic stimuli, pro-apoptotic Bcl-2 family members (such as Bax or Bak) initiate the mitochondrial dependent apoptotic pathway by causing a loss of outer mitochondrial membrane integrity [162]. This releases apoptogenic proteins located in the intermembrane space of mitochondria, such as cytochrome c, Smac/Diablo, and apoptosis inducing factor (AIF) into the cytosol [163, 164]. Cytochrome c, an electron carrier within the respiratory chain, interacts directly with Apaf-1 in the cytoplasm leading to the ATP dependent formation of a macromolecular complex known as the apoptosome [163, 164]. This complex recruits and activates the aspartyl directed protease caspase-9. Activated caspase-9 can activate additional caspase-9 molecules, as well as the downstream caspases such as caspase-3 or -7, resulting in morphological features of apoptosis. 33

50 Chapter 1 Smac/DIABLO, another mitochondrial protein released into the cytosol in response to apoptotic stimuli, promotes caspase activation by eliminating inhibitory of apoptosis protein (IAP) function [165]. AIF induces a caspase independent cell death and is critical for developmental apoptosis [166]. Anti-apoptotic members Bcl-2 and Bcl-XL inhibit mitochondrial dependent apoptosis by preventing Bax or Bak from disrupting the integrity of the outer mitochondrial membrane. Previous studies have shown that DNA damaging agents, serum deprivation, and endoplasmic reticulum stress agents trigger apoptosis through the mitochondrial dependent pathway. Fibroblasts from embryos of mice lacking either Bax and Bak genes or cells that over express BcL-XL or Bcl-2 are resistant to these apoptotic agents [167]. The mechanisms by which these apoptotic stimuli converge on Bax or Bak to activate mitochondrial dependent apoptosis remain unknown. The extrinsic pathway is initiated when a death ligand, such as FasL or TNF, interacts with its cell surface receptor, Fas (CD95) or TNF receptor (TNFR1/2) [168]. This results in the formation of a death-inducing signaling complex (DISC). The formation of DISC involves adaptor proteins such as FADD (Fas-associating protein with death domain) or TRADD (TNF receptor associating death domain) [169, 170]. These proteins are involved in the recruitment of pro-caspase-8 and its subsequent proteolytic activation. A variety of cell types undergoing apoptosis through this pathway show strong activation of caspase-8 and direct activation of caspase-3 [171]. In contrast, other cell types initially display a weak activation of caspase-8, which subsequently employs the mitochondria for amplification of the death signal. This process occurs by the caspase-8 dependent cleavage of Bid, a pro-apoptotic factor [172, 173]. A truncated Bid requires either Bax or Bak to induce the loss of outer mitochondrial membrane 34

51 Chapter 1 integrity leading to cytochrome c release and caspase-9 activation [174]. Thus, there is cross talk between the extrinsic and intrinsic pathways through truncated Bid. Several studies indicate that oxygen deprivation can induce apoptosis in a variety of cell types. As long as cells have an adequate supply of ATP during oxygen deprivation, apoptosis can be executed [175]. However, if cells are deprived of oxygen and glucose then cells undergo necrosis. The requirement for ATP to execute apoptosis during oxygen deprivation is attributed to energy dependent activation of caspases. Cells over-expressing the anti-apoptotic proteins Bcl- 2 or BcL-XL have been shown to prevent oxygen deprivation induced apoptosis by inhibiting the release of cytochrome c from the mitochondria [ ]. Fibroblasts from mice lacking both Bax and Bak genes are resistant to oxygen deprivation induced apoptosis [175]. Furthermore, the pro-apoptotic protein Bax translocates from the cytosol to the mitochondria during oxygen deprivation [179]. Cytochrome c is released and caspase-9 is activated in oxygen-deprived cells undergoing apoptosis. Cytochrome c is released independent of caspase activation since cytochrome c is still released in the presence of the caspase inhibitor zvad. Fibroblasts from caspase-9 or Apaf-1 deficient mice transformed with c-myc and H-ras are resistant to cell death during oxygen deprivation [180]. Consistent with this, Bid null fibroblasts are able to undergo apoptosis in response to oxygen deprivation indicating that the extrinsic pathway does not contribute to oxygen deprivation induced apoptosis [181]. In Jurkat cell lines, hypoxia-induced apoptosis was not affected by lack of caspase-8 or FADD, whereas overexpression of Bcl-2 or expression of dominant-negative caspase-9 mutant rendered the cells resistant to hypoxiainduced apoptosis [182]. Together, these results suggest that hypoxia-induced apoptosis mainly relies on intrinsic, mitochondrial pathways. 35

52 Chapter 1 The mitochondria are the central organelle in the intrinsic pathway. In some circumstances, however, the endoplasmic reticulum (ER) or sarcoplasmic reticulum in muscle cells plays an important role in the hypoxia-induced mitochondrial death pathway, as well as mediating cell death independently of mitochondria. Although the mechanisms by which the ER brings about cell death are poorly understood, increases in intracellular Ca 2+ appear to be central. ER Ca 2+ stores are thought to be increased by Bax and Bak, which are located at ER, as well as mitochondrial membranes [183, 184]. Increased ER Ca 2+ facilitates a more robust release of Ca 2+ into the cytoplasm on delivery of an apoptotic stimulus, and may activate several apoptotic mechanisms. First, mitochondrial Ca 2+ overload can trigger mitochondrial permeability transition pore (MPTP) opening and cytochrome c release [185]. Cytochrome c binds the inositol 1,4,5- trisphosphate (IP3) receptor, one of the ER Ca 2+ release channels, to further stimulate Ca 2+ release [186]. Second, increased intracellular Ca 2+ can activate calpain. Calpain can cleave Bid, providing another mechanism for cytochrome c release. Calpain activation also causes cleavage of procaspase-12 [187]. Caspase-12 has been shown in knockout mice to be required for apoptosis induced specifically by ER stress [188]. Cleaved caspase-12 translocates to the cytoplasm and activates caspase-9 independently of apoptosome formation [189, 190]. These events provide a mitochondria-independent mechanism for ER-mediated apoptosis. Some signals that activate the ER death pathways originate within this organelle itself, where a complex array of pathways mediate the unfolded protein and other ER stress responses [191]. In addition, given their roles in carrying upstream apoptotic stimuli to Bax and Bak at the mitochondria, BH3-only proteins would be anticipated to perform an analogous function in the ER pathway. BH3-only proteins Bik (Bcl-2 interacting killer) and Puma have been implicated in the ER death pathway [192]. It remains unclear, however, whether these proteins function to relay signals from the 36

53 Chapter 1 periphery to the ER and/or from the ER to mitochondria. However, upstream signals originating in the extrinsic pathway are known to be linked with the ER by Bap31 (B-cell receptor associated protein 31), an integral ER membrane protein that is cleaved by caspase-8 resulting in ER Ca 2+ release [193]. The apoptotic pathways described above have been schematically represented below. Figure 1.6 Pathways of apoptosis. Source: Gupta S et al. (2006) Lessons learned from apoptosis. 37

54 Chapter Regulation of apoptosis during hypoxia Both pro-and anti-apoptotic genes have been shown to be regulated by hypoxia. The molecular pathways triggering the apoptotic response to hypoxia are far from completely understood. Both p53 and Bcl-2 family members are involved in hypoxic activation of apoptosis. The PI-3 kinase pathway plays an important role during oxygen deprivation induced cell death. Electron transport inhibition also regulates hypoxia induced apoptosis Role of p53 in hypoxia-induced apoptosis The transcription factor p53 has been implicated in regulating oxygen-deprivation induced apoptosis. p53 can induce the expression of apoptotic genes such as Bax, NOXA, PUMA and PERP [194]. Oxygen deprivation leads to p53 protein stabilization [148]. Hypoxia causes p53 interaction with transcriptional repressor msin3a but not with the transcriptional activator p300 [128]. Also hypoxia causes localization of p53 to the surface of the mitochondria [195]. In addition, severe hypoxia causes p53 accumulation by down-regulating its negative regulator mdm2 and activation by post-translational modifications [196]. Hansson and colleagues demonstrated that the DNA-binding domain of p53 binds two specific motifs adjacent to and within the ODD domain of HIF-1, irrespective of HIF-1 hydroxylation status [197]. Moreover, HIF-1 null cells accumulate more p53 protein than wild-type cells at low oxygen concentrations, suggesting the expression of HIF-1 to cause p53 protein accumulation in hypoxic cells, and thus induce apoptosis through p53 activation [198]. The hypoxia-dependent post-transcriptional phosphorylation of p53 is by the ATR kinase, presumably due to replication arrest that occurs under hypoxia. Reoxygenation and the consequent DNA damage can then activate the ATM kinase which maintains p53 phosphorylation [152]. Hypoxic stress results in 38

55 Chapter 1 the association of p53 with transcriptional co-repressors and not co-activators capable of repressing anti-apoptotic molecules such as stathmin [128] Role of Bcl-2 family proteins in hypoxia-induced apoptosis Pro- and anti-apoptotic members of the Bcl-2 family play a role in hypoxia-induced apoptosis [175, 181]. BH3-only proapoptotic proteins, PUMA and Noxa, and BNIP3 and BNIP3L have emerged as potential initiators of apoptosis by low oxygen concentrations [ ]. HIF-1 has been shown to induce the expression of BNIP3 (formerly called Nip3), a proapoptotic member of the Bcl-2 family. BNIP3 heterodimerizes with Bcl-2/BcL-XL at both the mitochondrial and non-mitochondrial sites [202]. Removal of the BH3 domain does not inhibit apoptotic activity of BNIP3; instead, the transmembrane domain is critical for its function/activity. A direct role for HIF-1 in regulating sensitivity to oxygen deprivation induced apoptosis come from genetic studies using embryonic stem cells with HIF-1 deleted. HIF-1 null cells show a decrease in apoptosis compared to wild-type cells during oxygen deprivation [203]. Wild-type cells exhibit a decrease in Bcl-2 protein levels and an increase in p53 levels. In contrast, HIF-1 null cells display no changes in p53 or Bcl-2 protein levels during oxygen deprivation. Also, pancreatic cancer cells with constitutive expression of HIF-1 are resistant to apoptosis induced by oxygen deprivation when compared to cells without constitutive expression of HIF-1 [204]. Murine hepatoma cell lines without a functional HIF-1 display no difference in sensitivity to oxygen deprivation induced apoptosis. Thus, HIF-1 can have various effects on oxygen deprivation induced apoptosis depending on the cell type. Other anti-apoptotic proteins regulated by hypoxia include (i) ORP150, an ER associated heat shock protein, whose overexpression has been shown to prevent the release of cytochrome c 39

56 Chapter 1 and cell death in neurons deprived of oxygen. The precise mechanism by which ORP-150 inhibits cytochrome c release remains unknown. Presumably, ORP-150 prevents activation of pro-apoptotic Bcl-2 family members [205]; (ii) IAP-2, which inhibits caspase activity IAP-2 is induced in response to oxygen deprivation in a HIF-1 independent mechanism [206, 207] and (iii) RTP801, a recently identified and cloned HIF-1 target gene. Hypoxia induces RTP801 in a HIF-1 dependent manner. Over expression of RTP801 under normal oxygen conditions is toxic in neuron-like PC12 cells while it is protective against oxygen deprivation induced cell death in dividing PC12 cells and MCF-7 cells [208] Role of PI3-kinase pathway in hypoxia-induced apoptosis The phosphoinositide-3 kinase (PI3K)/Akt pathway is a potent mediator of cell survival signals [209]. Activated PI3K phosphorylates inositol phospholipids to phosphatidylinositol (3,4,5)-triphosphate (PIP-3). The increase in PIP-3 at the plasma membrane recruits Akt via its pleckstrin homology (PH) domain. Akt, activated upon phosphorylation, in turn, can phosphorylate and inactivate several substrates including the pro-apoptotic Bcl-2 family member Bad, the forkhead family transcription factor FKHRL1 and caspase-9 [ ]. Akt can also mediate cell survival through hexokinase, an enzyme involved in the first committed step in glycolysis [213]. Glycolysis utilizes glucose in the anaerobic production of ATP, which is necessary for cell survival. Both hexokinase II and the pro-apoptotic protein Bax can bind to mitochondria at the site of voltage dependent anion channel (VDAC). Binding of hexokinase to the VDAC can inhibit or block Bax binding, thereby preventing cytochrome c release and subsequent activation of apoptosis [214]. The tumor suppressor PTEN is involved in the negative regulation of the PI3K/Akt pathway [215]. PTEN is a dual specificity phosphatase, which is capable of dephosphorylating inositol phospholipids. Thus, PTEN activation results in decreased 40

57 Chapter 1 levels of PIP-3, leading to decreased Akt activity and increased apoptosis. PTEN null mouse embryonic fibroblasts exhibit constitutively elevated activity of Akt and display decreased sensitivity to cell death in response to a number of apoptotic stimuli, including UV irradiation [216]. In cardiac myocytes adenoviral gene transfer of activated Akt protects against oxygen deprivation induced apoptosis in vitro [217]. In contrast, over expression of activated Akt in Rat1a fibroblasts did not protect cells from oxygen deprivation induced apoptosis [175]. Other studies have shown that oxygen deprivation leads to Akt activation in PTEN null glioblastoma cell lines [218]. Expression of wild-type PTEN in these cell lines prevented Akt activation in response to oxygen deprivation. However, expression of PTEN at highly elevated levels did not alter sensitivity to oxygen deprivation induced apoptosis in PTEN null glioblastoma cell lines. Thus, the ability of Akt to prevent cell death during oxygen deprivation might be restricted to cell type Role of electron transport inhibition in hypoxia-induced apoptosis Electron transfer through the respiratory chain is coupled to the directional movement of protons across the inner mitochondrial membrane. This movement across the membrane establishes an electrochemical potential that provides the thermodynamic driving force for the F 1 F 0 -ATP synthase to generate ATP in the matrix. Hypoxia leads to an inhibition of the electron transport chain at cytochrome c oxidase, resulting in a decrease in inner mitochondrial membrane potential. This initial decrease in inner mitochondrial membrane potential due to electron transport inhibition during oxygen deprivation is the trigger for Bax or Bak activation [179]. Furthermore, mitochondrial membrane potential decreases in response to hypoxia prior to cytochrome c release [175]. Cells devoid of mitochondrial DNA ( o cells) do not undergo cell 41

58 Chapter 1 death in response to oxygen deprivation. Mitochondrial DNA encodes 13 polypeptides, including the three catalytic subunits of cytochrome c oxidase, whereas nuclear DNA encodes the proapoptotic protein cytochrome c. Therefore, o cells do not have a functional electron transport chain and must rely only on ATP derived from anaerobic glycolysis for survival and growth. The mitochondrial dependent cell death pathway is intact in o cells, as shown by their ability to undergo death in response to a variety of apoptotic stimuli such as doxorubicin, growth factor withdrawal, and staurosporine treatment [219, 220]. The inability of o cells to inhibit the electron transport chain during oxygen deprivation is one explanation as to why o cells are resistant to oxygen deprivation induced cell death. An alternate explanation is that o cells have adapted to glycolysis. The loss of mitochondrial generated ATP due to electron transport inhibition during oxygen deprivation could lead to activation of Bax or Bak. Since o cells have adapted to glycolysis and show no acute changes in mitochondrial ATP levels during oxygen deprivation, these cells might be resistant to hypoxia induced cell death. Presently, it remains unresolved whether adaptation to glycolysis or electron transport inhibition is sufficient to prevent hypoxia induced apoptosis Oxygen sensing mechanisms Cells generally respond to prolonged hypoxia by reducing energy consumption and up regulating ATP producing pathways [221]. The universal existence of these homeostatic processes implies that all cells have the ability to sense changes in oxygen concentration and their persistence in all mammalian species attests to their importance in determining survival [222, 223]. The specific tissue of the carotid body is comprised of groups of glomus cells, enveloped by glial-type sustentacular cells, and innervated by sensory nerve fibers. These units 42

59 Chapter 1 sense arterial po 2 and respond to hypoxia by initiating a variety of responses, which include the arterial chemoreflex, i.e., increasing firing activity in the carotid sinus nerve. Despite extensive work to identify the oxygen sensor in tissues whose specific role is to regulate the response to changes in oxygen supply such as the glomus cells of the carotid body, the molecular mechanism(s) involved have yet to be identified. It is likely that there are multiple O 2 sensors that respond to various po 2 levels, are differentially distributed and subserve different cellular processes [222]. Current hypotheses centre around a number of biomolecules that can bind to O 2, either as a reversible ligand, producing allosteric shifts of the sensor, as a substrate, capable of direct oxidation of the sensor or enzymatically converted to reactive oxygen species (ROS) which, in turn, mediate the action on the effector molecules. Accordingly, the following candidates for the mechanism of oxygen sensing have been proposed Evidence of heme as an oxygen sensor Figure 1.7 Heme sensor model. Source: Lopez-Barneo J (2001) Cellular mechanism of oxygen sensing. This ligand model is based on the observation that heme proteins, such as hemoglobin and cytochrome c-oxidase, bind O 2 reversibly and the resulting configurational changes are involved in the O 2 -dependent regulation of ion transporters and enzyme activity [224]. A heme based O 2 sensor has been proposed in erythropoeitin secreting cells because their response to hypoxia is reversed by carbon monoxide (CO) which binds with greater affinity to and displaces 43

60 Chapter 1 oxygen from heme groups and is mimicked by incubation of the cells with iron chelators or cobalt, which interfere with heme synthesis or render heme unable to bind O 2 [225]. The hemesensor model, however, lacks direct experimental support. Moreover, it is now known that CO can interact directly with HIF-1, preventing its dimerization, and that iron and cobalt have opposing effects on the interaction between pvhl and HIF-1 providing alternate explanations for the observations upon which the hypothesis is based. Although HIF-1 has a PAS domain that could bind a heme group, no direct interaction of HIF-1 with O 2 has been demonstrated [226]. The enzymatic production of reactive oxygen species (ROS) is sensitive to oxygen availability and, hence, is a potential signaling pathway for transmission of the hypoxic stimulus [227]. Although ROS can be produced at numerous cell sites and organelles, two systems have been investigated as potential O 2 sensors are NADPH oxidases and the mitochondrial electron transport chain NAD(P)H oxidases This plasma membrane-associated, phagocytic oxidase (NOX2) system is a multisubunit assembly consisting of a membrane-bound catalytic complex of gp91 phox and p22 phox subunits, which together form a flavo-cytochrome b558, and a cytosolic regulatory component consisting of p47 phox, p67 phox, as well as other regulatory units including the GTPases Rac-1 and Rac-2 [228]. In NOX1 and 3 other isoforms substitute for gp91 phox and possibly other sub-units but are mechanistically similar to NOX2; however NOX4 apparently does not require the cytosolic subunits for activation [229]. 44

61 Chapter 1 Figure 1.8 Structure of NAD(P)H oxidase Source: Tracy A. (2006) Chronic granulomatous disease, our understanding. This membrane model proposes that NAD(P)H-derived electrons are shuttled to O 2 by NAD(P)H oxidase at a rapid rate during normoxia, causing superoxide production. Under hypoxic conditions, limited oxygen supply decreases superoxide generation. Reduced H 2 O 2 production and a shift of cytosolic redox pairs to the reduced state ensues [230]. In pulmonary vascular smooth muscle, this has been proposed as the mechanism underlying hypoxic pulmonary vasoconstriction. According to this model, hypoxia inactivates redox-dependent membrane K + channel, results in membrane depolarization, opening of voltage-dependent Ca 2+ channels, entry of Ca 2+, and vasoconstriction [231, 232]. The findings that O 2 -sensitive K + currents are potentiated by exogenous H 2 O 2 and by activation of the oxidase with phorbol esters, and that regulation of K + channels by O 2 is abolished in the transgenic oxidase-deficient mouse with a null gp91 phox alleles provide experimental support for this proposal [233, 234]. 45

62 Chapter 1 Figure 1.9 NAD(P)H oxidase as oxygen sensor. Source: Lopez-Barneo J (2001) Cellular mechanism of oxygen sensing. There is strong evidence that NOX2 and a reduction in ROS underlies the response to hypoxia in neuroepithelial bodies, with consequent inhibition of K + channels and depolarization, an important component of this evidence being loss of O 2 sensing in gp91 phox -deficient mice [ ]. In contrast, O 2 sensing was retained in glomus cells, adrenomedullary chromaffin cells and lung from mice lacking gp91 phox, effectively ruling out any role for NOX2 in these tissues [ ]. Although knockout of p47 phox did not abolish O 2 sensing in either lung or glomus cells, in the former it caused inhibition of a rapid transient phase of HPV, whilst in glomus cells it enhanced O 2 sensitivity and abolished the hypoxia-induced elevation in ROS [233, 241]. It has been suggested that the latter represents an important modulatory role for a non-phagocytic NOX in glomus cells [234]. NOX4 has been shown to confer O 2 sensitivity upon TASK-1 channels in a model cell system, and TASK-1 has been implicated in O 2 sensing for both PASMCs and glomus cells [ ]. Thus, although viable as an O 2 sensor for specific cellular functions in select cell types, the persistence of robust hypoxic pulmonary vasoconstriction in the gp91 phox knockout mouse argues against a universal role [235, 238, 239, 245, 246]. 46

63 Chapter Mitochondria Mitochondria, the largest consumers of oxygen, play a vital role in determining the cytosolic po 2 and the O 2 gradient between alveoli and cytosol. They are also recognized as important signaling organelles [247]. To generate ATP by oxidative phosphorylation, mitochondria use oxygen as the final electron acceptor. Accordingly, the mitochondria are well positioned as the loci of cellular oxygen sensing [247]. The electron transport chain (ETC) is a multi-step redox process that occurs in the inner mitochondrial membrane. Figure 1.10 Mitochondrial Electron Transport Chain. Source: Fig , Biochemistry, 2 nd edition, R. Garrett and C. Grisham, Saunders Publishing 47

64 Chapter 1 The Kreb's cycle and -oxidation of fatty acids generate reduced nicotinamide adenosine dinucleotide (NADH) and flavin adenine dinucleotide (FADH 2 ) which are oxidised by the electron transport chain (ETC) in the mitochondrial inner membrane. The components of the ETC are outlined in Figure Briefly, oxidation of NADH in complex I and FADH 2 in complex II leads to transfer of 2 electrons to ubiquinone to form ubiquinol. This is reoxidized by complex III (cytochrome bc 1 ) in two stages. One electron is first removed by the Fe S group leaving ubisemiquinone, and transferred via cytochromes c1 and c to complex IV (cytochrome c oxidase, COX). The ubisemiquinone left behind is reoxidised to ubiquinone by cytochrome b L, which passes the remaining electron to cytochrome b H. This reduces ubiquinone first to ubisemiquinone and then back to ubiquinol, which re-enters complex III. Oxidation of one molecule of ubiquinol back to ubiquinone thus takes 2 cycles of complex III, with the sequential transfer of its 2 electrons to cytochrome c. These electrons are finally and sequentially transferred by cytochromes a and a3 in complex IV to O 2 to form water. The operation of complexes I, III and IV cause extrusion of protons thus generating the mitochondrial membrane potential ( m) and proton gradient ( ph) which drive the F 0 F 1 ATP synthetase (reviewed in [248]). The mechanisms of the ETC are such that at various points single electrons can be lost to molecular O 2 to form reactive O 2 species (ROS) in the form of superoxide, primarily from reduced flavins in complex I and ubisemiquinone in both Qo (intermembrane space) and Qi (matrix) sides of complex III; as much as 3% of electron flux through the ETC may be constitutively lost in this way [249]. Superoxide is rapidly dismuted to peroxide by cytosolic and mitochondrial superoxide dismutase (CuZnSOD and MnSOD respectively). There is a wide consensus that inhibitors of oxidative phosphorylation or procedures that modify mitochondrial function strongly affect O 2 sensing in PASMCs, glomus cells and 48

65 Chapter 1 adrenomedullary cells though not in neuroepithelial bodies [ ]. Significant controversy exists, however, concerning the signalling mechanisms that link mitochondrial function to the effectors, and there are currently three main hypotheses of mitochondrial O 2 sensing, involving cytosolic redox state, reactive O 2 species (ROS) and energy state respectively. Under normal physiological conditions, the ETC is a significant source of reactive oxygen species (ROS) [249]. Increased ROS generation occurs under hypoxic condition, when limited oxygen availability decreases the V max of cytochrome oxidase, causing iron to remain in the ferric (Fe 3+ ) state, thereby inhibiting prolyl hydroxylase activity HIF- stabilization and transactivation [255]. ROS generation and hypoxia-responsive gene expression occur simultaneously in cardiomyocytes and Hep3B cells incubated under hypoxic conditions and these effects are abolished in mitochondria-deficient cells. Others report that hypoxic responses can be blocked by proximal ETC inhibition and addition of exogenous scavengers of ROS [86]. Similarly, overexpression of enzymes that reduce ROS levels such as catalase or glutathione peroxidase inhibit hypoxic pulmonary vasoconstriction (HPV) and the hypoxia-induced elevation in intracellular [Ca 2+ ] and reverses hypoxia-induced HIF activation [256, 257]. However, the exact correlation between po 2 and ROS production, the downstream targets of ROS, and how ROS regulate these targets at the molecular level remain to be established. 49

66 Chapter Thesis objective The pulmonary and systemic circulations have evolved to preserve blood and oxygen supply to the tissues, and this is vital for survival in physiological as well as pathological conditions. By contraction and relaxation, VSMCs alter the luminal diameter, which enables blood vessels to maintain an appropriate blood pressure. VSMCs are also important in vessel remodeling in hypoxia induced disease states, such as arteriosclerosis and pulmonary hypertension. In these cases, hypoxia regulates both cell proliferation and/or cell survival to alter the structure of the vessels. Thus VSMCs are suited not only for short-term regulation of the vessel diameter, but also for long-term adaptation, via structural remodeling by changing cell number. The function of VSMCs, and their responses to hypoxia, the basis for several cardiopulmonary disorders, has been mostly studied in relation to the function of their adjacent structures in particular to the neighboring endothelial cells and fibroblasts. The direct response of VSMCs in either the systemic or the pulmonary circulation to varying oxygen concentrations remains largely unknown. Besides, these VSMCs experience a broad range of oxygen tensions across the vessel wall, with an oxygen concentration of about 11.2% at the lumen to about 2.2% at a depth of 150. Previous studies have shown hypoxia to act as a mitogen, or, in some other studies to induce growth arrest and cell death. Lack of consistency thus precludes a unifying hypothesis. There are several structural and physiological differences between the systemic and pulmonary circulations. The bioenergetic processes that maintain cellular integrity depend on a continuous supply of oxygen and substrates. To achieve this function, vascular smooth muscle cells (VSMCs), in both the circulations, have adapted to tolerate relatively prolonged periods of 50

67 Chapter 1 hypoxia whilst maintaining appropriate distribution and delivery of essential substrates to other tissues. However, there is a contrast in their basic response to hypoxia. Whereas in the systemic circulation vasodilatation to hypoxia (hypoxic systemic vasodilatation, HSV) improves blood supply and substrate delivery to hypoxic tissue, in the lung hypoxic pulmonary artery vasoconstriction (HPV) maintains ventilation/perfusion matching and arterial oxygenation. These could lead to differences in the proliferative and survival response of VSMCs from the systemic and pulmonary circulations. Chapter 2 reports the responses of human aortic smooth muscle cells to varying degrees of hypoxia and the oxygen regulation of their proliferation and survival. Chapter 3 studies human pulmonary artery smooth muscle cells their proliferative and survival response/s compared under normoxic and hypoxic conditions. Finally, based on the responses of aortic and pulmonary artery smooth muscle cells to varying oxygen concentrations, we propose a possible molecular mechanism that accounts for the divergent responses of the two cell types to hypoxia. 51

68 Chapter Aims and hypotheses Aims: Determine if hypoxia has different effects depending on degree to which cellular energy status is compromised in HASMCs and HPASMCs. Identify oxygen sensitive cell-cycle regulatory genes and potential regulatory pathways through gene expression profiling. Establish the relevance of the above in vitro findings to systemic vascular remodeling during hypoxia in vivo. Hypothesis 1: HASMC proliferation is enhanced or inhibited depending on the severity of hypoxia. Hypothesis 2: Proliferation and survival of PASMCs under hypoxic conditions are determined by their capacity to maintain cellular ATP levels and that differences between these cells and those from the systemic circulation reflect this parameter. 52

69 Chapter 2 Chapter 2 Oxygen regulation of systemic arterial smooth muscle cell proliferation and survival 53

70 Chapter Introduction Arterial smooth muscle cells are exposed to a broad range of oxygen concentrations. In the aortic wall O 2 concentrations from 11.2% at the lumen to 2.2% at a depth of 150 m[5, 6] have been recorded and longitudinal gradients of similar magnitude occur in the normal microcirculation[40]. Regions of severe hypoxia ( 1%O 2 ) exist subjacent to atherosclerotic plaques[7] and in arteries of hypertensive[8], and diabetic[9] rodents. Smooth muscle cell proliferation and survival are modulated by hypoxia [203, 258] leading to speculation about its role in vasculogenesis [11], vascular remodeling [10] and atherosclerosis [54, 203, 258, 259]. The effects of hypoxic incubation on vascular smooth muscle cells in culture, however, have been inconsistent; with both enhanced proliferation [258, 260, 261] and growth arrest and apoptosis [142, ] have been reported. More importantly, the effect of levels of hypoxemia relevant to human cardiopulmonary disease on systemic arterial smooth muscle cell turnover, in vivo, is unknown. Inconsistencies notwithstanding, the experimental evidence supports an important role for oxygen in regulating vascular smooth muscle cell growth and survival which, given their physiological and clinical relevance, requires clarification. Since cell replication is highly energy dependent [265] it is intuitive, though unproven, that hypoxia may have different effects depending on the degree to which cellular energy status is compromised. Furthermore, although a number of cell cycle regulatory genes are oxygen sensitive, discrepant results have prevented the development of a unifying hypothesis that accounts for the divergent effects observed. This study was carried out to determine if different responses are elicited in human aortic smooth muscle cells subjected to hypoxic incubation under conditions which do or do not result in cellular ATP depletion, whether these effects are relevant to vascular remodeling during hypoxia 54

71 Chapter 2 in vivo and to identify potential regulatory pathways using gene expression profiling in cells exposed to conditions that elicit discordant responses. 55

72 Chapter Materials and Methods Antibodies and Reagents: FITC (fluorescein isothiocyanate) -conjugated Ki67 antibody was obtained from Dako (Glostrup, Denmark), Hypoxia Inducible Factor 1- alpha (HIF-1 ) antibody from Novus Biologicals (Littleton, CO, USA) and Cell division cycle 6 (CDC6) antibody from Lab Vision (Fremont, CA, USA). Mini chromosome maintenance 2 (MCM2) and p21 antibodies were from BD Pharmingen (San Diego, CA, USA). p53 antibody was from Cell Signaling Technology (Danvers, MA,USA) and Telomerase Reverse Transcriptase (TERT) antibody from Calbiochem (San Diego, CA,USA). CaspACE FITC-VAD-fmk in situ marker and TUNEL kits were both from Promega (Madison, WI, USA). JC-1 labeling kit, ATP bioluminescence assay kit and TO- PRO-3 dye were purchased from Molecular Probes (Carlsbad, CA, USA). All other reagents were from Sigma (St. Louis, MO, USA). Cell Culture Studies: Human aortic smooth muscle cells (HASMC, Cambrex Bio Science Walkersville, MD, USA), were propagated to passage 6 in SMGM-2 medium (Cambrex) consisting of SmBM medium supplemented with single aliquots of 0.1% insulin, 0.2% hfgf-b, 0.1% GA-1000 (Gentamicin and Amphotericin B) and 5% v/v FBS, 0.1% hegf. Cells exposed to hypoxia were placed in a humidified Plexiglas chamber (Billups Rothberg, San Diego, CA, USA) maintained at 37 C and continuously flushed with gas mixtures containing 10%, 5%, 3% or 1% O 2, 5% CO 2, balance N 2. By convention, the term normoxic is applied to cells exposed to air/5% CO 2 (culture media O 2 concentration = 20.5%) under otherwise identical conditions. Upon reaching 70% confluence the media was changed and cells were incubated for a further 16 or 48 hrs under 56

73 Chapter 2 either normoxic or hypoxic conditions. Experiments were repeated three times using cells from at least 2 human donors with 6 replicates per observation. Cell Counting: After exposure to normoxia or hypoxia (10%, 5%, 3% or 1% O 2 ) for 16 or 48 hours, cells were washed twice with HBSS and detached with 0.25% trypsin and 0.02% EDTA. Cell number was determined by cell counting using a standard hemacytometer (American Optical, Buffalo, NY, USA) and cell viability was assessed by Trypan Blue exclusion. Initial cell counting studies indicate that the effects of 3% and 1% O 2 differ qualitatively (Figure 1A), therefore, further studies compared normoxic cells with cells incubated at these O 2 concentrations. The concentrations of dissolved O 2 in the culture medium, measured using the ISO2 dissolved oxygen meter (World Precision Instruments, Sarasota, FL, USA) were 20.5 ± 0.6%, 3.1 ± 0.4% O 2 and 1.2 ± 0.3%O 2, respectively under the three conditions. Steady state oxygen concentrations were achieved within 30 minutes. The effect of hypoxia on the response to platelet derived growth factor (PDGF) was assessed in HASMCs incubated under normoxia for 48 h following which, the cell culture medium was replaced with medium containing 0.5% FBS. 48 hours later, 10nM PDGF-BB (R&D Systems, Minneapolis, MA, USA) or the same volume of diluent (100 l SmBM-0.5% FBS) was added to the medium, and the cells were allowed to proliferate under either normoxic or hypoxic (1% or 3% O 2 ) conditions for 48 hours. [ 3 H]-Thymidine incorporation: HASMCs, at passage 6, were seeded at a density of 2x10 4 cells / well in Corning 24 well plates, and grown for 24 hours. The culture medium was then removed and cells were incubated in 1% FBS in SMGM-2 media for another 24 hours. 1 microcurie of [ 3 H]-Thymidine (specific activity 3.22TBq/mmol; Amersham) was added to each 57

74 Chapter 2 well and the cells were exposed to air, 1% O 2 or 3% O 2 for 16, 24, 48 or 72 hours. Following each exposure, the cells were washed twice with phosphate buffered saline (PBS), and fixed with ice-cold 10% (w/v) trichloroacetic acid (TCA) for 20 minutes. The resulting precipitate was solubilized in 0.1 N NaOH (0.5 ml/well) at 37 C. Solubilized DNA was transferred into scintillation vials and [ 3 H]-Thymidine incorporation was quantified by scintillation counting (Liquid Scintillation System, Beckman Instruments). Ki67 protein: The presence of Ki67 was detected in cells grown under normoxic or hypoxic (1% or 3% O 2 ) conditions for 16 or 48 hours. Detached cells were centrifuged at 1500 rpm for 10 minutes at 20 C. The supernatant was removed, the pellet suspended in 200 l of membrane shredding solution (99.5% v/v Ca 2+ and Mg 2+ free Dulbecco s PBS, 0.5% v/v NP-40, 0.5 mm Sodium-EDTA, 0.5% w/v BSA, 20μg/ml protease inhibitior, 0.2 mg/ml RNAse) and kept at room temperature for 15 minutes. Samples were incubated with 10 l of antibody in the dark for 30 minutes and analyzed by flow cytometry (Model Epics Altra, Beckman Coulter, Fullerton, CA, USA). A minimum of 10,000 events, per sample, was recorded and cell debris was excluded by adjusting the forward light scatter threshold setting. The number of cells positive for Ki67 was calculated using CELLQuest software (Becton Dickinson, St. Louis, MO, USA). Annexin V/Propidium Iodide labeling: The Roche Annexin V FLUOS staining kit was used to detect phosphatidylserine externalization (a marker of apoptosis) in HASMC, exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hrs. The cell suspension was centrifuged at 1500 rpm for 10 minutes at 4 C. The pellet was resuspended in 5 ml of cold PBS and centrifuged again. The supernatant was removed and the pellet suspended in 100 l of Annexin V- FLUOS labeling solution (20μl Annexin V-Fluos labeling reagent and 20μl Propidium Iodide 58

75 Chapter 2 (PI) solution per ml of incubation buffer) at 37 C. Labeled cells were analyzed by flow cytometry and the numbers of cells positive for either Annexin-V, or PI, or both, were calculated. Caspase activation: Caspase activation was detected in HASMCs exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hours. CaspACE FITC-VAD-fmk is a FITC conjugate of the cell permeable inhibitor of caspases. This structure allows delivery of the inhibitor into the cell where binding to activated caspase, serves as an in situ marker for apoptosis. About cells were incubated with 100 M FITC-VAD-FMK at room temperature in the dark for 20 min. Cells were then washed, resuspended in PBS and the percentage of cells positive for activated caspase quantified by flow cytometry. Terminal deoxynucleotidyl transferase-mediated dutp nick-end-labeling (TUNEL): HASMCs exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hours were centrifuged at 1500 rpm for 10 minutes at 4 C. The supernatant was removed and the cells fixed in 1% w/v iced paraformaldehyde for an hour. The cells were washed with PBS and 70% ethanol. DNA labeling solution (TdT enzyme and FITC-dUTP, DeadEnd TM Fluorometric TUNEL System, Roche, Basel, Switzerland) was added and samples kept at 37 o C for an hour, after which they were analyzed for DNA breaks by flow cytometry. DNA Content/Cell Cycle Analysis: HASMCs exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hours were trypsinized, pelleted, resuspended, and washed once in cold PBS. The cells were fixed with 70% ethanol and maintained at 4 C for 60 min. Ethanol was washed out and cells were resuspended in 1ml of PBS and 10 l RNAse, and incubated for 45 min at room temperature. PI (10 l) was added and cells were incubated in dark at room temperature for 30 59

76 Chapter 2 min prior to analysis by flow cytometry. In each experiment, 10,000 cells were counted. The amplitude of the fluorescent signal was analyzed to quantify DNA content in order to determine the fraction of the cell population in each phase of the cell cycle (G0/G1, S, and G2/M). Mitochondrial membrane potential: Depolarization of the mitochondrial membrane was detected by a cytofluorometric method using the potential-sensitive probe 5,5', 6,6'-tetrachloro- 1,1',3,3'-tetraethyl-benzimidazolylcarbocyanine iodide (JC-1). The JC-1 monomer enters the mitochondria at physiological membrane potentials [266] where, as a result of aggregation, its emitted wavelength changes from 530 nm (green) to 590 nm (orange) when excited at 490 nm. Disaggregation to the monomeric form during mitochondrial membrane depolarization is detected as an increase in green emission. HASMCs exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hours were incubated with JC-1 (10 g/ml) for 20 minutes in the dark, washed and resuspended in 1 ml PBS. The percentage of cells positive for JC-1 monomers was quantified by flow cytometry. Intracellular ATP Concentration: HASMCs exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hours were washed twice with ice-cold PBS and lysed by the addition of equal volumes of 3.6% perchloric acid. Samples were centrifuged, and ATP concentrations in the supernatants were determined using an ATP bioluminescence assay (Molecular Probes), according to the instructions provided by the manufacturer. The photometer was set for a 5-s delay period and a 5-s integration period. ATP levels were calculated using standard reference solutions corrected for background luminescence. Western blotting: Western analysis was used to quantify levels of the hypoxia-inducible transcription factor HIF-1, markers of cell division - CDC6[267] and MCM2[268], cell cycle 60

77 Chapter 2 regulatory proteins p21, p53 and telomerase subunit TERT in nuclear extracts from HASMCs exposed to normoxia or hypoxia (3% or 1% O 2 ) for 16 or 48 hours. Cells were lysed in buffer A (10 mm HEPES [ph 7.8], 10 mm KCl, 0.1 Mm EDTA, 1 mm dithiothreitol [DTT], 0.1% Nonidet P-40 [NP-40]) with protease inhibitors (5 μg of aprotinin per ml, 5 μg of pepstatin per ml, 5 μg of leupeptin per ml, 0.5 mm Pefabloc, 1 mm phenylmethylsulfonyl fluoride) and phosphatase inhibitors (10 mm sodium fluoride, 1mM sodium orthovanadate, and 20 mm - glycerophosphate). Nuclear proteins were then extracted with buffer B (50 mm HEPES [ph 7.8], 420 mm KCl, 0.1 mm EDTA, 1 mm DTT, 5 mm MgCl 2, 20% glycerol) containing both protease and phosphatase inhibitors. Equal amounts of protein extracted from HASMCs, incubated under normoxic and hypoxic (3% and 1% O 2 ) conditions for 16 and 48 hrs were loaded on 4-12% Tris-Glycine gels, separated by electrophoresis and transferred to nitrocellulose. Membranes were blocked with 5% milk overnight, and probed with anti-cdc6 (1: 2000), anti-mcm2 (1: 2000), anti-hif-1 (1:500), anti-p21 (1: 3000), anti-p53 (1: 1000) and anti-tert (1: 200). In all cases, protein concentration was determined by the Bradford assay and appropriate volumes of extraction buffer to produce constant protein loading in each lane were mixed with SDS loading buffer. Equality of protein loading and transfer efficiency were corroborated by full-lane densitometry of the Ponceau red-stained membranes. Immunoblots were probed with horseradish peroxidase (HRP)-donkey anti-rabbit IgG (1:1000 in blocking buffer) and visualized by enhanced chemiluminescence (ECL Plus kit, Amersham Biosciences, Buckinghamshire, UK). Band intensity was quantified by densitometry (Bio-Rad Laboratories, Mississauga, ON, Canada). Microarray analysis using Affymetrix Gene Chip hybridization: In three separate experiments total RNA was isolated from HASMCs exposed to normoxia, 3% O 2 and 1% O 2 for 16 and 48 61

78 Chapter 2 hours using Trizol Reagent (GIBCO/BRL). The quality of RNA was assessed using an Agilent 2100 Bioanalyzer (version A.02.01S1232, Agilent Technologies). Hybridizations were performed on the HG-U133A GeneChip Set with a total of 22,280 genes (Affymetrix, Santa Clara, CA, USA). Samples were prepared for hybridization (6 hybridizations per experiment) according to standard Affymetrix instructions and performed at the Toronto Genomic Core Centre at the Hospital for Sick Children. Experimental design, gene lists, hierarchical trees, chip hybridizations and statistical analyses were in compliance with the Minimum Information About a Microarray Experiment (MIAME) guidelines[269]. Data obtained from the GCOS (GeneChip Operating Software) analyses of the individual arrays were normalized using the RMA (Robust Multi-chips Analysis) method. After filtering, 2 way-anova (non equal variance) was performed and differentially regulated genes were clustered using GeneSpring 7.0 ( For details see the supplementary data (GEO Accession # GSE 4725). Animal studies: All protocols were in compliance with standards set by the Canadian Council of Animal Care and were approved by the institutional animal care committee. Male Sprague-Dawley rats ( g) were placed in a Plexiglas chamber into which the flow of air and nitrogen was controlled independently. In preliminary experiments, the arterial po 2 in rats, breathing a gas mixture containing 10% O 2, averaged 38 Torr (range Torr)[270]. Rats exposed to hypoxia breathed a gas mixture containing 10% O 2 for 48 hours. Normoxic control animals breathed room air under otherwise identical conditions. Data from each animal were averaged to serve as a single value for statistical analysis. 62

79 Chapter 2 Detection of Nonviable Aortic and Mesenteric Artery Smooth Muscle Cells: Nonviable cells were detected by their failure to exclude propidium iodide. As a positive control Lipopolysachharide (LPS) from Escherichia coli 055:B5 (0.1mg/kg body weight) was injected into the right jugular vein 72 hours prior to sacrifice, to induce apoptosis. At the end of the exposure period normoxic, LPS treated, and hypoxia exposed rats were anesthetized by intramuscular injection of 0.08ml/kg xylazine (20mg/ml) plus 0.72ml/kg ketamine (100mg/ml) followed by intravenous injections of 5 mol/kg propidium iodide. After 15 minutes, an incision was made at the left atrial appendage of the heart and flushed thoroughly with 300 ml PBS. En face sections of aorta and mesenteric artery were perfusion-fixed with 3.7% paraformaldehyde for 1 hour, washed with PBS and permeabilised with 0.2% Triton X-100. TO-PRO-3 was used for nuclear counterstaining. Sections were then mounted on glass slides with glycerol/phosphatebuffered saline, and viewed under a laser-scanning confocal microscope (BioRad Radiance, Hercules, CA, USA). Cells that did not exclude propidium iodide, as indicated by nuclear fluorescence, were considered nonviable. Detection of apoptotic cells: DNA fragmentation in aortic and mesenteric artery sections was detected by TUNEL (DeadEnd TM Fluorometric TUNEL System, Roche). Slides containing paraffin-embedded sections (normoxic and hypoxic rat aorta and mesenteric artery) were dewaxed, rehydrated, permeabilized with Proteinase K, preincubated with equilibration buffer and incubated with labeling solution (rtdt and nucleotide mix with fluorescein labeled dutp) for 1hour at 37 o C. The reaction was terminated by incubating samples in a stopping buffer for 30 minutes. After PBS washes and counterstaining with TO-PRO-3, the samples were mounted and examined by laser confocal microscopy. 63

80 Chapter 2 Detection of proliferating cells: Bromodeoxyuridine (BrdU) was infused subcutaneously using osmotic pumps (Model 2ML-2, Alza Corp, Palo Alto, CA). Pumps containing 0.32 g of BrdU in 2 ml vehicle (0.4%DMSO) were implanted intrascapularly in normoxic and hypoxiaexposed rats 48 hours prior to sacrifice. The rats received approximately 0.4mg BrdU/hour delivered continuously. At the end of the labeling period thoracic aorta and mesenteric artery segments were excised, fixed with paraformaldehyde, washed with PBS, dehydrated in graded ethanol (70-100%), cleared in xylene, and embedded in paraffin. 5μm thick sections were cut on an oscillating blade microtome (Leica, Wetzlar, Germany) and placed on coated glass microscope slides (Fisher Scientific, Pittsburgh, PA, USA). Dewaxed and rehydrated slides were incubated with 2M HCl for 1hour at 37 o C. The acid was neutralized by 0.1M borate buffer, ph 8.5. Following PBS washes, the slides were incubated with fluorescein labeled anti-brdu antibody (Roche) for 1hour at room temperature, protected from light. Slides were then washed with PBS and mounted for microscopy. Quantitative analysis: The number of TUNEL positive, BrdU and TO-PRO-3 labeled cells were determined in 3 aortic and mesenteric artery sections from each of 6 animals from each experimental group. Four to six randomly selected microscopic fields ranging from cells per field were counted for each section and the number of TUNEL- or BrdU-positive nuclei expressed as a percentage of the total number of nuclei. Nuclear density was measured was determined using Image J analysis software (NIH 2002, and expressed as the number of nuclei per μm 2 x Data from each animal in each group were averaged to serve as a single value for statistical analysis. Apoptotic index was evaluated for PI and TUNEL staining. 64

81 Chapter 2 Smooth muscle -actin staining: Aorta and mesenteric artery sections from normoxic and hypoxic rats were embedded (OCT compound; Miles Scientific; Naperville, IL) and quickly frozen in liquid nitrogen. Cryostat sections (100 thick) were stained overnight at 4 C with Cy3 conjugated monoclonal anti smooth muscle -actin (Sigma, St. Louis, MO, USA). Sections were counterstained with TO-PRO-3, washed with PBS and viewed under the confocal microscope. Data analysis: Data are presented as mean ± standard error of the mean of n observations with P < 0.05 considered significant. The significance of differences between individual means was determined by two-tailed Student s t test. Differences among multiple means were evaluated by analysis of variance corrected for multiple measures where appropriate and, when overall differences were detected, differences between individual means were evaluated post-hoc using the Student Neuman - Keuls procedure. 65

82 Chapter Results The proliferative response of HASMCs during normoxic and hypoxic incubation was evaluated by cell counting (Figure 2.1), [ 3 H]-thymidine incorporation (Figure 2.2) and nuclear levels of S phase proteins (Figure 2.3). As shown in figure 2.1A cell number is increased after incubation at 5% O 2 for 48 hours and after incubation at 3% O 2 for 16 hours and 48 hours. Incubation at 1% O 2 for either 16 or 48 hours reduced the total cell number. Similarly, in rat A7R5 smooth muscle cells, cell number increased 22.8± 6.5% and 31.4±7.3% after incubation at 3% O 2 for 16 and 48hrs, respectively (p <0.05 vs. normoxic control, for both) and decreased by 38.8±5.9% and 49.7±8.6% after incubation at 1% O 2 for 16 and 48hrs, respectively (p < 0.05 vs. normoxic control, for both). Trypan blue was excluded (cells are viable) in 97.1±0.9% of HASMCs after normoxic incubation and in 94±0.4% after 16hrs and 92.2±0.6% after 48 hrs of incubation at 1% O 2 (p < 0.05 vs. normoxic control values for both). HASMC viability did not differ between normoxic cells and cells incubated at 3% O 2. HASMC (synchronized by prior incubation with 0.5% FBS) cell numbers were decreased by incubation at 1% O 2 in the absence and in the presence of PDGF after 16 (18.5±4.6% and 19.7±3.8% decrease from normoxic control values, respectively) and 48 hrs (24.4±4.5% and 26.7±2.9% decrease from normoxic control values, respectively, Figure 2.1B). Incubation of HASMCs at 3% O 2, either in the absence or presence of PDGF caused an increase in cell number at both 16 (20.3±4.5% and 23.3±4.8% increase from normoxic control values, respectively) and 48 hrs (35.6±3.7% and 36.7±4.8% increase from normoxic control values, respectively, Figure 2.1C). 66

83 Chapter 2 Figure 2.1 (A) Effects of hypoxia on human aortic smooth muscle (HASMC) cell numbers after incubation for 16 and 48 hours * p <0.05 vs. corresponding normoxic control values, #,p <0.05 vs. corresponding 3% O 2 values. 67

84 Chapter 2 Figure 2.1 Effects of hypoxia on the proliferative response of HASMC cell numbers to 10nM PDGF-BB after incubation at 1% O 2 (B) and 3% O 2 (C) for 16 and 48 hours. n=6 per condition. * p<0.05 vs. corresponding normoxic control values. 68

85 Chapter 2 The effect of hypoxic incubation for up to 72 hours on [ 3 H]-thymidine incorporation, in HASMCs, is illustrated in Figure 2.2. Incubation at 1% O 2 decreased (Figure 2.2A) whereas 3% O 2 increased (Figure 2.2B) the rate of [ 3 H]-thymidine incorporation reflecting the effects of these conditions on the rate of DNA synthesis. The effects of hypoxia on other markers of cell proliferation, Ki67 (nuclear proliferation marker), CDC6[267] and MCM2[268] (both form the prereplication complex at the initiation site for DNA synthesis), were also studied (Figures 2.3 and 2.4). The percentage of cells staining positive for Ki67 decreased after incubation at 1% O 2, whereas 3% O 2 had the opposite effect (Figure 2.3 A and B). Similarly, CDC6 protein was decreased and increased in cells incubated at 1% and 3% O 2, respectively (Figure 2.4, A and B) and MCM2 protein was reduced after incubation at 1% O 2 (Figure 2.4, C and D). 69

86 Chapter 2 Figure 2.2 (A) ( 3 H)-Thymidine incorporation (counts per minute), in human aortic smooth muscle cells is decreased after incubation at 1% O 2 and (B) increased after incubation at 3% O 2 compared with the normoxic cells. n=6, * p <0.05 vs. corresponding normoxic control values. 70

87 Chapter 2 Figure 2.3 The percentage of cells positive for the Ki67 antigen is decreased after incubation at 1% O 2 (A) and increased after incubation at 3% O 2 (B). * p <0.05 vs. corresponding normoxic control values. 71

88 Chapter 2 Figure 2.4 CDC6 (A and B) and MCM2 (C and D) protein levels after normoxic and hypoxic (1% O 2 and 3% O 2 ) incubation. 1% O 2 decreased CDC6 protein level while CDC6 protein is increased in cells exposed to 3% O 2. 1% O 2 decreased MCM2 protein level with no change in cells exposed to 3% O 2. * p <0.05 vs. corresponding normoxic control values. 72

89 Chapter 2 To determine whether hypoxia alters HASMC proliferation through a general effect on cell cycle progression or through mechanisms specific to a particular cell cycle phase or checkpoint, DNA content was quantified by PI staining (Figure 2.5). After incubation at 1% O 2 there is accumulation of G 0 /G 1 phase cells and depletion of cells in S and G 2 /M phases (Figure 2.5A) indicating a delay in G 1 /S transition. Conversely, incubation at 3% O 2 results in a decrease in the percentage of cells in G 0 /G 1 phase, an increase in S phase cells and a corresponding increase in the percentage of cells undergoing mitosis (Figure 2.5B) suggesting acceleration of progression through the G 1 /S interphase. 94±0.8% of cells remained viable (did not stain for either Annexin V or PI) after normoxic incubation compared with 91.8±0.7%, and 83.7±0.8% following incubation at 1% O 2 for 16 and 48 hours, respectively (p < 0.05 vs. corresponding normoxic control values for both). After incubation at 3% O 2 the percentage of viable cells did not differ from the normoxic control value. 73

90 Chapter 2 % cells G0/G1 S G2/M 1% O 2 48hrs 91.8 * 5.1 * 3.1 1% O 2 16hrs 85.1 * 11.1 * % O % cells in cell cycle stage % cells G0/G1 S G2/M 3% O 2 48hrs 67.6 * 22.9 * 9.5 3% O 2 16hrs 75.3 * 16.9 * % O % cells in cell cycle stage Figure 2.5 Flow cytometric analysis of propidium iodide stained cells incubated to 70% confluence under normoxia, then incubated for a further 48 hours under normoxia or hypoxia (1% or 3% O 2 ) or for 32 hours under normoxia and 16 hours under hypoxia (1% or 3% O 2 ). (A) Incubation at 1% O 2 increases the percentage of cells at the G 0 /G 1 interphase compared with the normoxic cells, whereas (B) incubation at 3% O 2 increases the percentage of cells in G 2 /M at 16 and 48 hours, compared with the normoxic cells. n=6, * p<0.05 vs. corresponding normoxic control values. 74

91 Chapter 2 Since differences in cell number may reflect changes in the rate of cell death as well as proliferation, the prevalence of smooth muscle cell apoptosis after normoxic and hypoxic incubation was assessed in HASMCs by Annexin V / PI staining and TUNEL (Figure 2.6). The percentage of cells that stained with Annexin V-FITC only (early apoptotic cells) increased after incubation at 1% O 2 and decreased at 3% O 2 compared to the normoxic cells (Figure 2.6A). Similarly, caspase activity and the percentage of TUNEL positive cells are increased after incubation at 1% O 2 (Figure 2.6B and Figure 2.6C respectively). Figure 2.6 (A) Annexin V/PI staining indicate that apoptosis increased after incubation at 1% O 2 and decreased at 3% O 2. * p<0.05 vs. corresponding normoxic control values. 75

92 Chapter 2 Figure 2.6 (B) Caspase activity and (C) TUNEL indicate that apoptosis is increased after incubation at 1% O 2. * p<0.05 vs. corresponding normoxic control values. 76

93 Chapter 2 Depolarization of the mitochondrial membrane both results from and contributes to impairment of cellular ATP synthesis and is an early event in apoptotic cell death. Figure 2.7 illustrates the effect of hypoxic incubation on the percentage of cells positive for JC-1 monomers (an index of mitochondrial membrane depolarization). In cells incubated at 1% O 2 for 16 hours there is an increase in the percentage of cells positive for JC-1, indicating mitochondrial membrane depolarization, and a further increase after 48 hours (Figure 2.7 A). There was no change in the percentage of JC-1 monomer-positive cells after incubation at 3% O 2 (Figure 2.7 B). As shown in Figure 2.8 A, there is a decrease in cellular ATP concentration in HASMCs incubated for 16 hours at 1% O 2 and a further decrease after 48 hours. Incubation at 3% O 2 did not alter cellular ATP concentration (Figure 2.8, B). 77

94 Chapter 2 Figure 2.7 Increased JC-1 monomer formation indicates mitochondrial membrane depolarization after incubation at 1% O 2 (A) but not at 3% O 2 (B). * p<0.05 vs. corresponding normoxic control values. 78

95 Chapter 2 Figure 2.8 Cellular ATP concentration is decreased after incubation at 1% O 2 (A) but not 3% O 2 (B). * p <0.05 vs. corresponding normoxic control values. 79

96 Chapter 2 Gene expression profiling was used to identify oxygen regulated genes and, hence, potential regulatory pathways. Complete gene expression data are presented in online (GEO accession # GSE 4725). The fold changes from normoxic control values in the expression of genes with known pro- and antiproliferative and pro- and antiapoptotic roles, which demonstrated significant changes after hypoxic incubation, are presented in Table 2.1. Table 2.1 Normalized expression of genes of pro- and antiproliferative genes and pro- and antiapoptotic genes under normoxic and hypoxic conditions (1% O 2 and 3% O 2 ). Genebank # Gene 1% O 2 3% O 2 16hrs 48hrs 16hrs 48hrs Proproliferative genes M63889 Fibroblast growth factor receptor NM_00295 Replication protein A L14922 Replication factor C (activator 1) NM_00125 Cell division cycle NM_00479 Mitogen-activated protein kinase-activated protein kinase NM_00275 Mitogen-activated protein kinase 2 NM_ Mitogen-activated protein kinase kinase kinase kinase U29725 Mitogen-activated protein kinase

97 Chapter 2 Table 2.1 Normalized expression of genes of pro- and antiproliferative genes and pro- and antiapoptotic genes under normoxic and hypoxic conditions (1% O 2 and 3% O 2 ) contd. Genebank # Gene 1% O 2 3% O 2 16hrs 48hrs 16hrs 48hrs Antiproliferative genes NM_00039 Cyclin-dependent kinase inhibitor 1A, p NM_00561 Retinoblastoma-like 2 (p130) NM_00154 Inhibitor of growth family, member 1-like AA Bone morphogenetic protein NM_00447 Dual specificity phosphatase U16996 Dual specificity phosphatase AF Tuberous sclerosis Proapoptotic genes AF Tumor protein p NM_00436 Caspase AB Caspase 8 associated protein AF Caspase recruitment domain protein NM_ Serine/threonine kinase 17a NM_ Serine/threonine kinase 17b NM_ Apoptotic protease activating factor Antiapoptotic genes AF CASP8 and FADD-like apoptosis regulator U72398 BCL2-antagonist of cell death

98 Chapter 2 Figure 8 presents the effects of hypoxic incubation on nuclear levels of HIF-1, p21 and p53. HIF-1 levels are increased to a similar extent after incubation at 1% and 3% O 2. Levels of p21 and p53 proteins, inhibitory regulators of cell cycle progression, are increased after incubation at 1% O 2 with no change after incubation at 3% O 2. Levels of TERT protein, a component of the telomerase complex, that increaseed and enhanced HASMC survival following longer hypoxic epochs (>20 days), [271, 272] did not differ from the normoxic control values after 16 and 48 hours of incubation at 1% and 3% O 2 (data not shown). Figure 2.9 (A) Nuclear levels of HIF-1 protein after incubation of HASMCs under normoxic and hypoxic (1% or 3% O 2 ) conditions. Solid bars represent protein levels after normoxic incubation, open bars after incubation at 1% O 2 and hatched bars after incubation at 3% O 2. * p<0.05 vs. corresponding normoxic control values. 82

99 Chapter 2 Figure 2.9 Nuclear levels of p21 (B) and p53 (C) proteins after incubation of HASMCs under normoxic and hypoxic (1% or 3% O 2 ) conditions. Solid bars represent protein levels after normoxic incubation, open bars after incubation at 1% O 2 and hatched bars after incubation at 3% O 2. * p<0.05 vs. corresponding normoxic control values. 83

100 Chapter 2 As shown in Figure 2.10, exposure to hypoxia increased the number of propidium iodide staining cells (Figure 2.10 A) with a corresponding increase in the number of TUNEL positive nuclei (Figure 2.10 B) in both aortic and mesenteric artery sections. Normoxic Hypoxic Mesenteric artery Aorta Mesenteric artery Aorta Figure 2.10 (A) Propidium iodide staining of en face sections in paraffin embedded sections of normoxic and hypoxic (48hrs) rat aorta and mesenteric artery (40X magnification). 84

101 Normoxic Hypoxic Chapter 2 B Mesenteric artery Aorta Mesenteric artery Aorta C Aorta Mesenteric artery Figure 2.10 (B) TUNEL in paraffin embedded sections of normoxic and hypoxic (48hrs) rat aorta and mesenteric artery (40X magnification). (C) Quantitative analysis confirms increased cell death and apoptosis after hypoxic exposure. 85

102 Chapter 2 Figure 2.11 (panel A) shows that, in aortic and mesenteric artery sections from hypoxia exposed rats, there is an increase in the number of BrdU labeled cells compared with the normoxic control group. Compared to normoxic animals, aortic and mesenteric artery sections from hypoxia-exposed rats also demonstrate increased medial nuclear density (Figure 2.11, panel B). Smooth muscle -actin staining (Figure 2.11, Panel C) confirms that the increase in cellularity reflects an increase in medial smooth muscle. 86

103 Chapter 2 Aorta Mesenteric artery Normoxic Hypoxic Normoxic Hypoxic A B C Figure 2.11 (A) Immunohistochemical staining of incorporated BrdU in paraffin embedded sections of aorta and mesenteric artery from normoxic and hypoxia exposed (48hrs) rats (20X magnification). (B) Double staining with TO-PRO-3 and (C) -Smooth Muscle Actin in paraffin embedded sections of normoxic and hypoxic (48hrs) rat aorta and mesenteric artery shows increased cellularity and medial smooth muscle cell density after hypoxia (40X magnification). 87

104 Chapter 2 Nuclear density (Number of nuclei per μm 2 x 10-4 ) Figure 2.11 (D) Quantitative analysis of incorporated BrdU and TO-PRO-3 staining confirms increased proliferation (top) and nuclear density (bottom). 88

105 Chapter Discussion We found that in HASMCs in culture: 1) incubation at 3% O 2 enhances, whereas 1% O 2 inhibits DNA synthesis and the time-dependent increase in cell number; 2) incubation at 1% O 2 is associated with accumulation of cells in G1 phase of the cell cycle whereas 3% O 2 increased the percentage of cells in S and G2/M; 3) the prevalence of apoptosis is increased after incubation at 1% O 2 ; 4) cellular ATP levels are reduced and the mitochondrial membrane is depolarized after exposure to 1% but not 3% O 2 ; and 5) pro-and antiproliferative and pro- and antiapoptotic gene expression are tightly coordinated to effect directionally opposed responses within a narrow range of oxygen concentrations. In aorta and mesenteric arteries of rats breathing 10% O 2 (arterial PO 2 40 mmhg) both smooth muscle cell proliferation and apoptosis are increased. Hypoxia has been reported to enhance proliferation in pulmonary artery and aortic smooth muscle cells [10, 258, 260, 261, ]. In apparent contradiction, others have observed that hypoxic stress inhibits growth of these same cells [142, 147, 263]. The results of the current study show that cell numbers, [ 3 H]-Thymidine incorporation, biochemical markers of proliferation (Ki67[277], CDC6[267] and MCM2[268] protein levels) and the response to PDGF-BB are increased and decreased after incubation at 3% O 2 and 1% O 2, respectively. We conclude that hypoxia may either enhance or inhibit HASMC proliferation depending on its severity. Rather than being due to experimental error or differences in cell origin, the discrepancy in previous observations, therefore, reflects fundamental differences in the nature of the cellular response elicited by varying degrees of hypoxic stress. The effect persists during and is additive to the effects of extrinsic stimulation with PDGF, suggesting that the responses are 89

106 Chapter 2 mediated by nonconvergent pathways. Total cell number, however, reflects effects on both proliferation and cell death, and differential modulation of these processes by hypoxia and/or PDGF could account for these findings without inferring independent signaling mechanisms. Following incubation at 1% O 2 we found that the percentage of cells in G1 phase of the cell cycle is increased, with a corresponding depletion of cells in G2/M and S phases indicating a delay in progression through the G1/S interphase. In contrast, incubation at 3% O 2 results in a decrease in the number of cells in G1 phase, an increase DNA synthesis and in the percentage of mitotic cells, consistent with acceleration of the G1/S transition. The most frequently reported effect of hypoxia is delayed entry into S phase [142, 147, 148, , 278] as we observed in cells incubated at 1% O 2. Our finding that enhanced proliferation in cells incubated at 3% O 2 is also associated with an alteration in the rate of G1/S transition, albeit opposite in direction, is novel. It suggests that the bidirectional effects of hypoxia are integrated by events occurring at this checkpoint. During transition from G1 to S phase cyclin-dependent kinases (cdks) phosphorylate the retinoblastoma protein (Rb) displacing the E2F-1 transcription factor and activating expression of S phase genes required for DNA synthesis [145]. Activity of the cdks is dependent on association with their respective cyclins and regulated by endogenous cdk inhibitors (eg. p21 and p27). As shown in Table 2.1, genes whose products act at the G1/S transition are differentially regulated under the conditions studied. Both p53 and its transcriptional activation target p21, which inhibit G1/S transition, are increased after incubation at 1% O 2 and unchanged at 3% O 2. Moreover, genes with known roles in modulating p53/p21 activity demonstrate patterns consistent with their expected functional effects: Expression levels of replication protein A1, which binds p53 preventing activation of p21 transcription [279]; components of the 90

107 Chapter 2 extracellular signal regulated kinase cascade, which stimulate assembly of the cyclind CDK4/6 complex promoting G1/S progression and p27 degradation [280]; are enhanced after incubation at 3% O 2 and inhibited following incubation 1% O 2. Replication factor C (activator 1) 1[281], CDC6 [267] and MCM2 [268] are similarly affected. Conversely, expression of antiproliferative genes, Inhibitor of Growth Family, member 1 like, which binds and enhances p53 activity [282], Bone Marrow Morphogenetic Protein 2, which induces SMAD-mediated expression of p21 and p27 [283] and down regulates antiapoptotic Bcl-x L expression [284], and the dual specificity phosphatases 1 and 5, which deactivate ERKs [285], show the opposite pattern. Corresponding changes in Tuberous Sclerosis Complex-1, a negative regulator of mrna translation [286] (Table 2.1), suggest a mechanism by which the effects of hypoxia on protein and DNA replication may be coordinated. The results of three independent assays (Annexin V/PI, TUNEL and mitochondrial membrane depolarization) and the changes in pro- and antiapoptotic gene expression (Table 2.1) in the present study indicate that the rate of apoptosis is increased in HASMCs incubated at 1% O 2. The reduced cell number observed under this condition therefore reflects, in part, an enhanced rate of cell death. In fibroblasts and tumour cells, hypoxia of sufficient severity to cause ATP depletion impairs DNA repair and the increase in apoptosis in this setting is considered a protective mechanism to prevent the accumulation of hypoxia-induced mutations [201, 203, 287]. The increased rate of smooth muscle cell apoptosis during hypoxic epochs associated with reduced [ATP] i in the current study may serve a similar adaptive purpose. That hypoxia induces apoptosis at oxygen concentrations recorded in arteries affected by aneurismal dilatation and atherosclerotic plaques, [288] and that hypoxemia, in vivo, increases smooth muscle apoptosis in the walls of systemic arteries supports the suggestion that it may play a 91

108 Chapter 2 pathogenic role in arterial smooth muscle cell dysfunction [270] and loss in patients with cardiopulmonary disease and shock. Apoptosis may be triggered in response to stimuli extrinsic or intrinsic to the affected cell. Ligand binding to the TNF receptor superfamily such as Fas, through their association with the Fas associated death domain (FADD) protein, results in assembly of the death inducing signaling complex (DISC) which recruits and activates caspase-8. The intrinsic pathway is activated when mitochondrial membrane depolarization releases cytochrome C into the cytoplasm where it binds apoptotic protease activating factor-1 (Apaf-1) allowing it to activate pro-caspase-9. In the final common pathway caspase-9 (intrinsic) and caspase-8 (extrinsic) cleave and activate the effector protease, caspase-3[289]. p53, through interactions with Apaf-1 [180] and protective Bcl proteins, directly activates caspase-3 [290] and enhances permeability of the outer mitochondrial membrane [291]. Complementary regulation by death inhibitors and the balance between prosurvival and proapoptotic Bcl-2 family members[292] superimposes an additional level of control. Our finding (Table 1) that the expression of components of both the extrinsic (Caspase-8 associated protein 2, Casp8 and FADD-like apoptosis regulator) and intrinsic (caspase recruitment domain protein 7, apoptosis-inducing serine/threonine kinase 17a and b, BMP2, Apaf-1, Bcl2l1 (BCL2-antagonist of cell death) and p53 [148]) pathways as well as Caspase-3 are oxygen regulated, therefore, indicates that stimuli originating within the cellular microenvironment as well as the intrinsic response are important in hypoxic activation of the apoptotic program. Resistance to apoptosis has been reported in lung fibroblasts, A7r5 cells, rat kidney proximal tubular cells and rat pheochromocytoma PC12 cells during hypoxic incubation [207, 92

109 Chapter 2 293, 294]. In the current study the percentage of cells staining positive for Annexin V was decreased and the expression of antiapoptotic genes (CASP8 and FADD-like apoptosis regulator [295] and Bcl2l1, Bcl2 antagonist of cell death [296]) was increased following incubation at 3% O 2. Consistent with previous results in other cell types, therefore, hypoxia, at levels above those causing outright cytotoxicity, enhances smooth muscle cell survival. In the systemic circulation, protection from apoptotic cell death may be important in neovasoformation in response to oxygen delivery/requirement imbalance during development and in ischemic tissues where investiture of new endothelial networks by regulatory smooth muscle cells must occur under conditions of relative oxygen deficiency. The mitochondria are the site of oxidative phosphorylation, however, they must themselves consume ATP [297] to maintain the trans-mitochondrial membrane potential on which electron transport coupling depends [298]. Failure of efficient ATP synthesis and, consequently, ion-motive ATPase activity ultimately results in depolarization of the membrane, further ATP depletion, Ca 2+ influx, phospholipase and protease activation and the release of apoptotic factors [ ]. Little information is available concerning the degree of hypoxia required to initiate programmed cell death and it may vary among cell types [302]. Our present results show that incubation of HASMCs at 1% but not 3% O 2 causes mitochondrial membrane depolarization, therefore, the threshold between normal physiological functioning and cell destruction is exceptionally narrow in these cells, particularly when compared to the range of oxygen concentrations to which they are normally exposed [303]. Conflicts among previous reports are undoubtedly attributable, at least in part, to the lack of appreciation of this and the experimental rigor needed to separate these responses. 93

110 Chapter 2 Nuclear levels of HIF-1 were elevated to a similar extent in HASMCs incubated at 3% and 1% O 2 whereas the effects on HASMC proliferation were directionally opposed. This suggests that HIF-1 independent regulatory mechanisms predominate. It would be unusual, however, if the primary oxygen sensing mechanism in mammalian cells were not involved in such an important response. Moreover, many cell cycle associated genes contain functional hypoxia regulatory elements and HIF-1-regulated pathways that both enhance cell survival [304, 305] and, conversely, increase apoptotic cell death [147, 306] have been identified. A more complex role than can be accounted for by changes in HIF-1 levels must therefore be proposed in order to reconcile these observations. Differences between mrna expression of the HIF-1 regulated genes [87, 307] represented on the Affymetrix HG-U133A array in cells incubated under the two conditions (see supplemental data, GEO Accession # GSE 4725) support this notion. This is not surprising since HIF-1 is subject to extensive post-translational modification prior to nuclear translocation [56] and interacts with a multitude of coregulatory factors [147, 306, 308] offering many sites at which its function may be differentially affected by the two conditions. During the prenatal period, the systemic circulation undergoes continuous restructuring in response to the changing requirements of the developing tissues. In the mature circulation, hypoxemia, due to cardiopulmonary disease, elicits responses which redistribute blood flow and enhance the capacity for oxygen extraction [ ]. As the duration of hypoxia increases, however, systemic vascular smooth muscle and endothelial cell function are impaired [270, 312], limiting the efficacy of the acute responses. Concurrent structural remodeling thus plays an increasing role in maintaining the balance between oxygen delivery and metabolic demand[311]. Our current results indicate that this is facilitated by increases in the rates of both smooth muscle 94

111 Chapter 2 cell proliferation and death, a paradoxical state that will markedly enhance cell turnover. During remodeling, vascular cell replication and removal must be tightly controlled to ensure a degree of plasticity sufficient to achieve the required structural change while avoiding the accumulation of mutations and malignant transformation or the formation of abnormal structures that exacerbate circulatory dysfunction. The results of the present study indicate that the difference between oxygen concentrations that enhance smooth muscle cell proliferation and those that impair cellular energy status and trigger cell destruction is correspondingly small and well within the transmural and longitudinal gradients known to exist in the systemic circulation. Although both proliferation and apoptosis are enhanced in aortae and mesenteric arteries from hypoxia-exposed rats, the net effect is an increase in medial smooth muscle. Prolonged hypoxia of this severity results in a progressive loss of systemic arterial and arteriolar contractility [270, 313] with consequent impairment of the sympathetically-mediated reflexes that regulate blood flow distribution [314]. In this context, increased muscularity of the arterial wall can be viewed as a compensatory adaptation that preserves the capacity to regulate the systemic circulation. Vital organ function is highly intolerant of oxygen deprivation. Accordingly, mechanisms linking vascular cell turnover and the capacity for rapid structural change directly to oxygen concentration are required to avoid delays inherent in second messenger signaling. Our results indicate that pro- and antiproliferative and pro- and antiapoptotic gene expression are tightly coordinated to produce directionally opposed responses effected at the level of G1/S transition and involvement of both intrinsic and extrinsic apoptotic pathways. Further definition of the individual roles of these regulatory mechanisms will be valuable in identifying therapeutic targets in conditions in which enhanced plasticity of the 95

112 Chapter 2 vasculature may be exploited to alleviate tissue oxygen deficiency or in which over exuberant remodeling interferes with normal cardiovascular function. 96

113 Chapter 3 Chapter 3 Oxygen regulation of pulmonary artery smooth muscle cell proliferation and survival 97

114 Chapter Introduction Hypoxic pulmonary arterial hypertension (HPAH) contributes to morbidity and premature mortality in patients with cardiopulmonary diseases and responds poorly to treatment. [ ]. Increased smooth muscle in the pulmonary arterial wall is a hallmark of this condition and contributes to the increase in resistance to blood flow [315, 316, ]. The factors determining SMC proliferation and survival in the pulmonary circulation and, hence, susceptibility to and severity of HPAH are poorly understood. Studies to determine the mechanisms regulating these processes are, therefore, of high priority. Smooth muscle cells of diverse origin proliferate in response to moderate levels of hypoxia (2-3% O 2 ) [258, 323]. Proliferation and viability of systemic arterial SMCs are reduced, however, when hypoxia is sufficiently severe that the capacity to maintain intracellular ATP levels is impaired [323]. Pulmonary artery smooth muscle cells (PASMCs) differ from those in systemic arteries in several important respects. In addition to differences in their contractile responses to hypoxia (contraction in PASMCs vs. relaxation in systemic vascular SMCs), and vasoactive factors which may reflect differences in membrane ion channel expression [14, 324, 325], PASMCs are metabolically distinct. Mitochondria from these cells are relatively depolarized, display lower expression of proximal ETC components and a greater propensity to generate ROS at reduced oxygen tensions [326]. This could affect the response to hypoxic stress and would be particularly apparent in highly energy dependent responses such as proliferation. In vivo, hypoxia causes initial PASMC proliferation in rodents followed by a return to quiescence with continued hypoxic exposure under the combined influences of endotheliumderived mediators, circulating neurohumoral factors, and the direct effects of the hypoxic microenvironment [ ]. In cultured PASMCs from various species proliferation has been 98

115 Chapter 3 reported to be enhanced [276, 329, ], inhibited [10, 276, 329, 331, ] or unaltered [331, 335, 340, 341] during exposure to hypoxia of varying severity and duration (Table 3.1). This lack of consistency in experimental design among earlier studies has precluded the development of consensus regarding the direct role of ambient oxygen concentration. It has been equally difficult, for similar reasons, to establish the threshold at which hypoxia triggers cell death in PASMCs and whether this differs from that in SMCs derived from other tissues, although mechanisms that might confer protection from apoptosis in PASMCs have been proposed [284, ]. The current study was, therefore, carried out to determine if: 1) the effects of hypoxic incubation on human PASMC proliferation and survival differ from those in SMCs from the systemic circulation; 2) these responses differ under conditions which do or do not result in cellular ATP depletion; and 3) these effects are relevant to pulmonary vascular remodeling during hypoxia in vivo. 99

116 Chapter 3 Table 3.1 Influence of hypoxia on PASMC proliferation. First author (Reference) Species O 2 % Duration of hypoxia (days) Proliferation PASMC location Seeding density (cells/cm 2 ) FBS % Yang (329) Human (bronchial carcinoma) 0 2, 4, 6 Distal PA Eddahibi (339) Rat 0 1 Proximal PA Hassoun (340) 0 Proximal PA ,10 Proximal PA Cooper (276) Human 0 4 Not specified , 2, 5 Rose (10) Human, rabbit 1 1 Not specified 4000 Serum free Stotz (338) Rat 1 2 Microvascular Lu (336) Rat 2 1,2 Not specified Frid (331) Cow 3 2 Lower media, outer ~ media rounded epitheloid cells 3 2 All medial cells ~ Middle media, outer media spindle shaped cells ~ Lanner (335) Pig 3 2, 3, 4 Main branch of PA Main branch of PA Tamm (350) Human (peripheral lung cancer) 3 Not specified 5 Preston (337) Rat Lobar PA , 0.1 Dempsey (273) Cow 3 2, 3, 4 Not specified 25, Stiebellehner Cow 3 4 Distal PA (341) 3 Distal PA Ambalavalan Pig 1, 2, 3, 5, 7, 10 3 Not specified (332) (1, 3, 5, 7) Benitz (333) Cow 3,6, 9 10 Not specified PASMC: Pulmonary artery smooth muscle cell, FBS: Foetal bovine serum (concentration in cell culture media), : Increase in PASMC proliferation, : Decrease in PASMC proliferation, : No change in PASMC proliferation. 100

117 Chapter Materials and methods Antibodies and Reagents Hypoxia inducible factor 1- (HIF-1 ) antibody was purchased from Novus Biologicals (Littleton, CO), p21 antibody from BD Pharmingen (San Diego, CA) and p53 antibody from Cell Signaling Technology (Danvers, MA). CaspACE FITC-VAD-fmk in situ marker and TUNEL kits were both from Promega (Madison, WI). 5,5,6,6 -Tetrachloro-1,1,3,3 -tetraethylbenzimidazolylcarbocyanine iodide (JC-1) labeling kit, ATP bioluminescence assay kit, and TO- PRO-3 dye were purchased from Molecular Probes (Carlsbad, CA). All other reagents were from Sigma (St. Louis, MO). Cell Culture Studies: Human pulmonary artery smooth muscle cells (HPASMCs, Cambrex Bio Science Walkersville, MD, USA) were propagated to passage 6 in SMGM-2 medium (Cambrex) consisting of SmBM medium supplemented with single aliquots of 0.1% insulin, 0.2% hfgf-b, 0.1% GA-1000 (Gentamicin and Amphotericin B) and 5% v/v FBS, 0.1% hegf. Upon reaching 70% confluence the media was changed and cells were incubated for a further 16 or 48 hrs under either normoxic or hypoxic conditions. Cells exposed to hypoxia were placed in a humidified Plexiglas chamber (Billups Rothberg, San Diego, CA, USA) maintained at 37 C and continuously flushed with gas mixtures containing 10%, 5%, 3%, 1% or 0% O 2, 5% CO 2, balance N 2. Normoxic cells were exposed to air/5% CO2 under otherwise identical conditions. The concentrations of dissolved O2 in the culture medium, measured using the ISO2 dissolved oxygen meter (World Precision Instruments, Sarasota, FL, USA) were 20.5 ± 0.6%, 3.1 ± 0.4% O2, 1.2 ± 0.3% O2 and 0.3±0.2% O 2 when the chamber was flushed with air/5% CO 2, 3% O 2, 1% O 2 and 0% O 2 gas mixtures, respectively. Steady state oxygen concentrations were achieved 101

118 Chapter 3 within 40 minutes in each case. Experiments were repeated three times using cells from at least 3 human donors with 6 replicates per observation. Cell Counting: After exposure to normoxia or hypoxia (10%, 5%, 3%, 1% or 0% O2) for 16 or 48 hours, cells were washed twice with HBSS and detached with 0.25% trypsin and 0.02% EDTA and cell number was determined using a Coulter counter (Beckman Coulter, Inc, Fullerton, CA). Cell viability was assessed by Trypan Blue exclusion. BrdU incorporation/cell cycle analysis: Bromodeoxyuridine (BrdU), an analog of the DNA precursor thymidine is incorporated into newly synthesized DNA. Propidium iodide (PI) staining was simultaneously done to measure total DNA. Cells were pulse-labeled by the addition of BrdU to the culture medium to a final concentration of 10 μm during the last 16hrs or 48 hrs of normoxic or hypoxic incubation. At the end of the incubation period, cells were harvested and fixed for an hour in 70% ethanol followed by denaturation of DNA with 0.1 M HCl and neutralization with 0.1M borate buffer. Cells were then incubated for 45 min in a solution of PI (2.5 μg/ml) and RNase A (50 μg/ml) and with a fluorescein isothiocyanate-conjugated monoclonal anti-brdu antibody (diluted 1:1000, Pharmingen, San Diego, CA). The percentage of cells staining positive for BrdU uptake was determined by flow cytometry (BD FACScan flow cytometer, BD Biosciences, NJ) using CellQuest Software. Annexin V-Propidium Iodide labeling: To assess the effect of hypoxic incubation on apoptosis, the Roche (Basel, Switzerland) annexin V-fluorescence (Fluos) staining kit was used to detect phosphatidylserine externalization (an early event in apoptosis), and PI uptake (a marker of cell death) in HPASMC exposed to normoxia or hypoxia (3%, 1% or 0% O 2 ) for 16 or 48 h. Cell suspension was centrifuged at 1,500 rpm for 10 min at 4 C. The pellet was resuspended in 5 ml of cold PBS and centrifuged again. The supernatant was removed and the 102

119 Chapter 3 pellet suspended in 100 l of annexin V-Fluos labeling solution (20 l annexin V-Fluos labeling reagent and 20 l Propidium Iodide (PI) solution per milliliter of incubation buffer) at 37 C. Labeled cells were analyzed using flow cytometry, and the number of cells positive for Annexin V, PI, or both were calculated. Caspase activation: Caspase activation was detected in HPASMCs exposed to normoxia or hypoxia (3%, 1% or 0% O2) for 16 or 48 hours. CaspACE FITC-VAD-fmk is a FITC conjugate of the cell permeable inhibitor of caspases. This structure allows delivery of the inhibitor into the cell where binding to activated caspase, serves as an in situ marker for apoptosis. About cells were incubated with 100 μm FITC-VAD-FMK at room temperature in the dark for 20 min. Cells were then washed, resuspended in PBS and the percentage of cells positive for activated caspase quantified by flow cytometry. Mitochondrial membrane potential: Depolarization of the mitochondrial membrane, was detected using the potential-sensitive probe JC-1. JC-1 monomers enter the mitochondria at physiological membrane potentials (62), where, as a result of aggregation, the emitted wavelength changes from 530 nm (green) to 590 nm (orange) when excited at 490 nm. Disaggregation to the monomeric form during mitochondrial membrane depolarization is detected as an increase in green emission. HPASMCs exposed to normoxia or hypoxia (3%, 1% or 0% O 2 ) for 16 or 48 h were incubated with JC-1 (10 g/ml) for 20 min in the dark, washed, and resuspended in 1 ml PBS. The % cells positive for JC-1 monomer was assessed by flow cytometry. Intracellular ATP concentration: HPASMCs exposed to normoxia or hypoxia (3%, 1% or 0% O 2 ) for 16 or 48 h were washed twice with ice-cold PBS and lysed by adding equal volumes of 3.6% perchloric acid. Samples were centrifuged, and ATP concentrations in the supernatants 103

120 Chapter 3 were determined by using an ATP bioluminescence assay (Molecular Probes), according to the instructions provided by the manufacturer. The photometer (Multiskan EX Microplate Photometer, Thermo Labsystems, Philadelphia, PA) was set for a 5-s delay period and a 5-s integration period. ATP levels were calculated by using standard reference solutions corrected for background luminescence. Western blot analysis: Western blot analysis was used to quantify levels of HIF-1 and the cell-cycle regulatory proteins p21 and p53 in nuclear extracts from HPASMCs exposed to normoxia or hypoxia (3%, 1% or 0% O 2 ) for 16 or 48 h. Cells were lysed in buffer A [10 mm HEPES (ph 7.8), 10 mm KCl, 0.1 mm EDTA, 1mM dithiothreitol (DTT), and 0.1% Nonidet P- 40 (NP-40)] with protease inhibitors (5 g/ml aprotinin, 5 g/ml pepstatin, 5 g/ml leupeptin, 0.5 mm Pefabloc, and 1 mm phenylmethylsulfonyl fluoride) and phosphatase inhibitors (10 mm sodium fluoride, 1 mm sodium orthovanadate, and 20 mm glycerophosphate). Nuclear proteins were then extracted with buffer B [50 mm HEPES (ph 7.8), 420 mm KCl, 0.1 mm EDTA, 1 mm DTT, 5 mm MgCl2, and 20% glycerol], containing both protease and phosphatase inhibitors. Equal amounts of protein extracted from HPASMCs, incubated under normoxic and hypoxic (3%, 1% and 0% O 2 ) conditions for 16 and 48 h, were loaded on 4 12% Tris-glycine gels, separated by electrophoresis and transferred to nitrocellulose. Membranes were blocked with 5% milk overnight and probed with anti-hif-1 (1:500), anti-p21 (1:3,000), anti-p53 (1:1,000). In all cases, protein concentration was determined by the Bradford assay, and appropriate volumes of extraction buffer to produce constant protein loading in each lane were mixed with SDS loading buffer. Equality of protein loading and transfer efficiency were corroborated by full-lane densitometry of the Ponceau red-stained membranes. Immunoblots were probed with horseradish peroxidase-donkey anti-rabbit IgG (1:1,000 in blocking buffer) 104

121 Chapter 3 and visualized by enhanced chemiluminescence (ECL Plus kit, Amersham Biosciences). Band intensity was quantified by densitometry (Bio-Rad Laboratories, Mississauga, ON, Canada). Animal Studies: The effects of hypoxic exposure on pulmonary vascular smooth muscle cell proliferation and apoptosis in vivo were assessed in male Sprague-Dawley rats ( g). All protocols were in compliance with standards set by the Canadian Council of Animal Care, and were approved by the Institutional Animal Care Committee. Rats were placed in a Plexiglas chamber into which the flow of air and nitrogen was controlled independently. Rats exposed to hypoxia breathed a gas mixture containing 10% O 2 for 2 days, 7 days and 14 days. Normoxic control animals breathed room air under otherwise identical conditions. In preliminary experiments, the arterial PO 2 in rats breathing 10% O 2 averaged 38 Torr (range, Torr). At the end of the exposure period right and left main and first branch pulmonary arteries were excised immediately after decapitation, rinsed with PBS, fixed with paraformaldehyde, dehydrated in graded ethanol (70 100% ), cleared in xylene, and embedded in paraffin. Sections (5 m thick) were cut on an oscillating blade microtome (Leica, Wetzlar, Germany) and placed on coated glass microscope slides (Fisher Scientific, Pittsburgh, PA). Detection of apoptotic cells: DNA fragmentation was detected by TUNEL (DeadEnd Fluorometric TUNEL System, Roche). Slides containing paraffin-embedded sections were dewaxed, rehydrated, permeabilized with proteinase K, preincubated with equilibration buffer, and incubated with labeling solution (rtdt and nucleotide mixed with fluorescein-labeled dutp) for 1 h at 37 C. The reaction was terminated by incubating samples in a stopping buffer for 30 min. After PBS washes and counterstaining with TO-PRO-3, the samples were mounted and examined by laser confocal microscopy. 105

122 Chapter 3 Detection of proliferating cells: To assess the effects of hypoxia on pulmonary artery smooth muscle cell proliferation in vivo, bromodeoxyuridine (BrdU) uptake by these cells was assessed as a marker of de novo DNA synthesis. BrdU was infused subcutaneously using osmotic pumps (model no. 2ML-2; Alza Corp, Palo Alto, CA). Pumps containing 0.32 g of BrdU in 2 ml vehicle (0.4% DMSO) were implanted intrascapularly in normoxic and hypoxia-exposed rats 2 days, 7 days and 14 days before euthanasia. The rats received 0.4 mg BrdU/h delivered continuously. At the end of the labeling period, the right and left main and first branch pulmonary arteries were excised, fixed and embedded in paraffin. Dewaxed and rehydrated slides were incubated with 2 M HCl for 1 h at 37 C. The acid was neutralized by 0.1 M borate buffer, ph 8.5. Following PBS washes, the slides were incubated with fluorescein-labeled anti- BrdU antibody (Roche) for 1 h at room temperature, protected from light. Slides were then washed with PBS and mounted for confocal microscopy. Smooth muscle -actin staining: Immunohistochemical staining using anti alpha-smooth muscle actin antibody was carried out on the same slides to confirm localization of TUNEL and BrdU uptake to smooth muscle. Following TUNEL or BrdU labeling, sections were covered with blocking buffer (5% goat serum with PBS containing 1% BSA) for 30 minutes, following which they were stained with Cy3-conjugated monoclonal anti-smooth muscle -actin (1:200 in blocking buffer, Sigma) for 2 hours. Sections were counterstained with TO-PRO-3, washed with PBS, and viewed with a confocal microscope. Quantitative analysis: The number of TUNEL-positive, BrdU- and TO-PRO-3-labeled cells was determined in three sections from each of six animals from each experimental group. Four to six randomly selected microscopic fields ranging from cells per field were counted for each section, and the number of TUNEL or BrdU-positive nuclei was expressed as a 106

123 Chapter 3 percentage of the total number of nuclei. The diameter of a vessel (distance between external lamina) and the thickness of the tunica media (distance between the internal and external elastic lamina) were measured in main and first branch pulmonary artery sections stained for smooth muscle -actin, from digitized images using Image J analysis software (NIH 2002, Thickness was measured at equally spaced 10 points around the vessel wall and the percentage of medial wall thickness expressed as: % medial wall thickness = ([medial thickness x 2]/external diameter) x 100 Data from each animal in each group were averaged to serve as a single value for statistical analysis. Data Analysis: Data are presented as means ± SEM of n observations with P< 0.05 considered significant. The significance of differences between individual means was determined by two-tailed Student s t-test. Differences among multiple means were evaluated by analysis of variance (ANOVA) corrected for multiple measures, where appropriate, and, when overall differences were detected, differences between individual means were evaluated post hoc using the Student-Newman-Keuls procedure. 107

124 Chapter Results Figure 3.1A illustrates the effect of normoxic and hypoxic incubation on human pulmonary artery smooth muscle cell (HPASMC) number. After incubation at 5% O2 for 16 or 48 hours, there was a trend toward increased cell numbers which did not reach statistical significance. After incubation at 3% O2 for 16 or 48 hours, cell number was significantly increased, compared to the normoxic control condition. After incubation at 1% O2 for 16 hours cell number was increased compared to the normoxic control value, whereas after 48hrs cell number was reduced. Incubation at 0% O2 inhibited the increase in cell number at both 16 and 48hr. time points. As shown in Figure 3.1B, HPASMC viability (Trypan Blue exclusion) was not significantly affected by incubation at 10, 5, 3 or 1% O 2. Viability was decreased after incubation at 0% O 2 for 16 and 48 hours. 108

125 Chapter 3 Figure 3.1 (A) Effects of hypoxia on human pulmonary artery smooth muscle cell (HPASMC) numbers after incubation for 16 and 48 h. n= 6 per condition. *P <0.05 vs. corresponding normoxic control values. 109

126 Chapter 3 Figure 3.1 (B) Effects of hypoxia on human pulmonary artery smooth muscle cell (HPASMC) viability after incubation for 16 and 48 h. n= 6 per condition. *P <0.05 vs. corresponding normoxic control values. 110

127 Chapter 3 To confirm that PASMC proliferation was altered by hypoxic incubation the percentage of cells that incorporated BrdU after incubation at 3%, 1% and 0% O 2 for 16 and 48 hrs was assessed. As shown in Figure 3.2, the percentage of BrdU positive cells is increased after incubation at 3% O 2 for both 16 and 48 hours and at 1% O 2 for 16 hours. BrdU incorporation is decreased after incubation at 1% O 2 for 48 hours and at 0% O 2 for both16 and 48 hours. Figure 3.2 The % BrdU incorporated cells increased after incubation at 3% O 2 for 16 and 48 h, and after incubation at 1% O 2 for 16 h. BrdU incorporation decreased after incubation at 1% O 2 for 48 h and after incubation at 0% O 2 for 16 and 48 h. n= 6 per condition. *P <0.05 vs. corresponding normoxic control values. 111

128 Chapter 3 Since the effects of hypoxic incubation on PASMC cell numbers may also reflect changes in the rate of cell death, the percentage of PASMCs staining positive for Annexin V, Propidium iodide, activated caspases and JC-1 monomers was recorded after normoxic and hypoxic incubation and are presented in Figures 3.3 A-E respectively. The percentage of viable cells (staining negative for both Annexin V and PI) was unaffected by incubation at 3% and 1% O 2 but decreased significantly after incubation at 0% O 2. Similarly, the percentage of cells positive for activated caspases and for JC-1 monomers are unchanged after incubation at 3% or 1% O 2 but increased after incubation at 0% O 2 for 16 h and increased further after 48h. 112

129 Chapter 3 21% O 2, 16hrs 0% O 2, 16hrs Propidium iodide Propidium iodide 21% O 2, 48hrs 0% O 2, 48hrs Annexin V Figure 3.3 (A) Representative flow cytometric plots for Annexin V/PI staining after incubation at 0% O 2, 16 and 48 h, compared with the normoxic cells; n= 6. *P<0.05 vs. corresponding normoxic control values. 113

130 Chapter 3 Figure 3.3 (B) Bar graphs representing reduced number of Annexin V and PI negative cells (live cells) after incubation at 3%, 1% and 0% O 2, 16 and 48 hours, compared with the normoxic cells (open bars); n= 6. *P<0.05 vs. corresponding normoxic control values. 114

131 Chapter 3 21% O 2, 16hrs 0% O 2, 16hrs Counts 21% O 2, 48hrs 0% O 2, 48hrs FITC-Caspase Figure 3.3 (C) Representative flow cytometric plots for increased caspase activity after incubation at 0% O 2, 16 and 48 hours, compared with the normoxic cells (open bars); n= 6. *P<0.05 vs. corresponding normoxic control values. 115

132 D Chapter 3 E Figure 3.3 (D) Increased caspase activity and (E) increased 5,5,6,6 -tetrachloro- 1,1,3,3 -tetraethyl-benzimidazolylcarbocyanine iodide (JC-1) monomer formation (indicating mitochondrial membrane depolarization) after incubation at 0% O 2 compared with the normoxic cells (open bars); n=6. * P<0.05 vs. corresponding normoxic control values. 116

133 Chapter 3 To determine if hypoxia exerts a general effect on cell cycle progression or affects a specific cell cycle phase, DNA content was quantified by PI staining to assess the percentage of cells in G0/G1, S and G2/M phases. As illustrated in figure 3.4, incubation of HPASMCs at 3% O 2 (16 and 48hrs) decreased the percentage of cells in G0/G1 phase and increased the percentage of S-phase cells with a corresponding increase in the percentage of cells undergoing mitosis. This suggests acceleration of progression through the G1/S interphase. A similar distribution is observed after incubation at 1% O 2 for 16 hrs. After incubation at 1% O 2 for 48hrs and at 0% O 2 for 16 and 48hrs, however, there is accumulation of G0/G1phase cells and depletion of cells in S and G2/M phases, indicating a delay in transition through the G1/S interphase. 117

134 Chapter 3 Figure 3.4 Flow cytometric analysis of propidium iodide stained normoxic (N) and hypoxic (3%, 1% or 0% O 2 ) cells. Incubation at 3% O 2 16 and 48 h and 1% O 2 16 h increases the percentage of cells in G2/M, whereas incubation at 1% O 2 48 h and 0% O 2 16 and 48 h increases the percentage of cells at the G0/G1 interphase. n = 6, * P<0.05 vs. corresponding normoxic control values. 118

135 Chapter 3 p53 and its transcriptional target p21 are inhibitory regulators of cell cycle progression and are involved in hypoxia induced growth arrest in cancer cells [144]. Expression of these nuclear proteins were therefore, used as biochemical indices of the proliferative response. HIF- 1 nuclear translocation reflects activation of the cellular response to hypoxia and its nuclear levels were measured to confirm that the cells were responding, at, a molecular level, to the hypoxic stimulus. Levels of both p21 (Figure 5A) and p53 (Figure 5B) are increased after incubation at 0% O 2 but did not differ significantly from the normoxic control values after incubation at 3% O 2 or 1% O 2. Nuclear HIF-1 levels (Figure 5C) increased from the normoxic values after incubation at 3%, 1% and 0% O

136 Chapter 3 Figure 3.5 (A) Nuclear levels of p21 protein after incubation of HPASMCs under normoxic and hypoxic (3, 1 or 0% O 2 ) conditions. n= 6. *P <0.05 vs. corresponding normoxic control values. 120

137 Chapter 3 Figure 3.5 (B) Nuclear levels of p53 protein after incubation of HPASMCs under normoxic and hypoxic (3, 1 or 0% O 2 ) conditions. n= 6. *P <0.05 vs. corresponding normoxic control values. 121

138 Chapter 3 Figure 3.5 (C) Nuclear levels of HIF-1 protein after incubation of HPASMCs under normoxic and hypoxic (3, 1 or 0% O 2 ) conditions. n= 6. *P <0.05 vs. corresponding normoxic control values. 122

139 Chapter 3 As shown in Figure 3.6, intracellular ATP concentration remained unchanged in human pulmonary artery smooth muscle cells incubated at 3% O 2. After 48 hours of incubation at 1% O 2, ATP levels decreased (7.6 ± 2.8 % decrease from normoxic control value), however this change did not reach statistical significance (p = 0.096). In cells incubated at 0% O 2 cellular ATP concentration was reduced after 16 h and decreased further after 48 h (17.38 ± 4.5% and 22.5 ± 3.8% decrease vs. corresponding normoxic control values, respectively, p < 0.05 for both). Figure 3.6 Cellular ATP concentration is decreased after incubation at 0% O 2 compared with the normoxic cells (open bars). n= 6. *P<0.05 vs. corresponding normoxic control values. 123

140 Chapter 3 The effects of hypoxic exposure on pulmonary artery smooth muscle cell apoptosis and proliferation in rats was assessed to evaluate the effects of physiologically relevant levels of hypoxia on the pulmonary vasculature in vivo. TUNEL staining of pulmonary artery sections (Figure 3.7, A and B) was carried out after 2, 7 and 14 days of hypoxic exposure (10% FiO 2 ). In the main pulmonary artery and first branch pulmonary artery the apoptotic index was significantly increased after 7 days of hypoxic exposure and returned to normoxic control values after 14 days. Figure 3.8, A and B illustrate the effects of in vivo hypoxia for 2, 7 and 14 days on thickness of the pulmonary artery smooth muscle layer and BrdU incorporation. There is maximum BrdU incorporation after 7 days with a subsequent decrease in the number of BrdU staining cells to normoxic levels after 14 days, mirroring the effects of apoptosis. As illustrated in Table 3.2, medial wall thickness was increased after 7 days and remained at this level after 14 days of continuous hypoxic exposure. 124

141 Chapter 3 Normoxic Hypoxic 2 days 7 days 14 days Pulmonary artery branch Pulmonary artery (Main) Figure 3.7 (A) Representative images of TUNEL in paraffin-embedded sections of normoxic and hypoxic (2 days, 7 days, 14 days) rat pulmonary artery and pulmonary artery branch (x40 magnification). 125

142 Chapter 3 Figure 3.7 (B) Quantitative analysis of TUNEL positive cells; n=6 rats per group. *P<0.05 vs. corresponding normoxic control values. 126

143 Chapter 3 Normoxic Hypoxic 2 days 7 days 14 days Pulmonary artery branch Pulmonary artery (Main) Figure 3.8 (A) Representative images of immunohistochemical staining of incorporated BrdU (green) and -smooth muscle actin (red) in paraffin-embedded sections of pulmonary artery and pulmonary artery branch from normoxic and hypoxia (2 days, 7 days, 14 days) exposed rats (x40 magnification). 127

144 Chapter 3 Figure 3.8 (B) Quantitative analysis of BrdU positive cells; n=6 rats per group. *P<0.05 vs. corresponding normoxic control values. Table 3.1 Medial wall thickness (% ) of pulmonary artery and pulmonary artery branch from normoxic and hypoxia (2 days, 7 days, 14 days) exposed rats; n=6 rats per group. *P<0.05 vs. corresponding normoxic control values. Main Pulmonary artery Branch pulmonary artery Normoxic 5.8± ± day hypoxic 6.1± ± day hypoxic 10.4±0.9* 9.4±0.8* 14 day hypoxic 11.1±0.1* 9.9±0.5* 128

145 Chapter Discussion The main findings of this study are that: in HPASMCs in culture 1) incubation at 3% O 2 for 16 and 48hrs and at 1% O 2 for 16hrs increases cell number DNA synthesis and the percentage of cells in S and G2/M phases of the cell cycle; 2) proliferation and DNA synthesis are inhibited after incubation at 1% O 2 for 48hrs and at 0% O 2 for 16 and 48hrs with a corresponding accumulation of cells in G1 phase; 3) the prevalence of apoptosis and nuclear levels of the antiproliferative, proapoptotic factors p21 and p53 are unaffected by incubation at 3% and 1% O 2 but are increased after incubation at 0% O 2 ; and 4) cellular ATP levels are reduced, and the mitochondrial membrane is depolarized after incubation at 0% O 2. In the pulmonary artery of rats breathing 10% O 2 (arterial po 2 40 Torr), smooth muscle cell proliferation and apoptosis are increased during the first seven days of hypoxic exposure with the net effect being accumulation of smooth muscle in the vessel wall. A hallmark of hypoxic pulmonary hypertension is medial thickening in larger pulmonary arteries and muscularization of distal arteries not normally invested with smooth muscle [346, 347]. Hypoxia has long been known to stimulate PASMC proliferation in vivo although in most models, including that used in the current study, this is a transient effect being maximal after 7 days and then subsiding [348]. The mechanisms, by which the proliferative response is activated and, subsequently, regulated, are unresolved [349]. Hypoxia affects PASMC proliferation even in the absence of other cell types; therefore, although mitogens and cytokines from surrounding cells may modulate the response in vivo, both the sensory and effector mechanisms that underlie these effects are localized to the smooth muscle cell [323]. Studies of the influence of hypoxia on PASMC proliferation in vitro have yielded contradictory results [273, 276, 329, , 340, 341, 350]. Consequently, it remains 129

146 Chapter 3 unclear whether hypoxia has a direct mitogenic effect on PASMCs or acts as a comitogen acting in concert with hypoxia-induced release of proproliferative factors from adjacent endothelial cells or fibroblasts. Studies in which a positive correlation between acute hypoxia and PASMC proliferation was observed tended to use moderate levels of hypoxia (3 5% O 2 ) [276, , 336, 338, 350], whereas studies in which hypoxia caused a decrease in PASMC proliferation were carried out under conditions of severe hypoxia or anoxia (< 2% O 2 ) [329, 339, 340]. For example, Ambalavan et al. showed that swine proximal PASMC proliferated at oxygen concentrations of 5 10% [273, 332], whereas, Stotz et al. showed a 5 10% increase in proliferation rates of rat pulmonary microvascular SMC at 1% O 2 [338, 341]. Our current findings reconcile these discrepancies since they demonstrate that, rather than being due to variations in experimental conditions or species, hypoxia of differing severity may exert opposing effects depending on the capacity of these cells to maintain cellular energy status under the specific conditions studied. Previous studies have shown hypoxia induced G1 arrest in cell cycle of HPASMCs to be mediated by CDK inhibitor p21. In contrast, others have reported that cell cycle arrest at late G 1 is caused by p27 expression under severe hypoxia [142, ]. Li et al found that the oxygendependent checkpoint of the cell cycle is controlled by p27 expression, and that camp signaling also interferes with the cell cycle and p27 expression [354]. However, the precise mechanisms and interactions between the pathways activated by hypoxia, as well as the antiproliferative effects of p27 or p21 during hypoxic exposure in HPASMCs remain uncertain [355]. Our finding that HPASMC incubation at 1% O 2 for 48hrs and 0% O 2 for 16 and 48hrs, causes G 1 /S arrest is in line with those of others [258]. In addition, we also show that the directionally opposed effects of hypoxia on HPASMCs, proliferation or growth arrest, are integrated by events at G1/S. 130

147 Chapter 3 Proliferation of HPASMCs requires these normally quiescent cells to enter the cell cycle. The most important molecular process for cell cycle progression is retinoblastoma protein phosphorylation by cyclin-dependent kinase (CDK)-cyclin complexes, and CDK activities are mainly regulated by CDK inhibitors such as p21 and p27 [356]. In HPASMCs, both p53 and its transcriptional activation target p21, which inhibits G 1 /S transition, are increased after incubation at 1% O 2 48hrs and at 0% O 2, but remain unchanged at 3% O 2 and at 1% O 2 16hrs. HIF levels are also increased in HPASMCs post incubation at 3, 1 and 0% O 2, although the cell number increased at 1% O 2 16 hrs and cell number decreased at 1% O 2 48hrs. This indicates HIF-1 -independent regulatory mechanisms causing the directionally opposed proliferative responses of hypoxia. HIF-1 is the major transcription factor induced under hypoxia, and is known to regulate pathways that both enhance cell survival as well as induce cell death. Many cell cycle-associated genes containing functional hypoxia regulatory elements have been identified [ ]. Considering these, it is unlikely that there is no HIF-1 involvement at all in affecting the proliferative responses of HASMC or HPASMC. Further studies are needed to understand the possible interaction of HIF-1 with other coregulatory molecules. Apoptosis plays an important role in cell number control in various tissues and organs by balancing cell growth and multiplication. The remodeling in pulmonary vascular structure is mainly caused by imbalanced proliferation and apoptosis in pulmonary artery smooth muscle cells [ ]. An increase in PASMC proliferation and a decrease in PASMC apoptosis could concurrently mediate thickening of the pulmonary vasculature, which subsequently reduces the lumen diameter of pulmonary arteries, increasing pulmonary vascular resistance. It has been demonstrated that increased PASMC proliferation and/or inhibited PASMC apoptosis both 131

148 Chapter 3 contribute to induce pulmonary vascular medial thickening [360, 361], and this acquisition of resistance to apoptosis along with increased rates of VSMC proliferation appear to be necessary for neointima formation [ ]. The role of p53 in mediating VSMC apoptosis has been proposed [ ], though p53 alone does not induce apoptosis in either normal human or rat VSMCs in vitro or in vivo unless the cells are primed to die [ ], or massive expression is induced via adenovirus vectors [372]. The precise mechanisms involved in the regulation of PASMC proliferation and apoptosis in PAH are still incompletely understood. The precise level of hypoxia that induces apoptosis in HPASMCs, has also not been defined. The results of our apoptosis assays (Figure 3.3 A-E) show hypoxia induced cell death in HPASMCs only after incubation at 0% O 2. Viability is maintained at 3% O 2 and at 1% O 2. Although the cell number is reduced in HPASMCs after incubation at 1% O 2 48hrs, there is no change in apoptosis compared to normoxic condition. Together these results suggest that hypoxia in HPASMCs enhances smooth muscle cell survival, at levels above those that cause cytotoxicity. In the pulmonary circulation, this protection from apoptotic cell death and enhanced smooth muscle cell proliferation at moderate levels of hypoxia may be important in neovasoformation during development and in ischemic tissues, where growth of new endothelial networks by regulatory smooth muscle cells must occur under conditions of relative oxygen deficiency. Depolarization of the mitochondrial membrane only at 0% O 2 where apoptosis is increased suggests the involvement of the mitochondria mediated intrinsic pathway of apoptosis. Besides, the enhanced expression of p53 (at 0% O 2 ) could also play a role in inducing PASMC apoptosis. When the ratio of energy supply to energy demand decreases, as during severe hypoxia, homeostatic mechanisms attempt to match ATP production to ATP utilization [373, 374]. The 132

149 Chapter 3 hypothesis that changes in energy state signal pulmonary vasomotor responses to hypoxia has been considered by many investigators, but the role of energy state in these responses remains unclear [375]. In fibroblasts and tumor cells, severe hypoxia causes ATP depletion thus impairing DNA repair, but the role of intracellular ATP concentration ([ATP] i ) in determining the hypoxic response of HAPSMCs has not been studied earlier [376, 377]. [ATP] i in HPASMCs is maintained at normoxic levels after incubation at 3 and 1% O 2 and reduced after incubation at 0% O 2. ATP depletion is concurrent with reduced cell number and enhanced apoptosis and depolarization of the membrane potential suggesting that under hypoxia, HPASMCs continue to proliferate and/or survive as long as the [ATP] i is maintained. However, if the level of hypoxia is sufficient to cause depletion in intracellular ATP, the HPASMCs can no longer maintain the mitochondrial membrane potential, and the cells apoptose. As the duration of hypoxia is prolonged permanent structural remodeling occurs in the vasculature. This reflects changes in structure and biochemical phenotype of all of the cells that compose the pulmonary arteries. Universally observed is medial thickening and the appearance of SMC- like (based on -actin staining) cells in previously nonmuscularized vessels. Our in vivo data supports previous findings. In the current study, proliferation and apoptosis are both enhanced in the main and branch pulmonary artery after 2 days and 7 days of hypoxia and decreases after 14 days. This concurrent increase in cell proliferation as well as programmed cell death confers plasticity to the vessel wall, whereby required structural change is enabled while avoiding accumulation of mutations or abnormal vascular structure formation. 133

150 Chapter 4 Chapter 4 General Discussion and Conclusions 134

151 Chapter 4 VSMC proliferation has been recognized as central to the pathology of several cardiovascular diseases, such as atherosclerosis and hypertension. Hypoxia is an important regulator of physiologic processes, including erythropoiesis, angiogenesis, and glycolysis. In the vasculature, chronic hypoxia has been shown to cause proliferation of VSMCs, leading to vessel wall remodeling, a key pathophysiologic component of pulmonary hypertension. The mechanisms by which hypoxia regulates VSMC growth include direct cell cycle-specific effects, as well as indirect effects, via the regulation of VSMC mitogen production by neighboring endothelial cells. Hypoxia triggers a cellular adaptive response that is primarily mediated by the transcription factor hypoxia-inducible factor 1 (HIF-1). Expression of HIF-1 target genes serves to maintain cellular homeostasis. Transcriptional activation of hypoxia-responsive genes represents one major component of the vascular cell hypoxic response; however, the mechanisms regulating VSMC proliferation and survival in the vessel wall systemic or pulmonary, under hypoxia, remain to be elucidated. The normoxic pulmonary circulation is vasodilated and accommodates the entire cardiac output at much lower pressures than the systemic circulation. During hypoxia, the pulmonary arteries constrict, whereas systemic arteries, such as the aorta dilate. The mechanism of this opposing control of tone between the two vascular beds is unknown. Although the response of each bed to hypoxia is significantly modulated by the endothelium, the mechanism for the opposing responses to hypoxia appear to lie within the VSMCs. Hypoxia increases intracellular Ca 2+ and contracts PASMCs; in contrast, SMCs from systemic arteries display decreased intracellular Ca 2+ and relax in response to hypoxia. The present series of investigations compares the oxygen regulation of HASMC and HPASMC proliferation and survival. 135

152 Chapter 4 In both HASMCs and HPASMCs, incubation at various oxygen concentrations caused directionally opposite responses proliferation and growth arrest along with or without apoptosis, depending upon the ability of these cells to maintain intracellular ATP concentration. 3% O 2 incubation in both HPASMCs and HASMCs trigger a proliferative response without any change in viability compared to normoxic cells. 0% O 2 incubation diminishes cell numbers and enhances apoptosis in both HPASMCs and HASMCs. However, the difference in the hypoxic response of these two cell types is in the response at 1% O 2 incubation: (i) In HASMCs cell number is reduced after 16hrs and even more after 48hrs, while in HPASMCs, cell number increases after 16hrs and decreases only after 48hrs; (ii) Incubation at 1% O 2 in HASMCs, at both 16 and 48hrs, reduces cell viability, increases apoptosis, depolarizes the mitochondrial membrane potential and depletes intracellular ATP concentration. But in HPAMCs even though the number of cells was reduced at 1% O 2 for 48hrs, the viability along with mitochondrial membrane potential and intracellular ATP concentration is maintained. Our study has shown that hypoxia can increase or decrease cell number within the same cell type (systemic or pulmonary smooth muscle cells), and the fundamental differences in response depend on the severity of hypoxia that the cell is exposed to. The above responses also suggest that HPASMCs are better adapted to reduced oxygen concentrations than their systemic counterpart. In HASMCs an oxygen concentration of 1% O 2 16hrs was sufficient to cause a delay in cell cycle progression, with increased cells in G 1 /S phase. In HPASMCs, incubation at 1% O 2 for 16 hrs caused an opposite effect there is enhanced progression of cell cycle, with increased cells in G 2 /M and S phases. But even in HPASMCs, incubation at 1% O 2 for 48 hrs causes a delay in cell cycle progression, with increased cells in G 1 /S phase. Together the data suggest that 136

153 Chapter 4 in both HASMCs and in HPAMCs, the bidirectional effects of hypoxia are integrated by events occurring at the G 1 /S checkpoint, though the mechanism/s remain unclear. In HASMCs at levels where hypoxia causes a decrease in cell number (1% and 0% O 2 ) our results have shown apoptosis to be a contributing factor towards the decrease. However, in HPASMCs, at 1% O 2 for 48hrs, no apoptosis is observed though there is a decrease in cell number at this point. These results along with viability studies in HPASMCs, suggest that oxygen concentrations which are capable of inducing apoptosis in systemic arterial smooth muscle cells are not sufficiently hypoxic to induce cell death in pulmonary SMCs. In both the pulmonary and the systemic circulation, protection from apoptotic cell death and enhanced smooth muscle cell proliferation at moderate levels of hypoxia may be important in neovasoformation during development and in ischemic tissues, where investiture of new endothelial networks by regulatory smooth muscle cells must occur under conditions of relative oxygen deficiency. Possibly since pulmonary circulation carries deoxygenated blood, the smooth muscle cells of this region are genetically primed to survive under oxygen concentrations that are otherwise toxic to the systemic smooth muscle cells. Reduced sensitivity of pulmonary smooth muscle cells to hypoxia induced cell death provides the pulmonary vasculature an added advantage. Dead cells are generally replaced by fibrous or connective tissue, which would have otherwise impaired gas exchange and caused an early onset of pulmonary hypertension. Enhanced apoptosis in both HASMCs and HPASMCs under severe hypoxia is also an adaptive response that helps remove replication errors and prevents accumulation of hypoxia-induced mutations. 137

154 Chapter 4 Conclusions Systemic vascular smooth muscle cells experience a broad range of oxygen tensions under physiological (90 mmhg in aortic lumen; 20 mmhg at 150 m depth; 24 mmhg in terminal arterioles) and pathological (atherosclerosis, <7 mmhg) conditions. In earlier studies hypoxia has had an inconsistent effect on smooth muscle cell proliferation and survival. This lack of consistency has prevented the development of a unifying hypothesis regarding oxygen regulation of these processes in diseases in which arterial remodeling plays a significant pathophysiological role. Since cell replication is highly energy dependent it is intuitive, though unproven, that hypoxia may have different effects depending on the degree to which cellular energy status is compromised. On the other hand, many cell cycle associated genes are known to be induced at levels of hypoxia at which inhibition of ATP synthesis would not be predicted and a regulatory role beyond mere cytotoxicity merits consideration. This study was carried out to determine if different proliferative and apoptotic responses are elicited in human aortic smooth muscle cells (HASMCs) and human pulmonary artery smooth muscle cells (HPASMCs) subjected to hypoxic incubation under conditions which do or do not result in cellular ATP depletion, whether these effects are relevant to vascular remodeling during hypoxia in vivo and to identify potential regulatory pathways using gene expression profiling in cells exposed to conditions that elicit discordant responses. The novel findings of our study in HASMCs, presented in Chapter 2, are: 1. Hypoxia may enhance or inhibit HASMC proliferation and differences in the cellular response are elicited by varying degrees of hypoxic stress. 138

155 Chapter 4 2. The difference between O 2 concentrations that enhance SMC proliferation and those that impair cellular energy status and trigger cell destruction is small. 3. Pro- and antiproliferative and pro- and antiapoptotic gene expression are tightly coordinated in HASMCs during hypoxic incubation to produce directionally opposed responses, effected at the level of G 1 /S transition. 4. Our in vivo studies show increases in both SMC proliferation and death. This enhanced cell turnover confers plasticity sufficient to enable the required structural change while avoiding accumulation of mutations or abnormal vascular structure formation. 5. The net effect is an increased muscularity of the arterial wall, which is possibly a compensatory adaptation that preserves the capacity to regulate the systemic circulation. Hypoxic pulmonary arterial hypertension occurs in many cardiopulmonary diseases. Increased smooth muscle cell proliferation in the pulmonary arterial wall is a hallmark of this condition. The mechanism/s regulating these processes are yet unknown. The novel findings of our study in HPASMCs, presented in Chapter 3, are: 1. HPASMCs proliferated at O 2 concentrations which inhibited cell growth in HASMCs. 2. HPASMCs did not undergo apoptosis, maintained intracellular ATP concentration and mitochondrial membrane potential at O 2 concentrations which were cytotoxic to HASMCs. Together, our data suggest that the response to hypoxia, in either HASMCs or HPASMCs, depends on whether the hypoxic stress is sufficient to impair cellular energy status. In HPASMCs, ATP is preserved at a lower O 2 concentration and this permits the cells to maintain viability and proliferation at more severe levels of hypoxia than are tolerated by HASMCs. The differences in the hypoxic response could be a result of adaptation to exposure to 139

156 Chapter 4 the deoxygenated blood that the pulmonary circulation carries or due to differences in the embryonic origins of the cells that comprise the two circulations. In vivo, hypoxia induces a state of enhanced cell turnover and, hence, plasticity. Vital organ function is highly intolerant of oxygen deprivation. Accordingly, mechanisms linking vascular cell turnover and the capacity for rapid structural change directly to oxygen concentration are required to avoid delays inherent in second messenger signaling. Our in vitro studies show that systemic and pulmonary artery smooth muscle cells respond differently to hypoxia. HPASMCs are able to proliferate and remain viable at oxygen concentrations that induce growth arrest and/or apoptosis and in HASMCs. However, under reduced oxygen concentrations in both the cell types, systemic or pulmonary, cell viability is maintained until intracellular ATP levels are depleted. Our data do not address the mechanism(s) by which information about the site of origin of SMCs from systemic and pulmonary circulations is preserved from generation to generation (passage to passage) in in vitro cultures. A possible explanation for the different responses to hypoxia of the SMC when coming from two different sources (systemic vs. pulmonary) could be attributed to their diverse embryological origin. Arteries cranial to the heart are mostly products of the paired aortic arches, which course axially within the branchial arches, thus interconnecting the ventral aorta with paired dorsal aortas. The fourth pair of embryonic aortic arches becomes the aortic arch in the adult while the fifth pair forms the pulmonary arteries and ductus arteriosus. The dorsal aortas fuse into the single descending aorta [378]. During development of multicellular eukaryotic organisms, differences in genetic programming cause the cells to differentiate into different types, which perform different functions, and respond differently to the environmental stimuli. Thus, genes are silenced or 140

157 Chapter 4 activated in an epigenetically heritable fashion, giving cells a "memory" which determines their phenotype over subsequent cell divisions. This implies that in order to grow and maintain a specific, lineage-restricted state, such gene expression configurations need to be transmitted to daughter cells, without mutations of the DNA. Two types of epigenetic modifications can regulate chromatin conformation, thereby also regulating the transcriptional activity or silencing of specific genomic regions. These include (1) DNA methylation at cytosine residues of CpG dinucleotides in gene promoters, transposons, and imprinting control regions. In most cases, DNA methylation is associated with gene repression; and (2) histone modifications in the chromatin organization, such as methylation and acetylation, affecting the N-terminal tail of histones [379]. Studies to define which mechanisms are active in distinguishing the responses of PASMCs from SMCs derived from the systemic circulation are now indicated. Insight into the molecular mechanisms underlying these responses would aid in the development of pharmacological tools to modulate vascular remodeling in disease states. 141

158 Chapter 5 Chapter 5 Future directions 142

159 Chapter 5 Our studies in HASMCs (chapter 2) and HPASMCs (chapter 3) have shown that (i) incubation of these cells at varying oxygen concentrations cause divergent responses cell proliferation and enhanced cell survival on one hand and growth arrest and/or apoptosis at further reduced oxygen concentrations; (ii) cell viability under hypoxia remains unaffected as long as the cells are able to maintain their intracellular ATP concentration. Oxygen concentrations, at which intracellular ATP levels are diminished, cause the cells to undergo cell cycle growth arrest (at the G 1 /S interphase) and cell apoptosis; (iii) HPASMCs proliferated at oxygen concentrations which inhibited cell growth in HASMCs and (iv) HPASMCs did not undergo apoptosis and maintained their intracellular ATP levels and mitochondrial membrane potential at oxygen concentrations that are sufficient to deplete ATP and induce apoptosis in HASMCs. Based on these, it could be proposed that in HPASMCs, ATP is preserved at a lower O 2 concentration to enable the cells to maintain their viability, compared to HASMCs. In other words this difference/s in energetics in HASMCs and HPASMCs may play a key role in determining oxygen regulation of their proliferation and survival. The pulmonary circulation is adapted to function at a lower oxygen concentration compared to the systemic circulation. There is greater heterogeneity in the phenotype of pulmonary vascular SMCs compared to the systemic SMCs [325]. This varies with the size and location of the pulmonary artery (Table 5.1). K + channels are differentially distributed and there are cell-specific differences in the endothelin-1 (ET-1) system [380]. The vasoconstrictor response to ET-1 is mediated by ET A receptors in large pulmonary arteries but by ET B -like receptors in smaller muscular pulmonary arteries [381]. 143

160 Chapter 5 Table 5.1 Phenotypic heterogeneity in pulmonary vascular smooth muscle cells. Feature Main Pulmonary artery smooth muscle cells Distal Pulmonary artery smooth muscle cells Location Subendothelial Middle Media Outer media 3000μ 1500μ μ L1 L2 L3 L4 L3 L4 Morphology Round, small, irregularly shaped, interspersed in fragmented elastin Elongated, spindle shaped, oriented circumferentially between well developed and continuous elastic lamellae Large, spindle shaped cells, arranged in compact cell clusters, oriented longitudinally within vessel wall devoid of elastic lamellae Small, spindle shaped cells, oriented circumferentially in intersitial areas between L3 large cell clusters Same as Main Pulmonary Artery -SMA Calponin SM-MHC Metavinculin Desmin Caldesmon ND ND Hypoxia SMA: Smooth muscle actin, MHC: Myosin heavy chain, +: Positive, -: Negative, ND: Not detected, ++: Increase in proliferation. 144

161 Chapter 5 Significant mitochondrial diversity exists between pulmonary and systemic arteries, with direct imaging of rat vascular SMCs showing that PASMC mitochondria are more depolarized than ASMC mitochondria. Compared to systemic arteries, pulmonary artery mitochondria have higher rates of superoxide anion and H 2 O 2 production, lower expression of proximal electron transport chain components and higher expression of mitochondrial superoxide dismutase [326]. A deficiency in oxidative metabolism under hypoxic conditions will cause a cell to resort to glycolytic ATP production. Increased activity of rate-limiting glycolytic enzymes has been found in fibroblast, endothelial and kidney cell lines cultured under hypoxia. Considering these, a relevant question to address in future would be whether metabolic differences between systemic and pulmonary artery smooth muscle cells determine the differences in oxygen regulation of their proliferation and survival. It is possible that HPASMCs possess a greater aerobic capacity and/or an enhanced anaerobic capacity under hypoxia that allows these cells to maintain the intracellular ATP concentration and survive at levels of hypoxia that are cytotoxic to HASMCs. Preliminary studies were conducted to assess both the aerobic and anaerobic capacities of HASMCs and HPASMCs (Supplement). Real-time PCR was performed on DNA extracted from normoxic and hypoxic HASMCs and HPASMCs for three nuclear-specific genes in the promoter region, enos (endothelial nitric oxide synthase) promoter, inos (inducible nitric oxide synthase) promoter, VCAM1 (vascular cell adhesion molecule 1) promoter, and three mitochondrial-specific genes, mttrnl1 (mitochondrially encoded trna leucine 1), CO2 (cytochrome c oxidase II), and Dloop (non-coding region in mtdna controlling DNA replication). Mitochondrial copy number was assessed as an index of oxidative phosphorylative capacity in these cells (Supplementary Figure 1). To measure anaerobic glycolysis, protein 145

162 Chapter 5 levels of two key glycolytic enzymes namely Enolase and Phosphoglycerate kinase (PGK) (Supplementary Figure 2 A-D) and lactate concentration (Supplementary Figure 2 E and F) have been compared in normoxic and hypoxic HPASMCs and HASMCs. The effect of hypoxic incubation on mitochondrial DNA levels is illustrated in Supplementary Figure 1. In both HPASMCs and HASMCs, mitochondrial DNA levels decrease after hypoxic incubation. The decline correlates with the severity of the hypoxic exposure and is similar in the two cell types. Copy number values generated from standard curves (using plasmid DNA and genomic DNA) were also compared and similar results were obtained. Each human cell contains several hundreds to thousands of mitochondria and each mitochondrion has 2 to 10 copies of mitochondrial DNA (mtdna). The copy number of mtdna may reflect the abundance of mitochondria in a human cell [382]. Mitochondrial biogenesis is a highly regulated process and occurs on a regular basis in healthy cells, where it is controlled by the nuclear genome. Alteration of mitochondrial biogenesis and increased expression of nuclear genes encoding mitochondrial proteins are responses triggered by mitochondrial dysfunction or high energy demands found in pathophysiological conditions [383, 384]. The decrease in mtdna copy number has been observed in lung cancer, hepatocellular carcinoma and gastric cancers [385, 386]. Rapidly increased neuronal mitochondrial biogenesis has been observed after hypoxic-ischemic brain injury [387]. Mitochondrial biogenesis has not been studied earlier in HASMCs or HPAMSCs after incubation under various oxygen concentrations. Our results show that in both HASMCs and HPASMCs, there is a decrease in mitochondria. Since oxygen serves as a substrate in aerobic ATP production, it is quite obvious that under limited oxygen supply both the cell types would choose to reduce their aerobic capacity, and possibly switch their energy supply from aerobic metabolism to anaerobic glycolysis. The characteristics, including 146

163 Chapter 5 intron-less, without binding to histones, and inefficient mtdna proof-reading and DNA repair system, render mtdna more susceptible to oxidative damage than nuclear DNA [388]. Reduced mitochondria number could also be an adaptive response in both the cell types to prevent accumulation of mutations. ATP is the immediate source of metabolic energy whose hydrolysis causes protein synthesis, muscle contraction and ion transport across cell membrane. In the outer mitochondrial membrane, ATP phosphorylates creatine (Cr) by the creatine kinase (CPK) reaction: [ATP] + [Cr] [PCr] + [ADP] + [H + ] The phosphocreatine (PCr) shuttle carries the high energy bonds from the mitochondria to the sites of utilization where the CPK reaction is reversed (Lohman reaction) forming ATP and Cr [389]. Future studies should assess the ratios of PCr ([PCr]) or inorganic phosphate ([P i ]) concentration to [ATP] or creatine ([Cr]) concentration ([PCr]/[ATP], [P i ]/[ATP], and [PCr]/[Cr]), phosphorylation potential ( = [ATP]/[ADP][P i ], where [ADP] is ADP concentration). It is possible that HPASMCs have a greater potential to maintain its ATP through the Lohman reaction as above. Hypoxia is known to up regulate expression of several glycolytic enzymes to enhance anaerobic ATP production. Protein levels of glycolytic enzymes, PGK1 and enolase are increased after hypoxic incubation, in both HPASMCs (Suppl Fig 2 A and C) and HASMCs (Suppl Fig 2 B and D), suggesting enhanced glycolytic ATP production under hypoxic conditions in both cell types. Glycolysis in the cytoplasm produces the intermediate metabolite pyruvate. Under aerobic conditions, pyruvate is converted to acetyl CoA to enter the Kreb s cycle. Under anaerobic conditions, pyruvate is converted by lactate dehydrogenase to lactic acid. In aqueous solutions, lactic acid dissociates to lactate and H +. These H + can be used in the 147

164 Chapter 5 production of ATP by oxidative phosphorylation. Impairment of oxidative pathways, however, during lactate production results in a net gain of H + and acidosis occurs [390]. The lactate levels reflect the balance between lactate production and clearance. The normal plasma lactate concentration is mmol litre -1, while concentrations >5 mmol litre -1 are considered high enough to cause acidosis. In vascular smooth muscle, changes in intracellular ph (ph i ) can alter membrane potential, calcium homeostasis, and myosin light chain kinase activity [ ]. Altered ph i has been shown to change vasomotor tone in both systemic and pulmonary arteries and responses to hypoxia in isolated lungs [ ]. Lactate concentrations are increased under hypoxia in both HASMCs and HPASMCs. Incubation at 1% O 2 causes lactate to accumulate more (>5 mmol litre -1 ) in HASMCs than HPASMCs. A possible explanation could be that HPASMCs possess a better ability to prevent lactate accumulation and thereby acidosis under hypoxic conditions. Likely, this excess lactate accumulation, beginning at higher oxygen concentrations in HASMCs (1% O 2 ) compared to that in HPASMCs (0% O 2 ) that make the aortic cells more vulnerable to acidosis induced cell apoptosis than the pulmonary cells. Future studies will include determining the lactate/pyruvate ratio and rate of extracellular acidification in both HASMCs and HPASMCs. Factors other than enhanced lactic acid production might contribute to the fall of arterial ph i during hypoxia. Decreased mitochondrial electron transport and proton pumping could lead directly to cytoplasmic acidification [401]. Na + -H + exchange, an important component of ph i regulation in vascular smooth muscle, depends on activity of the Na + -K + pump, which requires energy for operation [391, 402, 403]. Thus under hypoxia when ATP concentrations are depleted, Na + -K + -ATPase activity could be limited, which would reduce the transmembrane sodium gradient, and decrease acid extrusion via Na + -H + exchange. In HPASMCs, [ATP] i is 148

165 Chapter 5 maintained at oxygen concentrations that are sufficient to deplete ATP levels in HASMCs. Hence in HPASMCs, Na + -H + exchange is maintained under hypoxia and onset of acidosis delayed compared to HASMCs. Alternatively, HPASMCs could genetically possess a greater ability to down regulate ATP utilization in the face of hypoxia or a cytochrome oxidase of greater oxygen affinity [404]. The two cell types might also differ in their glucose uptake. Role of glucose transporters and metabolic sensitivity to various modulators of glycolytic enzymes need to be assessed. Further investigation will determine which of these (or other) explanations is correct. Insight into these mechanisms could lead to the development of pharmacological tools for use in the treatment of diseases associated with hypoxia. 149

166 Supplement Supplement Materials and Methods Antibodies and Reagents Phosphoglycerate kinase (PGK) and Enolase antibodies were purchased from Santa Cruz Biotechnology, Inc (Santa Cruz, CA). Lactate assay kit was obtained from BioVision (Mountain View, CA). All other reagents were from Sigma (St. Louis, MO). Mitochondrial DNA content: Mitochondrial DNA (normalized for nuclear DNA) was quantified as a marker of mitochondria number. HASMCs and HPASMCs were exposed to 48 hours of normoxic or hypoxic (3%, 1%, and 0% O 2 ) incubation. Cells were harvested with 1ml of lysis buffer per plate (1M Tris (ph 8.0), 0.5M EDTA (ph 8.0), 10mg/ml RNase A, 10% SDS). Samples were incubated at 37 C for 2 hours, proteinase K (20mg/ml) was added and the lysate incubated at 50 C overnight. DNA was extracted using phenol/chloroform extraction and ethanol-precipitated overnight. DNA was pelleted instead of spooled to ensure a maximum yield including both nuclear and mitochondrial DNA, and resuspended in 30-80μl TE buffer. Real-time Polymerase Chain Reaction (PCR): Primer pairs specific for three Human mitochondrial genes and three nuclear chromosome genes were chosen for PCR, and cloned into the pcr II plasmid using the TA Cloning Kit Dual Promoter (Invitrogen; Carlsbad, CA). The three mitochondrial genes chosen were: trna leucine 1 (mttrnl1: sense mt3212f 5 - CACCCAAGAACAGGGTTTGT-3 ; antisense mt3319r 5 -TGGCCATGGGTATGTTGTTAA- 3 ); cytochrome c oxidase II (CO2: sense CO2F 5 -CCCCACATTAGGCTTAAAAACAGAT- 3 ; antisense CO2R 5 -TATACCCCCGGTCGTGTAGCGGT-3 ); and non-coding region in tdna controlling DNA replication (Dloop: sense DloopF 5-150

167 Supplement TATCTTTTGGCGGTATGCACTTTTAACAGT-3 ; antisense DloopR 5 - TGATGAGATTAGTAGTATGGGAGTGG-3 ). Primers for the three nuclear genes were chosen to amplify genomic sequences in their promoter regions The three nuclear genes chosen were inducible nitric oxide synthase (inos: sense hinostaqman 5 - TGAAGAGGCACCACACAGAGT-3 ; antisense hinostaqman3 - TGGTTTCCAAAGGGAGTGTCC-5 ), endothelial nitric oxide synthase (enos: sense henostaqman 5 - GTGGAGCTGAGGCTTTAGAGC-3 ; antisense henostaqman 3 - TTTCCTTAGGAAGAGGGAGGG-5 ) and vascular cell adhesion molecule 1 (VCAM1: sense 5 - ACTTGGCTGGGTGTCTGTTA -3 ; antisense VCAM 3 - GCGGAGTGAAATAGAAAGTC -5 ). The cloned plasmid sequences were verified by DNA sequencing (The Centre for Applied Genomics; Toronto, ON), amplified by maxiprep and used as plasmid standards for real-time PCR. DNA concentrations were quantified and diluted to a concentration of approximately 500 copies of nuclear DNA per μl. Real-time PCR was performed on the DNA samples in triplicate. Real-time PCR settings were 95 C for 10mins (Step I); 95 C for 15s, 60 C for 1 min (Step II x 40 cycles); 95 C for 15s, 60 C for 15s, 95 C for 15s (Step III). SYBR Green I dye (Applied Biosystems; Foster City, CA) was used to detect the PCR products, and the specificities of the amplicons were verified by comparing the T m for each amplicon, which were consistent across the different experiments. The number of copies of target sequence was determined by comparison with standard curves generated by both plasmid DNA and genomic DNA serial dilutions. Copy number values from mitochondrial genes were divided by that of nuclear genes to normalize the amount of mtdna to the total number of cells, for each cell type under each condition. The amount of mtdna for each hypoxic condition was divided by the amount of 151

168 Supplement mtdna in the normoxic control cells to determine the average fold change of mtdna levels. Lactate assay: Lactate is oxidized by lactate oxidase to generate a product, which interacts with the lactate probe (provided by manufacturer) to produce fluorescence (at Exitation/Emission = 535/590 nm). Samples were prepared in50 μl/well with Lactate Assay Buffer in a 96-well plate. A standard reference curve was used, corrected for background fluorescence for calculating lactate concentrations, according to the instructions provided by the manufacturer. Western blot analysis: Western blot analysis was used to quantify levels of glycolytic enzymes Phosphoglycerate kinase and Enolase in cytoplasmic extracts from HPASMCs and HASMCs exposed to normoxia or hypoxia (3%, 1% or 0% O 2 ) for 16 or 48 h. Cells were lysed in buffer A [10 mm HEPES (ph 7.8), 10 mm KCl, 0.1 mm EDTA, 1mM dithiothreitol (DTT), and 0.1% Nonidet P-40 (NP-40)] with protease inhibitors (5 g/ml aprotinin, 5 g/ml pepstatin, 5 g/ml leupeptin, 0.5 mm Pefabloc, and 1 mm phenylmethylsulfonyl fluoride) and phosphatase inhibitors (10 mm sodium fluoride, 1 mm sodium orthovanadate, and 20 mm glycerophosphate). Nuclear proteins were then extracted with buffer B [50 mm HEPES (ph 7.8), 420 mm KCl, 0.1 mm EDTA, 1 mm DTT, 5 mm MgCl2, and 20% glycerol], containing both protease and phosphatase inhibitors. Equal amounts of protein extracted from HPASMCs and HASMCs, incubated under normoxic and hypoxic (3%, 1% and 0% O 2 ) conditions for 16 and 48 h, were loaded on 4 12% Tris-glycine gels, separated by electrophoresis and transferred to nitrocellulose. Membranes were blocked with 5% milk overnight and probed with anti- Phosphoglycerate kinase (1:1000) and anti-enolase (1:500). In all cases, protein concentration was determined by the Bradford assay, and appropriate volumes of extraction buffer to produce constant protein loading in each lane were mixed with SDS loading buffer. Equality of protein loading and transfer efficiency were corroborated by full-lane densitometry of the Ponceau red- 152

169 Supplement stained membranes. Immunoblots were probed with horseradish peroxidase-donkey anti-rabbit IgG (1:1,000 in blocking buffer) and visualized by enhanced chemiluminescence (ECL Plus kit, Amersham Biosciences). Band intensity was quantified by densitometry (Bio-Rad Laboratories, Mississauga, ON, Canada). 153

170 Supplement HPASM C HASMC O 2 concentration O 2 concentration Figure S1 Average fold change of mtdna levels in two cell types (HASMC and HPASMC) under different hypoxic conditions (3%, 1%, and 0% O 2 ). Primer pairs specific for three human mitochondrial genes (trna leucine 1, cytochrome oxidase II, Dloop) and three nuclear chromosome genes (inos, enos, VCAM1) were chosen for PCR, and cloned into the pcr II plasmid. Copy number determined using plasmid DNA as standards. Results were averaged from 3 independent experiments. 154

171 A. HPASMC Supplement B. HASMC Figure S2 Cytoplasmic levels of Phosphoglycerate kinase protein in HPASMCs (A) and HASMCs (B) after incubation under normoxic and hypoxic (3, 1 or 0% O 2 ) conditions. n= 6. *P <0.05 vs. corresponding normoxic control values. 155