Adipose Tissue Function

Transcription

1 The Effect of Ageing on Perivascular Adipose Tissue Function A thesis submitted to The University of Manchester for the degree of Doctor of Philosophy (PhD) in the Faculty of Medical and Human Sciences Year of Submission: 2015 Heather M. Melrose School of Medicine 1

2 i) Contents i) Contents... 2 ii) List of Figures... 7 iii) List of Tables... 9 iv) Abstract v) Declaration vi) Copyright Statement vii) Acknowledgements and Dedication viii) The Author ix) Abbreviations Chapter 1: Introduction Cardiovascular Disease with Ageing Influences of Vascular Components on Vascular Tone Vascular Smooth Muscle Vascular Contractility with Ageing The Endothelium Endothelial Nitric Oxide Synthase Superoxide Prostaglandins Endothelium-Derived Hyperpolarising Factor Endothelin Age-Related Endothelial Dysfunction Perivascular Adipose Tissue Perivascular Adipose Tissue Structure PVAT Influence on Contractility Nitric Oxide Prostaglandins

3 Adipokines PVAT Dysfunction O-GlcNacylation Adenosine Monophosphate Kinase AMPK and Ageing Summary Hypothesis Aims Chapter 2: Materials and Methods Animals Phenotypic Measurements Measurement of Blood Pressure Measurement of Blood Glucose and Preparation of Blood/Plasma Samples Plasma Insulin Assay Insulin Resistance Calculation In Vitro Wire Myography Tissue Preparation Mounting vessels onto the wire myograph system Normalisation of vessels Myograph Protocol Exogenous PVAT Protocol Data Collection Pharmacological Agents Western Blotting Tissue Homogenisation Protein Assay

4 2.4.3 Preparation of Stain-Free Acrylamide Gels SDS Polyacrylamide Gel Electrophoresis (SDS-PAGE) Semi-Dry Protein Transfer Process Processing of Membranes Detection of Antibodies Quantification of Data Immunohistochemistry Sample processing Measurement of adipocyte size Statistical Analyses Chapter 3: Results The Effect of PVAT on Isolated Mesenteric Small Artery Contractility The Effect of PVAT on Isolated Mesenteric Small Artery Relaxation The Role of Nitric Oxide in PVAT Anti-Contractile Effect The Effect of U46619 and Phenylephrine on enos Phosphorylation The Role of Prostaglandins in PVAT Anti-Contractile Effect Factors Contributing to the Anti-Contractile Effect of PVAT The Effect of Ageing on the Cardiometabolic Profile of Male Wistar Rats The Effect of Ageing on Adipocyte Size The Effect of Ageing on Potassium-Mediated Constriction The Effect of Ageing on the PVAT Anti-Contractile Effect The Effect of Ageing on the Expression and Phosphorylation of enos in PVAT The Effect of Ageing on Endothelium-Dependent and Independent Relaxation The Contribution of Nitric Oxide to Endothelium-Dependent Relaxation with Ageing The Effect of Ageing on the Expression and Phosphorylation of enos in Mesenteric Arteries

5 3.15 The Effect of Pharmacological AMPK Activation in Young Mesenteric Arteries The Role of Nitric Oxide in AMPK Mediated Reductions in Contractility The Effect of Pharmacological AMPK Activation in Aged Mesenteric Arteries The Role of Nitric Oxide in AMPK-Mediated Effects on Contractility of Isolated Small Mesenteric Arteries of 24 Month Old Wistar Rats The Effect of Ageing on the Expression and Phosphorylation of AMPK in Mesenteric Arteries and PVAT The Effect of Ageing on O-GlcNac Modification of Proteins Chapter 4: Discussion Key Findings Vascular Contractions and Relaxation NO-Mediated Opposition of Contraction PVAT is the Primary Source of Anti-Contractile NO The Effect of COX-Derived Prostaglandins on Vascular Contractility Summary of Findings from Young Animals The Effect of Ageing on Cardiometabolic Parameters The Effect of Ageing on PVAT Anti-Contractile Effect The Effect of Ageing on Vascular Contractility The Effect of Ageing on enos Expression, Activation and Effects on Vascular Contractility The Effect of Ageing on Vascular Prostaglandins The Effect of Ageing and PVAT on Endothelial Function The Effect of Ageing on AMPK Expression, Activation and Effects on Vascular Contractility The Effect of Ageing on O-GlcNacylation Conclusions Limitations and Future Work

6 5 Bibliography Appendix Solution Recipes Myography Solutions Western Blotting Solutions Final Word Count: 38,004 6

7 ii) List of Figures Figure Title Page Figure 1.1 Demographics of cardiovascular disease prevalence with age 20 Figure 1.2 Disorders for which hypertension is an independent risk factor 21 Figure 1.3 Pathways involved in VSMC contraction 25 Figure 1.4 Reaction catalysed by the NOS enzymes 29 Figure 1.5 The protective role of nitric oxide in the vasculature 30 Figure 1.6 Effects of prostaglandins on vascular contractility 34 Figure 1.7 Vasodilator mechanisms of the endothelium 38 Figure 1.8 Vasodilator mechanisms of PVAT 43 Figure 1.9 Mechanisms contributing to age-related hypertension 51 Figure 2.1 Example of the insulin standard curve generated during the 54 insulin ELISA Figure 2.2 Schematic of a wire myograph chamber 56 Figure 2.3 Example of the BSA standard curve generated during the 62 Bradford Assay Figure 2.4 Schematic of a transfer sandwich 64 Figure 2.5 Example stain free blot 64 Figure 3.1 Effect of PVAT on the contractility of isolated mesenteric small 70 arteries Figure 3.2 Effect of PVAT on endothelium-dependent and -independent 71 relaxation in isolated mesenteric small arteries Figure 3.3 Effect of NOS inhibition on isolated mesenteric small artery 73 function Figure 3.4 Effect of U46619 and phenylephrine stimulation on enos 75 phosphorylation Figure 3.5 Effect of COX inhibition on isolated mesenteric small artery 76 function Figure 3.6 Effect of NOS and COX inhibition in exogenous PVAT applied to 79 isolated mesenteric small arteries 7

8 Figure 3.7 Phenotypic changes that occur in male Wistar rats with ageing 81 Figure 3.8 Correlation between ageing and phenotypic characteristics 82 Figure 3.9 Effect of ageing on HOMA-IR 84 Figure 3.10 Effect of ageing on adipocyte size 85 Figure 3.11 Effect of ageing on KPSS-mediated constriction of isolated 86 mesenteric small arteries Figure 3.12 Effect of ageing on the anti-contractile effect of PVAT in 88 response to U46619 in isolated mesenteric small arteries Figure 3.13 Effect of ageing on isolated mesenteric small artery 89 contractility in response to U46619 Figure 3.14 Effect of ageing on the anti-contractile effect of PVAT in 90 response to phenylephrine in isolated mesenteric small arteries Figure 3.15 Effect of ageing on isolated mesenteric small artery 91 contractility in response to phenylephrine Figure 3.16 Effect of NOS inhibition on isolated mesenteric small artery 92 contractility at 24 months Figure 3.17 Expression and phosphorylation of enos in PVAT 94 Figure 3.18 Effects of ageing on endothelium-dependent and - 96 independent relaxation of isolated mesenteric small arteries Figure 3.19 Correlation between ageing and endothelial function of 97 isolated mesenteric small arteries Figure 3.20 Effect of ageing on the contribution of NO to endotheliumdependent 98 relaxation of isolated mesenteric small arteries Figure 3.21 Expression and phosphorylation of enos in mesenteric small 100 arteries Figure 3.22 Effects of A incubation on AMPK Thr phosphorylation Figure 3.23 Effect of AMPK activation on contractility of isolated 103 mesenteric small arteries at 3 months Figure 3.24 Effect of NOS inhibition on the anti-contractile effect of AMPK activation in isolated mesenteric small arteries 105 8

9 Figure 3.25 Effect of AMPK activation on isolated mesenteric small artery 106 contractility at 24 months Figure 3.26 Effect of NOS inhibition on A mediated effects on 107 isolated mesenteric small artery contractility at 24 months Figure 3.27 Expression and phosphorylation of AMPK in mesenteric small 109 arteries Figure 3.28 Expression and phosphorylation of AMPK in PVAT 110 Figure 3.29 O-GlcNac modification of PVAT 111 Figure 4.1 Summary of contractile/anti-contractile mechanisms in small 124 arteries from young animals Figure 4.2 Mechanisms contributing to age-related hypertension 149 iii) List of Tables Table Title Page Table 2.1 Details of Primary Antibodies Used 65 9

10 iv) Abstract The Effect of Ageing on Perivascular Adipose Tissue Function Heather Melrose, The University of Manchester Doctor of Philosophy (PhD) in the Faculty of Medical and Human Sciences 14 th September 2015 Increasing age is the single biggest independent risk factor for cardiovascular disease, which is in turn the leading cause of morbidity and mortality worldwide. Ageing is associated with hypertension and metabolic changes which all increase the risk of the development of cardiovascular disease. In young, healthy individuals, perivascular adipose tissue (PVAT) secretes factors that can influence vascular contractility, exerting a net anti-contractile effect against numerous vasoconstrictors including the thromboxane A2 mimetic U46619 and α1-adrenoceptor phenylephrine. Whilst it is known that dysfunction in PVAT can contribute to obesity-related hypertension, little is known whether similar dysfunction occurs with ageing. In young Wistar rats, wire myography and pharmacological studies showed that the anti-contractile effect of PVAT in the presence of U46619 is dependent on both PVAT-derived nitric oxide and prostaglandins, whereas the anti-contractile effect in the presence of phenylephrine is nitric oxide independent. This finding was supported by Western blot experiments that showed increased phosphorylation of endothelial nitric oxide synthase (enos) in PVAT following U46619 incubation, but not phenylephrine. In the Wistar rat model of ageing used, wire myograph studies revealed that the PVAT anti-contractile effect in the presence of phenylephrine is preserved at 24 months of age, but in in the presence of U46619 is lost. Furthermore PVAT from aged animals had a deleterious effect on endothelial function, suggesting changes in its secreted factors. These changes are accompanied by alterations in the expression and activation of key enzymes in the nitric oxide synthesis pathway within the PVAT as measured by Western blot, as well as alterations in cardiometabolic phenotype including hypertension, hyperglycaemia and insulin resistance. Taken together these findings suggest that previously unidentified age-related PVAT dysfunction may contribute to agerelated hypertension and thus may provide a potential therapeutic target for future study. 10

11 v) Declaration No portion of the work referred to in this thesis has been submitted in support of an application for another degree or qualification from this or any other university or institute of learning, except the phenotypic data presented in Figures 3.7 and 3.8 which was jointly obtained by the author and Dr. Lucy Walton, who included it in her doctoral thesis entitled From molecules to tissues: characterising the relationship between structure and function in ageing arteries, submitted

12 vi) Copyright Statement i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the Copyright ) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. iii. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the Intellectual Property ) and any reproductions of copyright works in the thesis, for example graphs and tables ( Reproductions ), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. 12

13 iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property University IP Policy (see in any relevant Thesis restriction declarations deposited in the University Library, The University Library s regulations (see and in The University s policy on Presentation of Theses 13

14 vii) Acknowledgements and Dedication I'd rather be a could-be if I cannot be an are; because a could-be is a maybe who is reaching for a star. I'd rather be a has-been than a might-have-been, by far; for a might have-been has never been, but a has was once an are. -Milton Berle First and foremost I would like to acknowledge my parents for all of their support (emotional, financial and practical), not just throughout my PhD but my entire life. Your unwavering conviction in my ability to be a could be has gotten me where I am today. The quote above, by the way, way shamelessly stolen from the thesis of Dr. Sarah Withers who has, as needed, worn the hat of friend, mentor, idol, cheerleader, drill sergeant and tea buddy (mostly tea buddy) for half a decade. The most challenging half a decade of my life. It s been a rollercoaster and there s no-one I d rather share the carriage with. I, of course, want to thank my supervisors Dr. Austin, Dr. Edwards and Prof. Heagerty for their support, guidance and, most of all, patience throughout my PhD We got there in the end! To the many faces old and new who have passed through the lab, the tea team and beyond. Sharing a joke, attempting the Metro quiz, debating The Top Ten at Ten kept me sane and more importantly, smiling. And finally to my now husband, Lloyd. Your support at my lowest is what gives me the strength to climb to my highest. To turn the could be to an are. This work is for you. 14

15 viii) The Author The author, Heather Melrose, graduated from The University of Manchester with a first class degree in Biology with Industrial Experience (BSc) in This consisted of three years of lecture-based learning modules and a one year laboratory based project at AstraZeneca Pharmaceuticals. Experimental data presented in this thesis was obtained from September 2011 onwards, following an initial year consisting of three short research projects. All practical laboratory work was performed within the Institute of Cardiovascular Sciences at The University of Manchester. 15

16 ix) Abbreviations AC ACh AMP AMPK AMPKK ANOVA APS ATP BAT BH 4 BK Ca BSA Ca 2+ CaMKKβ camp cgmp CVD Adenylate cyclase Acetylcholine Adenosine monophosphate AMP-activated protein kinase AMPK-kinase Analysis of variance Ammonium persulfate Adenosine triphosphate Brown adipose tissue Tetrahydrobioterin Calcium-sensitive large-conductance potassium channel Bovine serum albumin Calcium Ca/calmodulin-dependent protein kinase kinase Cyclic adenosine 3,5 monophosphate Cyclic guanosine 3,5 monophosphate Cardiovascular disease Cx40 Connexin 40 EDTA EDHF EGTA ELISA enos GC H 2O 2 Ethylenediaminetetraacetic acid Endothelium-derived hyperpolarising factor Ethylene glycol-bis(2-aminoethyleether)-n,n,n,n Enzyme-linked immunosorbent assay Endothelial nitric oxide synthase Guanylate cyclase Hydrogen peroxide Hsp90 Heat shock protein 90 16

17 IK Calcium-sensitive potassium channel 3.1 inos IP 3 K ATP K Ca KCl K 2 EDTA KH 2PO 4 K v L-NMMA L-NNA MLC MLCK MLCP NaHCO 3 NO NOS nnos NS PE PG PGI 2 PIP 2 PKG PSS PVAT SDS-PAGE SNP Inducible nitric oxide synthase Inositol 1,4,5-trisphosphate Adenosine triphosphate-sensitive potassium channel Calcium-activated potassium channel Potassium chloride Ethylenediaminetetraacetic acid dipotassium salt Potassium dihydrogen orthophosphate Voltage-sensitive potassium channel ι-n G -monomethyl arginine citrate N ω -nitro-l-arginine Myosin light chain Myosin light chain kinase Myosin light chain phosphatase Sodium hydrogen carbonate Nitric oxide Nitric oxide synthase Neuronal nitric oxide synthase Not significant Phenylephrine Prostaglandin Prostacyclin Phosphatidylinositol 4,5-bisphosphate Protein kinase G Physiological saline solution Perivascular adipose tissue Sodium dodecyl sulfate polyacrylamide gel electrophoresis Sodium nitroprusside 17

18 SR TEMED TBS TBS-Tween TGF VSM VSMC WAT Sarcoplasmic reticulum Tetramethylethylenediamine Tris-buffered saline Tris buffered saline + 0.1% tween Transforming growth factor Vascular smooth muscle Vascular smooth muscle cell White adipose tissue 18

19 1 Chapter 1: Introduction 1.1 Cardiovascular Disease with Ageing Cardiovascular disease (CVD) including hypertension is one of the leading causes of morbidity and mortality worldwide 1. Whilst obesity, increased lipid levels and metabolic syndromes certainly confer major risks for the development of CVD 2, advancing age is unarguably the most important factor in determining the likelihood of CVD within a population. As Figure 1.1, based on data from the US, shows, increased age is coupled with an increased risk of hypertension, coronary heart disease, myocardial infarction and stroke. A report from the Office for National Statistics 3, UK, in 2012 found that age-standardised mortality rates were at the lowest since records began, an indication of our increasingly aged population worldwide 4. The same report also showed that the incidence of death due to circulatory disease in 2011 was 1,803 per million population, second only in the UK to cancer (the incidence of which was 2,023). The significant morbidity and mortality associated with CVD, the increasing incidence of CVD with age and our increasingly ageing population are likely to result in a significant burden on healthcare systems. Indeed an ageing population, in addition to an increasingly obese population 5, is potentially contributing to the rising incidence of hypertension globally. A study published in the Lancet in noted that in 2000 the estimated total number of adults with hypertension was 972 million, (approximately 26.4% of the population) a figure predicted to rise by 60% to approximately 1.56 billion (29.2%) by The occurrence of hypertension increases dramatically with 19

20 advanced age, with four times the incidence in those aged over 85 compared with those aged (Figure 1.1). Figure 1.1: Demographics of cardiovascular disease prevalence with age. A) Percentage of the US population (A) with high blood pressure (defined as systolic pressure 140mmHg or diastolic pressure 90mmHg), or taking antihypertensive medication; B) indicated that they had confirmed coronary heart disease, C) indicated they had suffered a confirmed myocardial infarction or D) indicated they had suffered a confirmed stroke. From NHANES Hypertension is an independent risk factor for a plethora of disorders, both in the vasculature and other organs (Figure 1.2). In the case of stroke this conferred risk is tangible; a decrease of 10 mmhg in systolic blood pressure is 20

21 associated with a reduction in stroke risk of around 33% in those aged years, with a continuous association down to levels of at least 115/75 mmhg 8. Figure 1.2: Disorders for which hypertension is an independent risk factor. Hypertension has been shown to be directly involved in the pathogenesis of numerous disorders, including: stroke 8, atherosclerosis 9, myocardial infarction 10, bone fractures 11, vascular dementia 12 and heart failure 13. There are many mechanisms by which hypertension can confer risks of CVD. Hypertension is associated with multiple changes throughout the vasculature, for example vascular remodelling (Reviewed by Mulvany, ), which increases peripheral resistance. Acutely, increases in blood pressure occur when the diameter of peripheral resistance arteries narrows (vasoconstriction); regulation of which is determined directly by vascular smooth muscle (VSM) contractility 15,16. Given the direct role of the VSM on regulating vessel constriction, the majority of investigations, in this study and those undertaken by others, are performed using isolated arterial segments. Given the potential healthcare burden, it is crucial that we develop an understanding of the mechanisms contributing to the development of age-related 21

22 hypertension. Key to this are age-related effects on 1) vascular smooth muscle contractility and 2) the factors that can influence this. The different components of the vasculature and their influence on vascular contractility (and thus their potential contribution to the pathophysiology of age-related hypertension) are discussed below. 1.2 Influences of Vascular Components on Vascular Tone The vasculature is subject to dynamic changes that occur with ageing in and in disease, on both structural and functional levels. Although other components of the vasculature such as the perivascular adipose tissue (PVAT) and endothelium can influence vessel contractility, it is ultimately the level of contraction of the vascular smooth muscle that dictates blood vessel diameter 17. In the case of the peripheral resistance vessels, this can have profound effects on systemic blood pressure The three major components of the vasculature, their impact on vessel constriction and their role in the pathogenesis of vascular diseases, such as hypertension, are described below Vascular Smooth Muscle Vascular smooth muscle cells (VSMCs) residing in the tunica media of the small resistance artery wall are responsible for directly levying the level of tone within the vessel by regulating lumen diameter, ultimately determining the pressure needed for blood to pass along the artery. Concerted contraction of VSMCs will cause the lumen of the blood vessel to narrow, thus increasing blood pressure within the artery. Small peripheral arteries are the chief source of peripheral vascular resistance and are key determinants of systemic blood pressure Thus the 22

23 level of contractility of the VSMCs of peripheral resistance vessels in health and disease will directly influence the level of blood pressure within the body. The process of VSMC contraction largely relies on increases in free intracellular calcium concentration ([Ca 2+ ]i). This free Ca 2+ binds the Ca 2+ binding protein calmodulin which in turn activates the myosin light chain kinase (MLCK). Activated MLCK (in the presence of ATP) then phosphorylates the myosin light chains (MLCs), which are regulatory subunits found on myosin heads. Once phosphorylated, MLCs form cross-bridges with actin filaments within the cell, causing VSMC contraction. Relaxation is achieved via dephosphorylation of the MLCs, by the myosin light chain phosphatase (MLCP). The increased [Ca 2+ ]i that leads to this contraction is the downstream effector of numerous signalling pathways and can be achieved either by increased influx of Ca 2+ into the cell or by localised release of Ca 2+ from intracellular stores such as the sarcoplasmic reticulum (SR). Ca 2+ can enter the cell through a variety of Ca 2+ channels, which in turn are activated by various stimuli. Some Ca 2+ channels on the VSMC membrane can be activated by specific agonists 21, whilst many are voltage sensitive, for example the L-Type Ca 2+ channel 22. Consequently, depolarisations of the VSMC membrane activate these channels, causing influx of extracellular Ca 2+ and thus VSMC contraction. In opposition of this, increases in [Ca 2+ ]i and changes in membrane potential also activate K + channels, such as the large conductance Ca 2+ activated K + channel (BKCa) 23, leading to VSMC hyperpolarisation. Increases in [Ca 2+ ]i, whilst unarguably important, are not the sole mechanism for induction of VSMC contraction. Some agonists cause contraction by increasing the Ca 2+ sensitivity of the VSMC contractile machinery, thus 23

24 enabling contractions at lower [Ca 2+ ]i. Ca 2+ sensitisation of the VSMC contractile machinery is achieved through activation of the RhoA/Rho-kinase pathway which inhibits dephosphorylation of the MLC by MLCP, maintaining contraction 24 VSMC contraction as described above can occur as the result of a variety of stimuli, both exogenous and endogenous. G-protein coupled receptors (GPCRs) residing on the VSMC membrane can elicit contraction through their coupling to different G proteins. GPCRs coupling to the Gq subunit elicit contractions through activation of phospholipase C (PLC), which catalyses the breakdown of phosphatidylinositol 4,5-bisphosphate (PIP2) to diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3). IP3 binding to its receptor on the membrane of the SR causes release of Ca 2+ from this intracellular store, increasing [Ca 2+ ]i and causing VSMC contraction as described above 25. Receptors coupling to this Gq mediated pathway include α-adrenoceptors 26, serotonin receptors 27, endothelin-1 receptors 28 (which also regulate VSMC contractility by antagonising the relaxatory effects of adenylate cyclase 29 ) and, to a lesser degree, the thromboxane A2 (TP) receptor 30. Activation of the G12-13 subunit activates RhoA and subsequently Rho-kinase, which increases the Ca 2+ sensitivity of the VSMC contractile machinery 31,32, as mentioned above. Activation of the G12-13 and subsequent Ca 2+ sensitisation of the VSMC contractile machinery is the primary signalling pathway coupled to the TP receptor 30,33. Many of the pathways described above are summarised in Figure

25 Figure 1.3: Pathways involved in VSMC contraction. VSMC contraction relies on the phosphorylation of the myosin light chain (MLC). When phosphorylated, MLC forms cross-bridges with actin filaments, causing contraction. Myosin light chain kinase (MLCK) is responsible for phosphorylating MLC, and is activated by calmodulin which in turn becomes active in the presence of Ca 2+, thus increases in [Ca 2+ ]i ultimately lead to contraction. Many signalling pathways result in increased [Ca 2+ ]i including Ca 2+ entry through channels such as the voltage gated Ca 2+ channel (VGCC) and the non-selective cation channel (NSCC). Alternatively, [Ca 2+ ]i can also be increased via release of Ca 2+ from intracellular Ca 2+ stores such as the sarcoplasmic reticulum (SR). This release can be triggered by agonist binding to G-protein coupled receptors (GPCRs) coupled to Gq proteins which augment formation of inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG) from PIP2 through activation of protein lipase C (PLC). IP3 binds its receptor on the SR causing local Ca 2+ release, whereas DAG causes Ca 2+ influx in addition to activating protein kinase C (PKC) which inhibits dephosphorylation of MLC by myosin light chain phosphatase (MLCP), opposing relaxation. GPCRs coupled to G12/13 also inhibit MLCP through activation of Ras homolog family member A (RhoA) and Rho-associated kinase (ROCK). One of the most important endogenous vasodilators, nitric oxide (NO), augments MLCP activity through increasing the production of the cyclic nucleotides cgmp and camp through activation of the enzymes guanylyl cyclase (GC) and adenylyl cyclase (AC) respectively. cgmp-mediated activation of protein kinase G (PKG) and camp -mediated activation of protein kinase A (PKA) both result in the phosphorylation and thus activation of MLCP which dephosphorylates MLC to effect VSMC relaxation. 25

26 Vascular Contractility with Ageing The effects of ageing on the production and magnitude of vascular contraction can vary depending on the pathway involved. Specific changes can be appreciated via the effects that ageing has on artery responses to different experimental agonists. Adrenoceptor agonists, for example, generally elicit similar contraction in the arteries of both young and old animals 34-36, suggesting a preservation of adrenoceptor-mediated contractility. Although the endogenous supply of adrenoceptor agonists is, however, altered with ageing. Increasing age has previously been shown to result in increased sympathetic nerve activity/influence on the vasculature 37-39, which is mediated via the release of endogenous adrenoceptor agonists. Increased sympathetic nervous system activity such as this has previously been shown to contribute to the pathogenesis of hypertension 40,41 and may therefore contribute to age-related vascular dysfunction. Unlike the generally preserved adrenoceptor-mediated contractility with ageing, TP receptor-elicited contractions induced by the thromboxane A2 mimetic U46619 have been shown to be diversely modulated by ageing, depending on the arteries investigated. For example, rat coronary arterioles exhibit markedly reduced contractility to U ; rat abdominal aortae a more marginal reduction in contractility 43 and rat thoracic aortae a preservation of contractility 44,45. Constrictions to serotonin have regularly been shown to be increased with ageing, in vessels including rat 46,47 and dog 48 coronary arterioles, as well as rat mesenteric arteries 48. However one study in rat basilar arteries did show a decrease in serotonin-mediated constriction

27 Endothelin-1 is a potent endogenous vasoconstrictor. One early study by Weinheimer et al. (1990) 50 provides insight into the effects of ageing on the contractions elicited by endothelin-1 in a variety of rat vascular beds. The EC50 for endothelin-induced contraction was significantly increased in both the aorta and renal arteries of aged rats, although the maximal response remained unchanged. In contrast, the maximal response of mesenteric arteries to endothelin-1 was increased with ageing, despite no change in EC50. Changes in age-related arterial contractility thus appear to be both agonist and tissue specific, with further investigation needed to determine the effects of ageing specifically on the small resistance arteries known to be key determinants of systemic blood pressure. Insight into changes occurring in these vessels may therefore also elucidate mechanisms involved in the pathogenesis of age-related hypertension. Furthermore, studies to date have been performed in animal models, with a need therefore for examination in human small resistance arteries. However, whilst the VSM is the direct mediator of resistance and thus blood pressure, the other tissues within the vascular environment, namely the endothelium lining the vessel lumen and the PVAT surrounding vessels (the focus of this investigation), can regulate the level of arterial contractility, either promoting or opposing contraction. The influence of these tissues on vascular contractility and their potential for dysfunction with ageing are discussed in more detail below The Endothelium The endothelium is a layer of simple squamous endothelial cells that lines the entire circulatory system. The endothelium is highly dynamic, exerting a 27

28 profound effect on vascular tone by secreting factors that modulate VSMC contractility through both autocrine and paracrine pathways. These secretory factors include both vasodilators such as endothelium-derived hyperpolarising factors (EDHF), the prostaglandin PGI2 and nitric oxide (NO) and vasoconstrictors such as endothelin-1, angiotensin II, thromboxane A2 and reactive oxygen species (Reviewed in Félétou & Vanhoutte, ). In addition to secreting soluble factors that diffuse locally to VSMCs to influence contraction, endothelial cells are also electrochemically coupled to adjacent VSMCs through myoendothelial gap junctions 52, which are permeable to ions and small molecules. These close connections are thought to be of crucial importance in the spread of hyperpolarisation from endothelial cells to VSMCs, the functional effect of which is described below in Section Whereas increases in VSMC [Ca 2+ ]I result in vasoconstriction, increases in endothelial [Ca 2 ]i stimulate the production of endothelium-derived factors which, on the whole, oppose VSMC contraction, ultimately resulting in vasodilation. The actions of some endothelium-derived vasoactive factors are described below, and the vasodilatory influence on VSM summarised in Figure Endothelial Nitric Oxide Synthase Nitric oxide (NO), first highlighted as an important endogenous vasodilator by Furchgott and Zawadzki in , is synthesised in endothelial cells by endothelial nitric oxide synthase (enos). enos is one of three members of the nitric oxide synthase (NOS) family, along with inducible NOS (inos) and neuronal (nnos), which all comprise two main domains: a flavin-containing reductase domain at the C-terminus and a haem-containing oxygenase domain at the N-terminus 54. These two domains perform catalytically discrete functions 28

29 that synergistically result in the production of NO through hydroxylation of L- arginine to L-citrulline via a N G -hydroxy-l-arginine intermediate which is itself also a substrate for NOS (Figure 1.4). Numerous co-factors, such as tetrahydrobiopterin (BH4), are required for this process, and disruption of their binding to enos can cause enos to become uncoupled, which is described in more detail in Section H2O H2O Figure 1.4: Reaction catalysed by the NOS enzymes. L-arginine is hydroxylated to N G -hydroxy- L-arginine by the NOS reductase domain, transferring electrons from 1.0mol of NADPH to cytochrome c via the flavins flavin adenine nucleotide (FAD) and flavin mononucleotide (FMN). The oxidase domain of NOS then oxidates N G -hydroxy-l-arginine to L-citrulline and NO, using a single electron from 0.5mol of NADPH. NO synthesised by enos in the endothelium diffuses to the VSM where it activates soluble guanylyl cyclase and thus augments production of the intracellular second messenger cgmp 55 which activates cgmp-dependent protein kinase G (PKG). PKG activation opposes vascular smooth muscle contraction through a variety of mechanisms (reviewed by Lincoln et al., ), including: inhibition of IP3 and its receptor; desensitisation of contractile proteins; inhibition of Ca 2+ -activated K + -channels; activation of Ca 2+ -ATPase activity; and 29

30 activation of the MLCP 57, resulting in dephosphorylation of the MLC 58. The potency of NO as a vasodilator is illustrated by the robust smooth muscle relaxations 59 and reductions in blood pressure that are induced in humans 60 by the nitric oxide donor sodium nitroprusside (SNP). In addition to direct effects on vascular contractility, NO is also protective against numerous mediators of CVD, summarised in Figure 1.5. Figure 1.5: The protective role of NO in the vasculature. NO is protective against several pathological processes within the vasculature including vascular smooth muscle cell proliferation and thrombus formation. Benefits reviewed by Lei et al. (2013) 61. Numerous stimuli can promote the production of NO, including but by no means limited to: pharmacological intervention (e.g. stimulation of β- adrenoceptors by agonists such as noradrenaline 62 ), physical induction (e.g. endothelial shear stress 63 ) and direct activation (e.g. phosphorylation at Ser

31 64 by kinases such as adenosine monophosphate kinase (AMPK) 65 ), discussed in more detail below in Section Superoxide In health, enos predominantly produces NO, as described above. However in certain pathophysiological conditions, enos can become uncoupled, instead producing an increased proportion of free radicals, including superoxide 66. The major mechanism for enos uncoupling is a reduction in the bioavailability of the co-factor BH4, which is reduced by electron transfer from enos flavins following its protonation to BH3.H +. If this protonated BH4 is unavailable, electron transfer from enos flavins becomes uncoupled from N G - hydroxy-l-arginine oxidation, leading to a dissociation of the ferrous-dioxygen complex 67 and production of superoxide by the enzyme s haem domain 68. Superoxide is known to induce vasoconstriction following its reaction with NO, directly through formation of peroxynitrite and hydrogen peroxide, which can cause vasoconstriction through downstream production of vasoconstrictor prostaglandins and indirectly by reducing the bioavailability of NO Prostaglandins Prostaglandins (PGs) were among the first endogenously produced vasoactive substances identified 73 and can have diverse actions within the vasculature (Figure 1.6). PGs are synthesised via the metabolism of cellmembrane-derived arachidonic acid by enzymes with cyclooxygenase (COX) activity. There are two main isoforms of COX enzyme; COX-1 which was first 31

32 isolated independently by three groups 74-76, and COX-2, which was identified slightly later 77. Both COX isoforms share a high level of homology (65%), however their expression, activation and regulation can differ vastly, even within the same cell type 78. Their differential substrate preferences (COX-1 primarily metabolises arachidonic acid whereas COX-2 metabolises both arachidonic acid and 2- arachidonyl glycerol equally) means that COX-2 is capable of producing a subset of metabolites that COX-1 generally does not 79. These PGs interact with PG receptors (IP, EP, DP, FP and TP), which are present in differing proportions on VSMCs, endothelial cells and PVAT adipocytes, to produce downstream effects such as vasoconstriction or dilation. PGI2, a potent endogenous vasodilator and one of the most abundant PGs produced by the vasculature, is synthesised in the endothelium 80 (and also PVAT 81,82 ). PGI2 binding to its IP receptor on VSMCs activates the Gs subunit, resulting in activation of adenylate cyclase, increases in camp and ultimately vasorelaxation 83. Endothelial-derived PGF2α binding to VSMC FP receptors causes an increase in [Ca 2+ ]i that causes vasoconstriction; however FP receptors have also been shown to be present on vascular endothelial cells, where their activation can induce vasodilation through augmented production of NO 84. PGE2 has divergent effects on vascular contractility depending on the receptor subtype, of which there are 4: EP1, EP2, EP3, and EP4 83. VSMC EP1 activation results in vasoconstriction through increased [Ca 2+ ]i following PLC activation and also by increased Ca 2+ sensitivity through Rho activation. EP3 activation results in VSMC contraction through decreases in intracellular camp

33 VSMC EP2 or EP4 activation, however, both result in vasorelaxation by activation of adenylate cyclase (AC) and subsequent increases in camp 83. EP4 receptor activation can also increase NO bioavailability though enos activation 86, which could occur through its ability to activate adenosine monophosphate kinase (AMPK) 87, a key upstream kinase known to phosphorylate enos. Although PGD2 is synthesised in the endothelium of the vasculature, where its production is upregulated in response to shear stress 88, it appears to be involved more in the regulation of allergic responses and inflammation 89 than in vascular contractility. However it can elicit both NO- and endothelium-dependent vasorelaxation 90, as well as VSMC contractions, the latter via activation of TP receptors 91. The PG thromboxane A2 is an endogenous vasoconstrictor that acts through binding TP receptors on VSMCs; TP receptors are also present on perivascular adipocytes 92 and endothelial cells 93. VSMC TP receptors couple to both Gq and G12-13 subunits, which induce vasoconstriction through both increases in [Ca 2+ ]i and, predominantly, Ca 2+ sensitivity of the contractile machinery. Other PGs can also bind and activate TP receptors, including PGD2 and PGE2, and TP receptor activation can also result in increases in PGE2 production 94, which can then go on to exert the downstream effects of either vasoconstriction or vasodilation, as described above. PGF2α can also activate TP receptors. In addition to producing vasoactive PGs, COX activation also produces superoxide anions, which can reduce NO bioavailability through COX-mediated co-oxidation of substances such as NADPH 95. Thus COX activity in the vasculature can have diverse effects on vascular contractility. 33

34 Figure 1.6: Effects of prostaglandins on vascular contractility. Prostaglandins produced within the perivascular environment can cause vasoconstriction or vasodilation depending on the downstream signalling pathways triggered by their receptors. PGI2 binding to the IP receptor will result in vasodilation and thromboxane A2 binding to the TP receptor will result in vasoconstriction whereas PGF2α, PGE2 and PGD2 can cause either vasodilation or vasoconstriction depending on the receptor activated Endothelium-Derived Hyperpolarising Factor In addition to the release of NO and PGs, the endothelium can also oppose VSMC contraction via endothelium-derived hyperpolarising factors (EDHFs). The ability of muscarinic stimulation of the endothelium to evoke a hyperpolarisation (and thus relaxation) in VSMCs was first identified by Kuriyama et al. in EDHF-mediated dilations occur due to a hyperpolarisation of endothelial cells that subsequently spreads to vascular smooth muscle cells. There are numerous candidates for EDHF, including several diffusible factors, such as arachidonic acid-derived epoxyeicosatrienoic acids (EETs), hydrogen peroxide (H2O2), and K + ions, as well as spread of the hyperpolarisation from endothelial cells through 34

35 myoendothelial gap junctions 97. Numerous pharmacological experiments have provided evidence for numerous characteristics of EHDF-mediated relaxation, including it being: Sensitive to Removal of the endothelium 98 Non-Selective calcium-activated potassium channel (KCa) blockers 98 Blockade of endothelial cell connexion 40 (Cx40) 99 Depolarisation with potassium chloride (KCl) Gap-junction inhibition 100 Insensitive to ATP-sensitive potassium channel (KATP)-specific blockers 98 Calcium-activated large-conductance potassium channel (BKCa)-specific blockers 101 Voltage-activated potassium channel (Kv)-specific blockers 102 NOS-inhibitors 103 COX-inhibitors 103 It is also known that endothelium-dependent hyperpolarisations of VSMCs can be induced by activators of the endothelial calcium-activated smallconductance (SKCa) or intermediate-conductance (IKCa) channels (but not of the myocyte BKCa) 104 and increases in endothelial cell [Ca]i 105. Furthermore, KCa channels necessary for EDHF-mediated relaxations of VSMCs were shown to reside on endothelial cells and not VSMCs 106. Thus taken together, it appears that the EDHF is likely to be an efflux of K + through endothelial KCa channels 106, in addition to the direct spreading of hyperpolarisation from endothelial cells to VSMCs via myoendothelial gap junctions 99. The importance of myoendothelial gap junctions to EDHF is highlighted by the relationship between their abundance 35

36 and the contribution of EDHF to endothelium-dependent relaxation, which increases in parallel as vessel diameter decreases Endothelin-1 Although in health the endothelium exerts a net anti-contractile effect on VSMCs, it does secrete a range of pro-contractile factors, including the potent vasoconstrictor endothelin Endothelin-1 binding to its ETA receptor on VSMCs results in a long-lasting vasoconstriction that occurs via increases in [Ca 2+ ]i downstream of numerous G protein-dependent and -independent pathways, including PLC-mediated production of IP3 109, RhoA activation 110 and the mitogen-activated protein kinase (MAPK) pathway 111. Endothelin-1 also activates pathways that counteract its effects on constriction, providing a feedback mechanism. This is achieved through binding of endothelin-1 to ETB receptors present on the endothelial cells themselves, resulting in increases in [Ca 2+ ]i, here leading to the synthesis and secretion of NO 112 and PGI Age-Related Endothelial Dysfunction As previously discussed, one of the most important mediators of the endothelium s influence on the level of constriction imposed by the VSM, is the endogenous vasodilator NO. Reduced bioavailability of NO is a hallmark of both ageing and CVD and can occur through a variety of mechanisms, many involving oxidative stress. This compromise of the NO-mediated relaxation pathway is termed endothelial dysfunction and correlates positively with cardiovascular risk 114, at least in large arteries. There is a correlation between several hallmarks of endothelial dysfunction and ageing, including reduced NO bioavailability 115, reduced shear stress 116 and 36

37 increased release of pro-atherogenic factors from the endothelium such as endothelin There is also a wealth of evidence for the relationship between ageing and enos uncoupling, as described above, which in turn can contribute to both oxidative stress and reductions in NO bioavailability 117. In rat models of ageing, numerous rat vascular beds have shown evidence for enos uncoupling, including mesenteric arteries 118, the aorta 119, coronary arterioles 120, skeletal muscle resistance arterioles 121 and the carotid artery 119. Furthermore, isolated human endothelial cell lines have been shown to exhibit reduced enos function with ageing 122,123. In addition to reductions in the bioavailability of NO, the endothelium has also been shown to have augmented secretion of contractile factors in ageing, such as endothelin-1 124, reactive oxygen species 125 and, perhaps most importantly, COX-derived prostaglandins, thought to be predominantly COX-1 derived 126. Thus the efficacy of the endothelium as an anti-contractile influence on vascular function is compromised with ageing. However, despite this apparent paradigm of age-related endothelial dysfunction, several studies demonstrate maintained endothelium-dependent vasodilation in PVAT-denuded arteries with ageing. For example ageing has no effect on ACh-induced relaxation of coronary arteries from 24 m.o. Fisher 344 rats 127, carotid arteries from lean Zucker rats 128 or mesenteric arteries from an accelerated senescence mouse model of ageing 129. This is indicative of the lack of coherence, and the need for further clarification, in our knowledge of the effect of ageing on the vasculature 37

38 Figure 1.7: Vasodilator mechanisms of the endothelium. Endothelial cells secrete numerous factors that oppose vascular smooth muscle cell (VSMC) contraction including nitric oxide (NO) and PGI2 which diffuse to VSMCs to cause increases in intracellular cgmp and camp respectively that result in vasorelaxation via PKG and PKA respectively. The endothelium also produces PGE2 which augments NO production by activation of endothelial nitric oxide synthase (enos). 38

39 1.2.3 Perivascular Adipose Tissue The tissue surrounding the arteries, termed perivascular adipose tissue (PVAT) can also influence the level of contraction imposed by the VSM. By secreting adipokines and other factors that act both directly on VSMCs and indirectly via the endothelium to promote or oppose contraction, PVAT can regulate vascular contractility and thus potentially affect systemic blood pressure Perivascular Adipose Tissue Structure Once thought to be an inert reservoir for lipid storage and structural support, the perivascular adipose tissue surrounding arteries was shown to be a dynamic endocrine organ capable of influencing vascular tone by Soltis & Cassis in PVAT is a heterogeneous tissue comprised of mature adipocytes containing lipid droplets; immune cells such as macrophages and T-lymphocytes; fibroblasts, microvessels and nerves 136. PVAT is present in lean subjects but increases in mass with increasing adiposity Adipose tissue is traditionally divided into two distinct types: white adipose tissue (WAT) and brown adipose tissue (BAT). WAT functions essentially as a fat depot with white adipocytes containing large lipid droplets consisting of triglycerides, whereas BAT is responsible for the generation of body heat in response to cold temperatures, or non-shivering thermogenesis 140. PVAT has features of either class depending on its anatomical location; abdominal PVAT (such as that surrounding mesenteric resistance arteries) predominantly shows the characteristics of WAT and thoracic PVAT that of BAT 141. In some cases, PVAT appears to have the characteristics of WAT while expressing genes predominantly expressed in BAT

40 PVAT Influence on Contractility There are numerous secretory factors produced by PVAT that can directly influence vascular contractility, either inducing or opposing contraction 143. It has been known for over two decades 135 that in young, healthy individuals these factors result in a beneficial net anti-contractile effect on resistance arteries. This anti-contractile effect has been shown to occur in response to numerous vasoconstrictors and in numerous vascular beds. In response to the thromboxane A2 mimetic U46619, both the mouse 144 and rat 145 aorta exhibit significantly reduced contractility in the presence of their surrounding PVAT. Rat aortic PVAT has also been shown to exert anti-contractile effects in response to both serotonin and the α1-adrenceptor agonist phenylephrine that were highest in the presence of the endothelium, but still significant in its absence. 130 In addition to the aorta, the presence of PVAT has also been shown to attenuate phenylephrine-induced responses in the mesenteric arteries of both male 146 and female 147 rats. Constrictions elicited with a less specific adrenoceptor agonist, noradrenaline, are also significantly reduced by the presence of PVAT; an effect that has been demonstrated in resistance arteries from mice 148,76, rats 149,150 and, crucially, humans 151. Some of the key vasoactive substances secreted by PVAT are described below, and summarised in Figure Nitric Oxide Adipocytes have previously been shown to express enos 152,153. This has not yet been confirmed specifically in perivascular adipocytes, but it is known that NO is an essential component of the PVAT anti-contractile effect. Blockade of NO production through NOS inhibition abolishes the PVAT anti-contractile effect 40

41 in numerous arterial preparations, including the rat aorta 130,154, mouse mesenteric arteries 150 and, importantly, human resistance arteries obtained from gluteal fat biopsies 151. This effect of NOS inhibition on the PVAT anti-contractile effect provides evidence that NO is a major component of the influence of PVAT on arterial contractility. Crucially, whether the NO involved in the PVAT anticontractile effect is principally derived directly from the PVAT itself, or is produced by the endothelium in response to PVAT-derived factors is yet to be determined. Whilst the production of NO from the PVAT may be contributing to the PVAT anti-contractile effect, expression of enos also has the potential for enos uncoupling and the production of deleterious superoxide, as described above in the endothelium. Indeed superoxide secreted by PVAT has been shown to cause vasoconstriction of rat mesenteric arteries 155. Furthermore, chronic PVAT release of superoxide can also affect vascular structure by promoting arterial stiffness through hypertrophy and increased collagen deposition Prostaglandins Like the endothelium, PVAT also produces a range of PGs 81,82, that act, as described earlier, to influence vascular tone. Both COX1 and COX2 isoforms are expressed in adipocytes 157,158 and stimulate the production of numerous vasoactive PGs, including PGI2, PGE2, PGD2, PGF2α, and thromboxane A The range of PGs produced can vary depending on vessel type 82, and it has been shown that PVAT-derived PGs may contribute to the PVAT anti-contractile effect in saphenous vein, but play little role in PVAT effects in the internal mammary artery 82. Little is known about the contribution, if any, that PVAT-derived PGs have to the PVAT anti-contractile effect of small resistance arteries, such as those in the mesenteric bed. 41

42 Adipokines Adipokines is a collective term for numerous factors secreted by adipocytes, including leptin 160, adiponectin 161, components of the reninangiotensin system 162 and cytokines such as tumour necrosis factor α (TNFα) 163 and IL Through the secretion of these adipokines, adipose tissue can have various paracrine and autocrine effects throughout the body including within the vasculature. One of the most important vasoactive adipokines is adiponectin, which diffuses across the local vascular environment to activate its receptor (AdipoR1) on VSMCs, endothelial cells and perivascular adipocytes themselves. Activation of PVAT or endothelial AdipoR1 receptors ultimately results in phosphorylation and thus activation of enos 165 (through activation of its upstream kinase AMPK, described in more detail below in Section 1.3.3), causing increased production of vasodilatory NO. Vice versa, enos activity is also thought to augment adiponectin production from adipocytes 166. VSMC AdipoR1 activation also stimulates vasorelaxation, thought to occur via the large-conductance BKCa channel. Activation of BKCa channels results in K + -efflux, membrane hyperpolarisation, subsequent closure of voltage-gated Ca 2+ -channels and ultimately vascular relaxation. Experimentally, direct application of adiponectin to a vessel will result in a relaxation that is abolished in the presence of the BKCa channel blocker iberiotoxin, and is absent in BKCachannel knockout mice 150. The importance of adiponectin signalling in vascular reactivity is highlighted by dysfunction in endothelium-dependent vasodilation seen in hypoadiponectinaemia 167, which is known to occur in obesity 168, and the restoration of vascular function in obesity by exogenous globular adiponectin application

43 Figure 1.8: Vasodilator mechanisms of PVAT. Perivascular adipocytes secrete numerous factors that oppose vascular smooth muscle cell (VSMC) contraction including nitric oxide (NO) and PGI 2 which diffuse to VSMCs to cause increases in intracellular cgmp and camp respectively that result in vasorelaxation via PKG and PKA respectively. PVAT also produces PGE 2 which augments NO production by activation of endothelial nitric oxide synthase (enos), and adiponectin which binds to its AdipoR1 receptor, activating adenosine monophosphate kinase (AMPK) which also activates enos to increase NO production, as well as directly causing vasorelaxation, potentially through interaction with K + channels of VSMCs. 43

44 PVAT Dysfunction PVAT dysfunction is known to occur in a variety of pathologies, including obesity and hypertension. In obese humans, the beneficial anti-contractile effect of PVAT on small artery constriction is lost 151, despite there being an increase in PVAT mass. This deleterious change in function is most likely due to changes in vascular adipocyte morphology and their surrounding environment that lead to hypoxia and altered adipokine secretion that consequently increase vascular tone 151 and lead to the production of other hypoxic challenges creating a viscous cycle of hypoxia and inflammation. Alterations in PVAT function have also been implicated in other diseases, such as hypertension, type 2 diabetes and atherosclerosis. In models of hypertension, PVAT has been shown to undergo changes in structure and function, including adipocyte expansion 170, reduction in the endotheliumdependent and independent relaxation of PVAT-intact vessels 170 and reductions in the PVAT anti-contractile effect PVAT has also been strongly linked with insulin resistance 173,174 leading to the suggestion that PVAT dysfunction may be both a consequence and a contributing factor to the pathogenesis of type 2 diabetes 175,176. In atherosclerosis, PVAT inflammation has been shown to induce the production of transforming growth factor beta (TGF-β), a known mediator of atherosclerosis and recruitment of inflammatory cells such as macrophages to the site of inflammation 136. Little is known about the effects of ageing on PVAT specifically, principally only that in a mouse model of ageing, age-related increases in superoxide signalling within PVAT surrounding the aorta promote arterial stiffness as a result 44

45 of VSMC hypertrophy and increased collagen deposition 156. Nothing, however, is known about the influence of ageing on PVAT effects on vascular contractility. More generally, increasing age has been associated with alterations in the size, distribution and function of adipose tissues 177, as well as changes in adipocyte morphology, secretion profile and gene expression 178,179. In the setting of increasing age, adipokine secretion profiles can be profoundly altered 180. For example, activity of the inflammatory cytokine TNFα has been shown to be upregulated in both the epididymal and retroperitoneal adipose tissue of aged Wistar and Fisher 344 rats 181. Indeed other inflammatory cytokines such as IL-6 and IL-1β have also all been shown to be upregulated in the adipose tissue of aged mice 182. It is unclear of the effects of ageing on one important basally secreted adipokine, adiponectin, which has previously been shown to have both reduced 183 and increased 184 plasma concentrations with ageing. However, ultimately, changes in adipokine secretion with ageing are coupled with increased levels of inflammation 185 and oxidative stress 186, in a similar 187 but genetically distinct 178 mechanism to the changes to adipose tissue seen in obesity. As a result of these changes, the paracrine and autocrine actions of adipose tissue can therefore be dramatically altered with ageing 177, O-GlcNacylation With changes in adipose tissue function often come changes in metabolic parameters, such as glucose metabolism, which are known to be affected by age 189. Alterations in metabolic status can increase the risk of CVD and, furthermore, can have effects at the molecular level, causing protein posttranslation modifications, such as O-GlcNacylation. This particular modification 45

46 involves the O-linked attachment of β-n-acetylglucosamine to serine and threonine residues of nuclear and cytoplasmic proteins 190. Unlike other forms of glycosylation, O-GlcNacylation occurs through an enzyme-dependent mechanism, mediated by the activity of O-GlcNac transferase. The level of O- GlcNacylation in a tissue increases as extracellular glucose or insulin concentrations increase 191 and is involved in many of the pathophysiological effects of diabetes and insulin resistance 192,193, as well as increasing vascular contractility as a result of increased MLCK and decreased MLCP activity 194. Both enos and its upstream kinase AMPK, present in the PVAT, are highly susceptible to hyperglycaemia-induced O-GlcNacylation, which is known to interfere with their ability to be phosphorylated and thus activated The level of O-GlcNacylation has been shown to be increased in several tissues with ageing, including heart, aorta, brain and skeletal muscle 198. Thus, ageing may lead to increased O-GlcNacylation of these key enzymes, reducing their capacity to oppose vascular contraction and potentially impairing PVAT function. The dysfunction occurring within adipose depots with ageing and in disease suggests the potential for PVAT dysfunction with age, especially since PVAT-specific dysfunction has been identified in the setting of numerous diseases. Furthermore, within the vascular environment there is considerable overlap in the secretion profiles of PVAT and the endothelium, with many of the shared components being the target of age-related deteriorations in endothelial function. This overlap in substrates for dysfunction, coupled with the knowledge that dysfunction occurs in adipose tissue at large with ageing together provide the basis for the focus of this investigation, the effects of ageing on PVAT modulation of vascular tone. 46

47 1.2.5 Adenosine Monophosphate Kinase Common to the endothelium and the PVAT is the importance of the production of NO by enos. One key kinase known to phosphorylate enos at its Ser 1177 activation site is AMPK 65. More recently 199, AMPK has also been shown to phosphorylate enos at another activation site - Ser 633 and in the absence of Ca 2+ /calmodulin 65, the inhibitory site Thr 495, although this latter activity was shown in isolated enzyme preparations 200. Another suggested mechanism for AMPKmediated augmentation of enos activity is AMPK enhancement of Hsp binding to enos, which is needed for maximal enos activity 202. AMPK is itself activated via phosphorylation of Thr 172 by the upstream kinases liver kinase beta 1 (LKB1) 203 and Ca 2+ /calmodulin-dependent protein kinase kinase β (CaMKKβ) 204 (suggesting that AMPK may be activated in response to changes in [Ca 2+ ]i 205 ). AMPK activity can be inhibited by phosphorylation of Ser 485/491, which has been shown to prevent LKB1-mediated activation 206. In addition to stimulation of NO production by enos, AMPK activation can also directly influence VSMC contractility through direct interaction with the contractile machinery. AMPK activation has been shown to inhibit MLC phosphorylation 207,208. These direct and indirect effects on VSMC contractility can be observed experimentally through exogenous activation of AMPK. This can be achieved by vessel incubation with AMPK activators, such as 5-Aminoimidazole- 4-carboxamide ribonucleotide (AICAR) 209 or the thienopyridine A Incubation with either of these activators has widely been shown to reduce the contractility of vessels, including the mouse 211,212 and rat aorta 213, rat mesenteric 47

48 artery 214, rat cremaster muscle arteries 215 and rat uterine artery 216. Furthermore, acute activation of AMPK has been shown to lower the blood pressure in mice 212. Adipocytes have been shown to express AMPK 217. However, whilst it has been suggested that perivascular adipocytes specifically express AMPK 218, to our knowledge this has not yet been confirmed. Despite this lack of confirmation, pharmacological activation of AMPK has been shown to have profound effects on PVAT, including increased secretion of the beneficial PVAT-derived adipokine adiponectin, which in turn activates AMPK through binding to the AdipoR1 219 receptor on the endothelium or adjacent PVAT adipocytes. Furthermore, adiponectin-induced relaxation is also inhibited by the AMPK inhibitor dorsomorphin and mimicked by exogenous AMPK activators, suggesting a role for the enzyme in adiponectin s effects on BKCa currents, and ultimately vascular tone, most likely due to an indirect elevation in local calcium (A. H. Weston, personal communication) AMPK and Ageing AMPK activation has numerous beneficial effects within the vasculature which may both contribute to vascular dysfunction with ageing if downregulated, and provide a potential mechanism for reversing dysfunction if stimulated (potentially therapeutically). As discussed, AMPK can augment NO production. This can occur both directly through enos phosphorylation, and indirectly via its protective effects against oxidative stress. These effects are achieved via a widerange of mechanisms 220, including inhibition ROS production by NADPHoxidases 221, thus eliminating ROS generation at its source. AMPK activation therefore has multiple beneficial effects on NO bioavailability, which in turn bolster advantageous vasodilatory responses. 48

49 It is also well known that AMPK also has beneficial effects on vascular smooth muscle cell proliferation 222,223,224, which is increased with ageing 225,226 and is responsible for age-related vessel wall thickening 227,228. This thickening is linked to vascular stiffening 229 and can cause narrowing of the lumen, leading to hypertension. AMPK is a known to be involved in mediating the ageing process within a range of tissues, and its activation has been shown to actually increase longevity 230,231. However, less is known regarding the influence of ageing on AMPK activity. Several studies have highlighted what seems to be a general trend: a reduction in AMPK activity with ageing. For example, Reznick et al. (2007) 232 demonstrated that administration of an exogenous AMPK activator or physical exercise both clearly increased AMPK activity in the skeletal muscle of young rats, but failed to elicit this response in old rats. Furthermore, Qiang et al. (2007) 233 showed an age-related impairment of AMPK activation and subsequent suppression of insulin-stimulated glucose uptake in old rats compared with their younger counterparts. Both of these studies, however, were performed on skeletal muscle and hitherto little is known about the effects of ageing on vascular AMPK. If AMPK proves to be downregulated in a similar fashion within the vasculature with ageing, it may therefore provide a potential therapeutic target in the restoration of healthy vascular function and thus, potentially regulation of blood pressure. Whilst global upregulation of AMPK may seem beneficial, given the effects of downregulation with ageing, further investigation is needed to determine the pleiotropic effects of AMPK activation in such a setting. 49

50 1.3 Summary Peripheral resistance vessels are important mediators of vascular resistance and thus arterial and systemic blood pressure. The level of contraction of resistance arteries in turn is regulated by the PVAT surrounding the vessel and the endothelium lining the lumen of the vessel. Changes in the structure and function of the different components of the vasculature can therefore impact arterial tone and thus systemic blood pressure. Ageing is linked to a multitude of dynamic changes within the vasculature that can contribute to CVD, including hypertension. Although much is known about the effects of ageing on the VSM and endothelium, little is currently known about the effects of ageing specifically on PVAT. A brief summary of some of the mechanisms described above known to contribute to age-related hypertension is shown in Figure Hypothesis Given the potential for changes in the balance between vasodilator and vasoconstrictors mediating PVAT s influence on vascular contractility discussed above, the hypothesis of this project was that: the known alterations in vascular function that occur with ageing are, at least in part, associated with alterations in the anti-contractile effect of PVAT. 1.5 Aims To investigate this hypothesis, the specific aims of the project were: To characterise the involvement of NO, AMPK and prostaglandins in the PVAT anti-contractile effect 50

51 To identify the changes in metabolic profile that occur with ageing To identify the functional changes that occur in the vasculature with ageing: o The effects of ageing on vascular contractility o The effects of ageing on PVAT-mediated effects on vascular function To investigate how NOS and AMPK expression/activation is altered with ageing Figure 1.9: Mechanisms contributing to age-related hypertension. Ageing is known to cause changes to vascular smooth muscle cells (VSMCs) and the endothelium which in turn contribute to changes in vascular function, such as vascular stiffening, inflammation and oxidative stress, all of which can contribute to the pathogenesis of hypertension. Little is currently known about the effects of ageing on perivascular adipose tissue (PVAT) function, and how this may contribute to the development of age-related hypertension. 51

52 2 Chapter 2: Materials and Methods 2.1 Animals All animals used were male Wistar rats, housed under a 12 hour light-dark cycle with food and water available ad libitum. Animals used were aged approximately 3 months old (m.o.), 12 m.o., 18 m.o. or 24 m.o. and were killed using CO2, followed by cervical dislocation (in accordance with Schedule 1 of the Animal (Scientific Procedures) act of 1986). Due to the availability of animals, experiments were performed at different periods of time. Prior to sacrifice rats were fasted overnight ( 16 hours), weighed and blood glucose measurements taken (method described below). Following confirmation of death, the entire mesenteric bed was then removed and immediately placed in ice-cold physiological saline solution (PSS) (See Section 6.1 for recipe). 2.2 Phenotypic Measurements All animals had a number of phenotypic measurements taken, including body weight, blood pressure, blood glucose and plasma insulin to allow assessment of their metabolic state Measurement of Blood Pressure Unrestrained and un-anaesthetised rats had their blood pressure measured using a tail cuff (15mm diameter) attached to a non-invasive blood pressure measuring machine (Panlab LE5002 NIBP, Harvard Apparatus, UK) 234. Before measurements were taken all rats were familiarised with the tail cuff procedure; each animal underwent a cycle of 20 inflation and deflations on at least 2 separate occasions, performed at the same time of day. When obtaining 52

53 final values, once 20 systolic blood pressure readings were generated, readings were placed in numerical order and the middle 12 readings averaged. The final values used were then the averages of two separate occasions of blood pressure measurement (i.e. two separate days). The reproducibility (coefficient of variance) of blood pressure measurements of the age-groups ranged from and was calculated using all blood pressure values obtained from 3m.o. rats, using the following equation: Standard Deviation Mean Measurement of Blood Glucose and Preparation of Blood/Plasma Samples Samples of blood (approximately 2ml) were taken during cardiac puncture using a BD Microlance 3 25g needle 235 and tested for blood glucose levels using a Contour blood glucose monitor and Contour blood glucose test strips (Bayer, UK). The remaining blood sample was then transferred to a falcon tube coated with heparin before being spun at 10,000g at 4 C for 10 minutes. The plasma supernatant was then transferred to Eppendorf tubes, snap frozen in liquid nitrogen and stored at -80 C for future use in the plasma insulin assay Plasma Insulin Assay Plasma insulin from frozen plasma samples was measured using a Rat Insulin ELISA assay kit (ALPCO; Product number 80-INSRT-E01, E10). This assay is sandwich type a ligand binding assay and thus responses measured have a sigmoidal relationship to insulin concentration. Small samples (10µl) of 53

54 plasma, negative controls or standards were mixed with 75µl of the conjugate provided and added to a 96-well plate coated with insulin-specific monoclonal antibodies. After incubation for 2hrs to allow the antibodies to bind to the insulin protein present in the samples, wells were washed with the wash buffer provided and blotted dry. A soluble colorimetric substrate for horseradish peroxidase (HRP) enzyme was then added that produces colour where antibody is bound to insulin protein. The intensity of the colour produced is directly proportional to the concentration of insulin present within the samples. After 15 min incubation with the substrate the Stop Solution was added and the optical density at 450nm measured using a spectrophotometer (BioTek EL800, Northstar Scientific, UK). Sample insulin concentrations were then estimated from the line fitted to the points of the standard curve (Figure 2.1). The sensitivity of the assay was 0.124ng/ml. Figure 2.1: Example of the insulin standard curve generated during the insulin ELISA. Sample insulin concentrations are calculated using the line fitted to the points of the standard curve 54

55 2.2.4 Insulin Resistance Calculation Homeostatic model assessment of insulin resistance (HOMA-IR) was used as an indication of insulin resistance and calculated using the following formula 236 : Glucose (mmol/l) Insulin (U/l) In Vitro Wire Myography Wire myography was used to assess the contractile function of isolated mesenteric small arteries in vitro Tissue Preparation Following overnight fasting, animals were killed using CO2 followed by cervical dislocation (in accordance with Schedule 1 of the Animal (Scientific Procedures) Act of 1986). The entire mesenteric bed was removed and immediately placed in ice cold physiological saline solution (PSS) solution (composition given in Appendix). The entire mesenteric bed was pinned out onto a Sylgard base within a petri dish and viewed using a (Technival 2, Carl Zeiss Jenna, Germany). Third order arteries (diameter approximately 200µm, length approximately 2-3mm) were dissected either with (+PVAT) or without (-PVAT) 55

56 perivascular fat intact. Vessels were kept in PSS on ice (approximately 4 C) until required for experimental use Mounting vessels onto the wire myograph system Vessel segments were mounted onto two 40µm wires within a wire myograph bath chamber (Model 610 Version 2.2, Danish Myo Technology, Denmark) containing 6ml PSS as shown in (Figure 2.2) either with (Figure 2.2, A) or without (Figure 2.2, B) PVAT intact. Vessels were then measured using a microscope eyepiece graticule. Chambers were then placed onto the myograph where they were kept at 37 C and supplied with 5% CO2/air throughout the duration of the experiment. Vessels were initially left for 30 min under no tension to equilibrate before normalisation. A B Figure 2.2: Schematic of a wire myograph chamber. The arterial segment either with (A) or without (B) PVAT is mounted onto two 40µm wires, which are in turn attached to opposing jaws and secured under screws. One jaw was attached to a micrometer and the other to a mechanical transducer. Any contraction undergone by the artery thus creates tension on the wires and is transmitted to the computer via a force transducer and experiments recorded using Powerlab software (AD Instruments, Australia) 56

57 2.3.3 Normalisation of vessels After a 30 min equilibration period vessels were then normalised to 90% of target maximal transmural tension, defined as 100mmHg (13.3kPa) for rat mesenteric arteries 237. Normalising to a set proportional tension allows for comparison between vessels of different lengths and diameters and expression of results in mn/mm. Normalisation involves distending the vessel by increasing the distance between the jaws (4µm increments) in a stepwise manner by moving the micrometer to elicit increases in force reading. The Danish Myo Technology Normalization Module (Danish Myo Technology, Denmark) then calculates the tension-force relationship for the vessel using the following equation: T n = (Y n Y 0 ) 2δ (a 1 a 2 ) (Tn is tension at the nth micrometer position corresponding to the micrometer reading at said position (Yn). Yo is micrometer reading when the wires are just separated and thus no tension is applied to the vessel wall; δ is microscope eyepiece reticule calibration factor (mm per division) and a1 and a2 are vessel end points observed when measuring vessel length using the microscope eyepiece graticule). The program also calculates vessel internal circumference at the nth micrometer reading using the following equation: IC n = 205.6µm + (2 (X n X 0 )) (ICn is internal circumference at the nth micrometer position; Xn is the micrometer reading at said position and Xo is the micrometer reading when the wires are just separated). For each micrometer position an estimate of internal pressure (effect pressure) is calculated using the Laplace equation: 57

58 P n = T n ( IC n 2π ) (Pn is effective pressure at the nth micrometer position). As the stepwise vessel distension is continued, an exponential curve is fitted and distension is continued until the effective pressure exceeds the target transmural pressure (100mmHg for rat mesenteric arteries). Once exceeded, the point of intersection between the function of the exponential internal circumference pressure data curve and the function of the 100mmHg isobar is classed as IC100 (the internal circumference at maximal tension). The internal circumference at 90% maximal tension and thus the corresponding micrometer reading is then calculated using the following equations: IC 90 = 0.9 IC 100 X 90 = X 0 + (IC 90 IC 0 ) 2 (IC90 is internal circumference at 90% maximal tension; X90 is micrometer reading that corresponds to 90% maximal tension) Myograph Protocol Vessels from rats aged 3, 12, 18 and 24 months were challenged with a variety of agonists to assess contractiity in the presence of various agonists. Vessels both with and without PVAT were used to assess the effect of PVAT on vessel contractility. Following a 30 min equilibration period after normalisation, initial viability and endothelial integrity were assessed using 3 x 10-5 mol/l phenylephrine and 1 x 10-5 mol/l acetylcholine (ACh) respectively. Vessels were excluded if constriction to 3 x 10-5 mol/l phenylephrine was <0.9mN/mm or relaxation to 1 x 10-5 mol/l acetylcholine was <80%. Vessels were then incubated for 60 min in the presence/absence of specific agonists/inhibitors: 58

59 +/- 1 x 10-5 mol/l A to activate AMPK +/- 5 x 10-5 mol/l L-NNA to inhibit NOS activity, thus preventing NO production +/- 1 x 10-5 mol/l indomethacin to inhibit COX activity, thus preventing prostaglandin production Dose responses to the thromboxane A2 mimetic U46619 (1 x x 10-6 mol/l) and the α1-adrenoceptor agonist phenylephrine (1 x x 10-5 mol/l) were then constructed, in addition to constriction with high potassium (6 x 10-2 mol/l potassium-enriched PSS (KPSS) (composition given in Appendix). Endothelium-dependent vasodilation following agonist/inhibitor incubation was assesed by applying 1 x 10-5 mol/l ACh to vessels pre-constricted with phenylephrine (3 x 10-5 mol/l). Endothelium-independent relaxation was then assessed by application of 1 x 10-2 mol/l sodium nitroprusside (SNP) to vessels preconstricted with 1 x 10-5 mol/l U Vessels were then washed four times with Ca 2+ -free PSS to reveal any inherent vessel tone Exogenous PVAT Protocol To allow for specific inhibition of enzymes (NOS/COX) with the PVAT alone, exogenous PVAT experiments were performed. For exogenous PVAT experiments, small sections of PVAT were collected and incubated +/- 5 x 10-5 mol/l L-NNA (to inhibit NOS and thus NO production from the PVAT) or 1 x 10-5 mol/l indomethacin (to inhibit COX and thus prostaglandin production from the PVAT) for 60 min. Exogenous PVAT sections were then suspended within the bath in close proximity to the vessels using 40µm wire. The myography protocol described above was then performed. 59

60 2.3.6 Data Collection Values were obtained from the LabChart file created and were taken from the point at which tension had reached a plateau phase after agonist addition. During times of vasomotion the average between peak and trough was used. The statistical methods used to analyse the data obtained are detailed below in Section Pharmacological Agents The AMPK inhibitor A76992 (Cat. No. 3336), NOS inhibitor Nω-nitro-Larginine (L-NNA) (Cat. No. 0664) and thromboxane A2 receptor antagonist U46619 (Cat. No. 1932) were all purchased from Tocris Bioscience (UK). The α1- adrenoceptor agonist phenylephrine (Cat. No. 6126), nitric oxide donor sodium nitroprusside (Cat. No ) and endothelium-dependent vasodilator acetylcholine chloride (Cat. No. A6625) were purchased from Sigma-Aldrich (UK). 2.4 Western Blotting Tissue Homogenisation The arteries and PVAT from an entire mesenteric bed were collected from animals at each age point and snap frozen in liquid nitrogen and stored at -80 C until needed. For agonist-stimulation experiments used to examine the effects of specific agonists on the expression/phosphorylation of key proteins, PVAT or mesenteric arteries from an entire mesenteric bed were incubated at room temperature for 60 min with either 1 x 10-5 mol/l A769662, 3 x 10-6 mol/l U46619 or 1 x 10-5 mol/l phenylephrine before being snap frozen. Frozen samples were 60

61 homogenised in RIPA buffer (Sigma-Aldrich, UK) (90µl for mesenteric small artery samples, 0.8µl/mg for PVAT samples) containing a Protease Inhibitor Cocktail (complete ULTRA Tablets, Mini, EDTA-free, EASYpack, Roche) at 4 C using a manual glass-glass homogeniser. Samples were rotated for 30 min at 4 C before being centrifuged at 12000rpm for 5 min at 4 C and the supernatant collected Protein Assay Samples were assessed for their protein concentration to allow for equal loading of samples into gels which in turn allowed for equal relative loading of the protein of interest. Protein concentrations were measured using the Bradford test; a colorimetric protein assay that allows the quantification of total protein in a sample by measuring the binding of Coomassie brilliant blue G-250 to protein 238. Bovine Serum Albumin (BSA) standards to produce a standard curve were prepared at concentrations of 0, 0.125, 0.25 and 0.5mg/ml. A 10µl aliquot of each sample s supernatant (diluted 1:10 or 1:100 with dh2o if too concentrated to be compared against the standard curve) or BSA standard were added to 990µl of dilute Bradford Reagent (diluted 1:25 with dh2o) and 200µl of this mixture added to a 96 well plate. The absorbance of each sample or the standard curve points was then measured at 595nm using a plate reader (Biorad, UK). Protein concentrations of the samples were then calculated by reference to the BSA standard curve derived from an equation of the line fitted to the standard curve points (Figure 2.3). 61

62 Figure 2.3: Example of the BSA standard curve generated during the Bradford assay. Protein sample concentrations were calculated using the line fitted to the points of the standard curve. All samples were then diluted (either 1, 2 or 4µg/µl depending on the initial concentrations to allow for easier aliquoting) using RIPA/protease inhibitor cocktail and Laemmli buffer (added to a final concentration of 1 in 5) Preparation of Stain-Free Acrylamide Gels Gels were cast (10-well, 12%, 0.75mm thick sodium dodecyl sulfate (SDS) polyacrylamide gels employing stain free technology (See Section 18.2 for recipe)) using the Mini-PROTEAN Tetra Handcast system (Biorad, UK). Stain free technology utilises a gel formulation including a trihalo compound that reacts covalently with tryptophan residues in the sample proteins when exposed to ultraviolet (UV) irradiation. This reaction causes proteins to fluoresce under UV excitation to allow the assessment of protein loading and transfer as well as the 62

63 normalisation of antibody signals to the total protein present in each sample if required SDS Polyacrylamide Gel Electrophoresis (SDS-PAGE) Samples of either homogenised arteries or homogenised PVAT were heated to 95 C to denature proteins and 40µg loaded into the wells of the SDS stain free polyacrylamide gel. Each gel contained samples from each age group to allow for comparisons between gels. Gels were assembled into a BioRad Mini PROTEAN Tetra Cell gel electrophoresis tank filled with Running Buffer (see Section 18.2 for recipe) and sample proteins separated according to size across an electrochemical gradient (60mV while travelling through the stacking portion of the gel, 120mV whilst travelling through the separating portion of the gel). Once proteins were sufficiently separated gels were removed from the tank and images were taken under UV-excitation using the Gel Doc XR+ System (BioRad, UK) to confirm successful protein separation Semi-Dry Protein Transfer Process Gels were then assembled into a transfer sandwich as illustrated below (Figure 2.4). Membranes used were low-fluorescence polyvinylidine fluoride (LF- PVDF), activated for approximately 30 seconds in methanol and then soaked in transfer buffer (see Section 18.2 for recipe) to remove hydrophobicity. Transfer stacks (multiple layers of filter paper) were also soaked in transfer buffer prior to use. Assembled transfer sandwiches were then placed into a Trans-Blot Turbo Transfer system (BioRad, UK) and gently compressed using a roller to remove any air bubbles. Proteins were then transferred electrophoretically for 15 min at 25 volts and 1.3 amps. 63

64 Figure 2.4: Schematic of a transfer sandwich The SDS-PAGE gel is placed onto an activated LV-PVDF membrane which is in turn between two transfer stacks soaked in transfer buffer Processing of Membranes Once transfer was complete, images were taken under UV-excitation using the Gel Doc XR+ System (BioRad, UK) to confirm successful transfer of proteins from gel to membrane and to allow for normalisation to total protein content. Figure 2.5 is an example of an image obtained during this UV-excitation visualisation step. MSA PVAT kd Figure 2.5: Example stain-free blot. The trihalo compound within the gel reacts covalently with tryptophan residues in the samples when exposed to ultraviolet (UV) irradiation, allowing visualisation of all proteins within a sample in addition to the protein size ladder. 64

65 Membranes were then incubated in blocking buffer (5% BSA (Sigma- Aldrich, UK) solution) for 60 minutes at room temperature on a gel rocker (Gyro Rocker, Stuart Scientific, UK). Membranes were then incubated overnight at 4 C with the relevant primary antibody (detailed in Table 2.1) diluted in 0.1% BSA, 0.1% Tween-TBS to a final volume of 10ml. Following overnight incubation, membranes were washed 3 times for 10 minutes in 0.1% Tween-TBS before being incubated with an anti-rabbit secondary antibody (Jackson/ ; Lot ; final concentration 40ng/ml; diluted in 0.1% BSA, 0.1% Tween-TBS) for 60 minutes at room temperature. At the end of secondary antibody incubation membranes were again washed 3 times for 10 minutes in 0.1% Tween-TBS. Table 2.1: Details of Primary Antibodies Used: Antibody target, species, supplier, product number, batch number and final concentration used. Antibody Target Species Supplier/Product No. Final Concentration Used AMPKα Rabbit Cell Signalling/ ng/ml p-ampkα (Thr 172 ) Rabbit Cell Signalling/2535S 260ng/ml enos Rabbit Santa Cruz/SC ng/ml p-enos (Ser 1177 ) Rabbit Cell Signalling/9571S 395ng/ml The secondary antibody used (Donkey Anti-Rabbit, , Conc: 400ng/ml Jackson Immunoresearch, USA) was conjugated to a horseradish peroxidase (HRP) enzyme which catalyses a reaction when a chemiluminescent substrate is added (Clarity Western ECL Substrate Kit, BioRad, UK) to produce 65

66 a luminescent signal proportional to the amount of primary antibody and thus target protein concentration Detection of Antibodies After the chemiluminescent substrate had been added to the membrane and incubated in the absence of light at room temperature for 4 minutes, blots were placed in the Gel Doc XR+ System (BioRad, UK) and exposed until a clear signal was observed without any points of overexposure. (Typically 30-60s for AMPK, p-ampk and enos, 300s for p-enos) Quantification of Data Total protein concentrations for each sample were calculated from the stain free images obtained (Figure 2.5), using ImageLab V5.2.1 for Windows (BioRad, UK). The software allowed for quantification of the total protein present in each individual lane. The densities of individual antibody-bound protein bands within the samples were then normalised to the total protein present in the individual sample, allowing for normalisation without the need to further probe for housekeeping genes. 12, 18 and 24 m.o. samples were expressed as a proportion of control (3 m.o.) to show changes with ageing. 2.5 Immunohistochemistry Sample processing Samples of arteries with and without PVAT were placed in 4% paraformaldehyde (Sigma-Aldrich, UK) for 4 hours at room temperature then transferred to phosphate buffered saline (PBS) solution (Sigma-Aldrich, UK) and 66

67 stored at 4 C until processed. Samples were placed in a tissue processor (Shandon Citadel 2000, Shandon, USA), which progressed samples through an 11-hour cycle comprising a stepwise increase in ethanol concentration, xylene and finally molten wax. Samples were then manually embedded into paraffin wax using a tissue embedder (Shandon Histocentre 2, Shandon, USA) and 5µm sections cut using a microtome (Leica, UK) and attached to polylysine-coated slides (VWR International, USA) Measurement of adipocyte size The histology of PVAT tissue was assessed following haematoxylin and eosin (H&E) staining, which was performed using an automated staining system. Slide-mounted sections were de-waxed by immersion in xylene and rehydrated by progression through decreasing concentrations of ethanol and finally dh2o. Mayer s haemotoxylin stain was then applied to the slides for 2 minutes before washing with dh2o and application of eosin Y solution for a further 2 minutes. Once staining was complete slides were then dehydrated by progression through increasing concentrations of ethanol and finally xylene. Glass cover slips were then applied to the slides using DPX mounting medium (VWR International, USA) and slides visualised using a light microscope (DM5000, Leica, UK). Images of each slide were obtained (DFC450, Leica, UK) and used to calculate adipocyte size using ImageJ V1.47 for Windows (NIH, USA) to trace the adipocyte cell margins (100 consecutive adipocytes per image). 2.6 Statistical Analyses Data are presented as mean±sem with number of animals used (functional experiments) or replicates (Western blot experiments) presented as n= x. 67

68 Data from myograph recordings were analysed and statistical significance determined using (repeated measures) two-way analysis of variance (response in aged vessels versus corresponding response in vessels from 3 m.o. rats) followed by a Bonferroni post test to determine significant differences between results from different aged groups. Power analysis to determine sample sizes for these experiments utilised an expected standard deviation (derived using results from analogous studies) 34,36,138,147,241 with a minimal effect size of interest set at 20%. Using these parameters, a sample size of 6-9 rats per group (aged and control) was calculated to be necessary to achieve 80% power to detect the target effect size (assuming a single-tailed test). Where sample sizes fall below this size, care should be taken in drawing firm conclusions from the data, although trends can be observed and interpreted cautiously. Normal distribution of data was assessed using the Shapiro-Wilk normality testing and where normally distributed, significance was either investigated using one-way ANOVA, two-way ANOVA (with Bonferroni post-hoc testing) or unpaired t-tests. Non-normally distributed data were assessed using the non-parametric Mann Whitney test. Correlations and linear regressions were also performed and statistical significance assessed using the Pearson Correlation Coefficient (R). All statistical analyses were performed using GraphPad Prism V6.01 for Windows (GraphPad Software, USA). Data were considered to be significantly different (*) if, following the appropriate statistical test, a probability value 0.05 was obtained considered significant. 68

69 3 Chapter 3: Results 3.1 The Effect of PVAT on Isolated Mesenteric Small Artery Contractility The anti-contractile capacity of PVAT from young (3 m.o.) male Wistar rats was examined in response to a variety of vasoconstrictors. Small arteries with their PVAT intact constricted to a significantly less degree than those with their PVAT removed in response to increasing concentrations of the thromboxane A2 mimetic U46619 (Figure 3.1, A) and the α1 adrenoceptor agonist phenylephrine (Figure 3.1, B) (p and p= respectively, 2-way ANOVA). There were no significant differences between contractions to high K + (Figure 3.1, C) (NS, 2- way ANOVA). 3.2 The Effect of PVAT on Isolated Mesenteric Small Artery Relaxation PVAT had no effect on the relaxant responses of the small arteries; including both endothelium-dependent, acetylcholine (ACh) induced relaxation (Figure 3.2, A) (NS, 2-way ANOVA) and endothelium-independent, sodium nitroprusside (SNP) induced relaxation (Figure 3.2, B) in small arteries preconstricted with 3 x 10-6 mol/l U46619 (NS, unpaired t-test). PVAT also had no effect on endothelium-dependent, 1 x 10-5 mol/l ACh induced relaxation in small arteries pre-constricted with 3 x 10-5 mol/l phenylephrine (Figure 3.2, C) (NS, unpaired t-test). 69

70 A B C * * ** n=6 n=6 C Figure 3.1: Effect of PVAT on the contractility of isolated mesenteric small arteries. The presence of PVAT had a significant anticontractile effect in small arteries exposed to increasing concentrations of (A) U46619 (p 0.001, 2-way ANOVA) and (B) phenylephrine (p=0.0397, 2-way ANOVA). The presence of PVAT had no effect on contractility to 6 x 10-2 mol/l KPSS (C) (NS, unpaired t-test). n=6 n=6 70

71 Dilation (% constriction to 3 x 10-6 mol/l U46619) Dilation (% constriction to 3 x 10-6 mol/l U46619) Dilation (% constriction to 3 x 10-5 mol/l phenylephrine) A B C -PVAT +PVAT -PVAT +PVAT n=6 n=6 n=6 n=6 n=6 n=6 Figure 3.2: Effect of PVAT on endothelium-dependent and -independent relaxation in isolated mesenteric small arteries. At 3 months, the presence of PVAT had no effect on either endothelium-dependent relaxation to ACh (A) (NS, 2-way ANOVA). or endothelium independent relaxation to 1 x 10-4 mol/l SNP (B) of vessels pre-constricted with 3 x 10-6 mol/l U46619 (NS, unpaired t-test), or on endothelium-dependent relaxation to 1 x 10-5 mol/l ACh of small arteries pre-constricted with 3 x 10-5 mol/l phenylephrine (C) (NS, unpaired t-test). 71

72 3.3 The Role of Nitric Oxide in PVAT Anti-Contractile Effect Nitric oxide synthase (NOS) inhibition was used to investigate the role of NO in the anti-contractile effect of PVAT. In the absence of PVAT, L-NNA incubation had no significant effect on small artery contractility to U46619; however in the presence of PVAT, inhibition of NO production not only reversed the anti-contractile effect, but potentiated contractility to U46619 to a level significantly higher than small arteries in the absence of PVAT (Figure 3.3, A) (p 0.001, 2-way ANOVA). Inhibition of NO production had no effect on small artery contractility to phenylephrine, either with or without PVAT present (Figure 3.3, B) (NS, 2-way ANOVA). Inhibition of NO production had no effect on small artery constriction to 6 x 10-2 mol/l KPSS in either the presence or absence of PVAT (Figure 3.3, C) (NS, 1-way ANOVA). 3.4 The Effect of U46619 and Phenylephrine on enos Phosphorylation Western blotting was used to determine the effects of U46619 and phenylephrine on enos phosphorylation, measured by phosphorylated:total ratio of enos. Incubation with U46619 had no effect on enos phosphorylation in mesenteric small artery samples (Figure 3.4, A) (NS, 1-way ANOVA), but increased (although not significantly) enos phosphorylation in PVAT samples (Figure 3.4, B) (NS, 1-way ANOVA). Phenylephrine incubation resulted in a nonsignificant decrease in enos phosphorylation in mesenteric small artery samples (Figure 3.4, A) (NS, 1-way ANOVA) but had no effect on enos phosphorylation in PVAT (Figure 3.4, B) (NS, 1-way ANOVA). 72

73 ** * Figure 3.3: Effect of NOS inhibition on isolated mesenteric small artery function. Inhibition of nitric oxide synthase with 5 x 10-5 L- NNA in the absence of PVAT had no significant effect on small artery contractions to either (A) U46619 or (B) phenylephrine. In the presence of PVAT, however, inhibition of NO production significantly increased contractility to U46619 (A) (p 0.001, 2-way ANOVA) but not phenylephrine (B) (NS, 2-way ANOVA). Inhibition of NO synthase also had no effect on constrictions to KPSS in either the presence or absence of PVAT (C) (NS, 1-way ANOVA). 73

74 3.5 The Role of Prostaglandins in PVAT Anti-Contractile Effect COX-derived prostaglandins are known to influence vascular contractility. To investigate the contribution of COX-derived prostaglandins in contractile responses to U46619 and phenylephrine, and potentially the anti-contractile effect of PVAT, small arteries in the presence and absence of PVAT were preincubated with indomethacin before U46619 and phenylephrine dose responses performed. COX inhibition significantly reduced contractility to U46619 in both the presence (p 0.01, 2-way ANOVA) and absence (p , 2-way ANOVA) of PVAT (Figure 3.5, A). In response to phenylephrine, COX inhibition significantly reduced contractility in the presence of PVAT (p=0.0388, 2-way ANOVA), but had no effect in PVAT-denuded vessels (Figure 3.5, B). In summary, these initial experiments demonstrate an anti-contractile effect of PVAT across a range of concentrations of U46619 and phenylephrine, with no effect of PVAT on either endothelium-dependent or independent relaxation. The anti-contractile effect of PVAT in the presence of U46619, but not phenylephrine, was lost during NO synthesis inhibition, and potentiated during either NO and cyclooxygenase inhibition for both U46619 and phenylephrine 74

75 A B n=3 n=3 n=3 n=3 n=3 n=3 Figure 3.4: Effect of U46619 and phenylephrine stimulation on enos phosphorylation. Incubation with 3 x 10-6 mol/l U46619 had no effect on enos phosphorylated:total ratio in mesenteric small arteries (MSA) (A) but increased enos phosphorylated:total ratio in PVAT (although increase was not statistically significant) (NS, 1-way ANOVA). Incubation with 1 x 10-5 mol/l phenylephrine resulted in a nonsignificant decrease in enos phosphorylated:total ratio in mesenteric small arteries (A), and had no effect on enos phosphorylated:total ratio in PVAT (B) (NS, 1-way ANOVA). 75

76 ** **** * ** **** *** * ** * * Figure 3.5: Effect of COX inhibition on isolated mesenteric small artery function. Inhibition of cyclooxygenase (COX) with 1 x 10-5 mol/l indomethacin reduced small artery contractility to U46619 in both the presence and absence of PVAT (A) (-PVAT vs. -PVAT + indomethacin p ; +PVAT vs. +PVAT + indomethacin p 0.01, 2-way ANOVA). In response to phenylephrine, COX inhibition significantly reduced contractility in the presence of PVAT (P=0.0388, 2-way ANOVA), but had no effect on contractility in the absence of PVAT (B) (NS, 2-way ANOVA). 76

77 3.6 Factors Contributing to the Anti-Contractile Effect of PVAT To confirm the anti-contractile effect of PVAT was due to a releasable factor and not simply a barrier effect, an exogenous PVAT system was used, where sections of PVAT were suspended in the bath of a PVAT-denuded small artery and dose-responses to U46619 and phenylephrine performed. Exogenous PVAT was shown to exert an anti-contractile effect in response to both U46619 (Figure 3.6, A) (p 0.01, 2-way ANOVA) and phenylephrine (Figure 3.6, B) (p=0.0437, 2-way ANOVA) similar to those seen in PVAT-intact small arteries (Figure 3.1). NO is synthesised in both the endothelium and PVAT. To investigate the source of the NO essential for the PVAT anti-contractile effect during constriction with U46619, sections of exogenous PVAT were incubated with the NOS inhibitor L-NNA before being suspended in the bath of a PVAT-denuded small artery. Similar to the effect seen in PVAT-intact small arteries, incubation of exogenous PVAT with L-NNA prevented the anti-contractile effect of PVAT (p , 2-way ANOVA), and actually increased contractility to U46619 (Figure 3.6, A) (p=0.0472, 2-way ANOVA), but had no effect on contractility to phenylephrine (Figure 3.6, B) (NS, 2-way ANOVA). To determine the contribution of PVATderived prostaglandins to the PVAT anti-contractile effect, exogenous PVAT sections were incubated with the COX inhibitor indomethacin. The anti-contractile effect of exogenous PVAT in response to U46619 was lost following incubation of exogenous PVAT with indomethacin (Figure 3.6, C) (-PVAT vs. +exogenous PVAT + indomethacin NS, 2-way ANOVA). In contrast, indomethacin-incubated exogenous PVAT significantly reduced small artery contractility in response to 77

78 increasing concentrations of phenylephrine (Figure 3.6, D) (p 0.001, 2-way ANOVA). A dual incubation of exogenous PVAT with both L-NNA and indomethacin also had differing effects on small artery contractility in response to U46619 and phenylephrine. During a U46619 dose response, small arteries constricted to a level similar to a PVAT-denuded small artery (Figure 3.6, E) whereas small arteries exposed to phenylephrine constricted to a level similar to a small artery in the presence of control exogenous PVAT (Figure 3.6, F). In summary, these results suggest that the NO essential for the PVAT anti-contractile effect in the presence of U46619 was released from PVAT, rather than the endothelium. They also suggest that PVAT-derived prostaglandins too contribute to the PVAT anti-contractile effect in response to U46619, but are perhaps contributing to constriction in response to phenylephrine. 78

79 * *** NS * *** ** **** A B *** ** **** * * * C D **** **** E F ** *** *** Figure 3.6: Effect of NOS and COX inhibition in exogenous PVAT applied to isolated mesenteric small arteries. Application of exogenous PVAT pre-incubated with 5 x 10-5 mol/l L-NNA (to inhibit NOS) significantly increased small artery contractility to U46619 versus a PVAT-denuded small artery (A) (-PVAT vs. +exogenous PVAT p 0.01; -PVAT vs. +exogenous PVAT +L-NNA p=0.0472; +exogenous PVAT vs. +exogenous PVAT +L-NNA p , 2-way ANOVA) but had no effect in response to phenylephrine (B) (-PVAT vs. +exogenous PVAT p=0.0437, 2-way ANOVA). Application of exogenous PVAT incubated with indomethacin significantly increased small artery contractility compared with control exogenous PVAT in response to U46619 (C) (p=0.0416, 2-way ANOVA), but significantly decreased contractility in response to phenylephrine (D) (p 0.001, 2-way ANOVA). Small arteries in the presence of exogenous PVAT co-incubated with both L-NNA an indomethacin constricted to a similar level as PVAT-denuded small arteries alone in response to U46619 (E) (p=0.0179, 2-way ANOVA), and to a similar level as PVAT-intact small arteries in response to phenylephrine (F). 79

80 3.7 The Effect of Ageing on the Cardiometabolic Profile of Male Wistar Rats A Wistar rat model of ageing was used to investigate the effects of ageing on metabolic parameters and vascular function. A variety of cardiometabolic parameters were therefore measured in male Wistar rats at 3 m.o., 12 m.o., 18 m.o., and 24 m.o.. Body weight increased significantly from an average of 293.2g ±23.8 (n=10) at 3 m.o. to an average of 513.6g ±18.7 (n=10) at 12 m.o. at which time rats had reached full adult body weight, but then remained relatively stable from m.o., with no further significant differences (Figure 3.7, A). Systolic blood pressure was significantly increased from an average of 140mmHg ±3.4 (n=10) at 3 m.o. compared with an average of 153mmHg ±5.4 (n=10) at 24 m.o. (Figure 3.7, B) (p 0.05, 1-way ANOVA). Blood glucose appeared increased in all age groups compared with 3 m.o. (Figure 3.7, C), however this was only significant at 18 m.o. (p 0.05, 1-way ANOVA), with no significant differences between the other age groups. Plasma insulin was significantly elevated at 24 m.o. compared with 3 m.o. (Figure 3.7, D) (p 0.05, 1-way ANOVA) again with no significant differences between the other age groups. Ageing was positively correlated with increasing blood pressure (p=0.0290, non-linear regression), in addition to increased blood glucose (p=0.0194, non-linear regression) and plasma insulin (Figure 3.8) (p=0.0489, non-linear regression). Increased blood pressure was not significantly correlated with body weight from 12 m.o. onwards (once rats had reached adult body weight) suggesting that changes were age- and not weight-dependent (data not shown, Pearson correlation coefficient=0.232). 80

81 **** **** **** * Figure 3.7: Phenotypic changes that occur in male Wistar rats with ageing. A) Body weight had increased significantly at 12 m.o. as rats reached full adult body weight, then remained stable (and significantly larger than 3 m.o.) from m.o. (All ages vs. 3 m.o. p ). B) Systolic blood pressure was unchanged from 3-18 m.o. but was then significantly increased at 24 m.o. compared to 3 m.o * * (p 0.05). C) Blood glucose was only significantly different between 3 and 18 months (p 0.05). D) Plasma insulin was significantly elevated at 24 m.o. compared to 3 m.o. (p 0.05). All data analysed using 1-way ANOVA. 81

82 Figure 3.8: Correlation between ageing and phenotypic characteristics. Ageing was positively correlated with increased systolic blood pressure (A) (p=0.0290), blood glucose (B) (p=0.0194), and plasma insulin (C) (p=0.0489). All data analysed using non-linear regression. R=Pearson correlation coefficient, n=10 for all groups. 82

83 HOMA-IR was calculated from the blood glucose and plasma insulin levels of rats aged 3, 12, 18 and 24 m.o. as a measure of insulin resistance. Rats aged 24 months had a significantly increased HOMA-IR value compared with rats aged 3 months (Figure 3.9, A) (p 0.01, Mann Whitney test). Furthermore, ageing showed a significant positive correlation with HOMA-IR (Figure 3.9, B) (p=0.0120, non-linear regression). 3.8 The Effect of Ageing on Adipocyte Size As changes in adipocyte morphology are known to occur with changes in metabolic profile 242, sections of PVAT were stained with haematoxylin and eosin solution to allow visualisation of adipose tissue morphology and assessment of adipocyte size. PVAT adipocyte size increased progressively from 3 to 18 m.o. (Figure 3.10, A-C) before decreasing again at 24 m.o. (Figure 3.10, D). Adipocytes from 24 m.o. PVAT were, however, still significantly larger than those from 3 m.o. rats (Figure 3.10, E) (p , 1-way ANOVA). 3.9 The Effect of Ageing on Potassium-Mediated Constriction Small artery contractility to 6 x 10-2 mol/l KCl was similar for both PVATintact and PVAT-denuded small arteries at 3, 12 and 24 m.o (Figure 3.11) (NS, 1-way ANOVA). The constrictions elicited by 6 x 10-2 mol/l KCl were unchanged with ageing in both the presence and absence of PVAT (Figure 3.11) (NS, 1-way ANOVA). 83

84 Figure 3.9: Effect of ageing on HOMA-IR. At 24 m.o., HOMA-IR (A) was significantly higher than at 3 m.o. (p 0.01, Mann Whitney test). Ageing was significantly positively correlated with increased HOMA-IR (B). (p=0.0120, non-linear regression, R=Pearson correlation coefficient, n=10 for all groups) 84

85 A 3 months months 12 B 12 months months C 18 months D months E Figure 3.10: Effect of ageing on adipocyte size. Adipocyte Figure 3.10: The effect of ageing size increased progressively on adipocyte size. Adipocyte size from 3 (A) to 12 m.o. (B) and increased from 12 progressively to 18 m.o. from 3 (C). (A) to 12 Adipocyte m.o. (B) and from size 12 to then 18 m.o. (C). decreased Adipocyte from size 18 then to 24 decreased m.o. (D), although adipocytes from 18 to 24 m.o. (D), although remained significantly larger adipocytes remained significantly larger than 3 m.o. (E). All data than analysed at 3 m.o. using (E) (p , 1-way 1-way ANOVA). ANOVA, all significant differences p

86 n=6 n=6 n=10 n=9 n=8 n=8 n=8 Figure 3.11: Effect of ageing on KPSS-mediated constriction of mesenteric arteries. There were no significant differences between contractions elicited by 6 x 10-2 mol/l KCl in PVAT-denuded or PVAT intact small arteries at 3, 12 or 24 m.o.. There were also no significant differences between the constrictions at 3, 12 and 24 m.o. (NS, 1-way ANOVA). 86

87 3.10 The Effect of Ageing on the PVAT Anti-Contractile Effect When stimulated with increasing concentrations of U46619, the presence of PVAT had a pronounced anti-contractile effect on small arteries from 3 m.o. rats (p , 2-way ANOVA). At 12 m.o., the anti-contractile influence of PVAT on U46619 contraction was still present (p=0.0260, 2-way ANOVA), although the magnitude was markedly decreased compared with 3 m.o. arteries. At 24 m.o. this anti-contractile effect has been lost completely (Figure 3.12) (NS, 2-way ANOVA). Small artery contractility in both the presence and absence of PVAT was significantly reduced at 12 m.o. (Figure 3.13, A & B) (p and p , respectively, 2-way ANOVA), as was contractility in the absence of PVAT at 24 m.o. compared with 3 m.o. (Figure 3.13, A) (p , 2-way ANOVA). Contractility to U46619 in the presence of PVAT at 24 m.o., however, was significantly increased compared with 12 m.o. (Figure 3.13, B) (p , 2-way ANOVA). Overall, this resulted in a negative correlation between increasing age and decreasing small artery contractility in response to U46619 in the absence of PVAT (p=0.0213, non-linear regression), but not in the presence of PVAT (NS, non-linear regression) (Figure 3.13, C & D). The reduced contractility of PVAT-intact small arteries versus PVATdenuded small arteries in response to increasing concentrations of phenylephrine was maintained at both 12 and 24 m.o. (Figure 3.14) (p and p=0.0259, respectively, 2-way ANOVA), and ageing was not correlated with changes in small artery in contractility to phenylephrine in either the presence or absence of PVAT (Figure 3.15) (NS, non-linear regression). 87

88 *** *** A 33 months months 24 months B C * * ** * 24 months Figure 3.12: Effect of ageing C on the anti-contractile effect of PVAT in response to U46619 in isolated mesenteric small arteries. In response to increasing concentrations of U46619, the presence of PVAT exerted a significant anti-contractile effect in small arteries from animals aged 3 m.o. (A) (p<0.001) and 12 m.o. (B) (p=0.0260). At 24 m.o. the presence of PVAT had no effect on small artery contractility (C) (NS). All data analysed using 2-way ANOVA. 88

89 *** **** **** A C -PVAT -PVAT B +PVAT Figure 3.13: Effect of ageing on **** **** isolated mesenteric small artery **** **** R = *** ** * D R = **** **** **** ** contractility in response to U Small artery contractilities to U46619 in both the absence (A) and presence (B) of PVAT were significantly reduced at both 12 and 24 m.o. compared to 3 m.o. (2-way ANOVA). There was a significant negative correlation between age and maximum response to U46619 in the absence of PVAT (C) (p=0.0213, nonlinear regression, R= Pearson correlation co-efficient, n=10 for all groups), but no correlation between age and maximum response to U46619 in the presence of PVAT (D) (NS, non-linear regression, R=Pearson correlation coefficient, n=10 for all groups).

90 * A 3 months 12 months 3 months B 12 months 24 months C 24 months ** * *** **** **** * Figure 3.14: Effect of ageing on the anti-contractile effect of PVAT in isolated mesenteric small arteries exposed to phenylephrine. 24 months Over a range of concentration C of phenylephrine, the presence of PVAT exerted a significant anti-contractile effect in small arteries from animals aged 3 m.o. (A) (p=0.0397), 12 m.o. (B) (p ) and 24 m.o. (C) (p=0.0259). All data analysed using 2-way ANOVA. * 90

91 ** ** * R = R = A -PVAT B +PVAT Figure 3.15: Effect of ageing on isolated mesenteric small artery ** ** contractility to phenylephrine in isolated mesenteric small arteries. **** *** * PVAT-denuded small artery contractility to increasing concentrations of phenylephrine (A) was significantly increased at 24 m.o. compared with 12 m.o. (p=0.0296), but was not significantly different between 3 m.o. and 24 m.o. In C D PVAT-intact small arteries (B), constriction to phenylephrine was R = R = significantly reduced at 12 m.o. compared with both 3 and 24 m.o. (vs. 3 m.o. p=0.0395, vs. 24 m.o. p ). There was no significant correlation between ageing and small artery contractility either in the presence (C) or absence (D) of PVAT (NS, non-linear 91 regression, R=Pearson correlation coefficient, n=10 for all groups).

92 A 24 months 24 months B Figure 3.16: Effect of NOS inhibition on isolated mesenteric small artery contractility at 24 months. At 24 months, blockade of NO production with 5 x 10-5 mol/l L-NNA had no effect in either PVAT or +PVAT mesenteric small arteries in response to either U46619 (A) or phenylephrine (B) (NS, 2-way ANOVA). 92

93 In contrast to small arteries from young animals, (Figure 3.3, A), inhibition of NO synthesis with 5 x 10-5 mol/l L-NNA had no effect on small artery contractility to U46619, either in the presence or absence of PVAT (Figure 3.16) in animals aged 24 m.o. (NS, 2-way ANOVA) The Effect of Ageing on the Expression and Phosphorylation of enos in PVAT As experiments using exogenous PVAT suggested that PVAT-derived NOS is essential for the U46619-mediated PVAT anti-contractile effect, expression and phosphorylation (Ser 1177 ) of enos in PVAT from rats aged 3, 12, 18 and 24 months were investigated using Western blot. Total enos expression was non-significantly increased at 24 m.o. compared with 3 m.o. (Figure 3.17, A), while the total amount of phosphorylated enos remained constant (Figure 3.17, B). This resulted in a non-significant decrease in the phosphorylated:total ratio of enos in the PVAT of 24 m.o. rats compared with 3 m.o. rats (Figure 3.17, C) The Effect of Ageing on Endothelium-Dependent and Independent Relaxation Percentage relaxation in response to 1 x 10-5 mol/l acetylcholine was used to assess endothelium-dependent relaxation, which remained unchanged with ageing in the absence of PVAT, but tended to decline with age in the presence of PVAT in small arteries pre-constricted with either U46619 (Figure 3.18, A) or phenylephrine (Figure 3.18, B). 93

94 kDa 100kDa 75kDa kDa 100kDa 75kDa Figure 3.17: Expression and phosphorylation of enos in PVAT. Expression of enos in PVAT appeared to be increased at 12, 18 and 24 m.o. compared with 3 m.o., however this was not significant (A). Levels of phosphorylated enos also appeared increased at 12 and 18 m.o. compared with 3 m.o. but this was insignificant, and were unchanged at 24 m.o. (B). There was a slight nonsignificant decrease in the ratio of phosphorylated:total enos at 18 and 24 m.o. compared with 3 m.o., but no change at 12 m.o. compared with 3 m.o. (C). (All data analysed using 1-way ANOVA; n=3 for all groups). 94

95 Endothelium-independent relaxation, defined as the percentage relaxation in response to 1 x 10-4 mol/l SNP, was unchanged with age in the both the presence and absence of PVAT (Figure 3.18, C). At 3 and 12 m.o., the presence of PVAT had no effect on ACh-induced relaxation (Figure 3.18 A). However, the maintenance of ACh-induced relaxation in the absence of PVAT and concomitant decline in the presence of PVAT with ageing suggest that there was a significant decrease in endothelium-dependent relaxation in the presence of PVAT at 24 m.o. (Figure 3.18, A) (p 0.05, 1-way ANOVA). The decline in endothelial function in the presence of PVAT was also evidenced in the significant negative correlation between increasing age and endothelium-dependent relaxation in PVAT-intact (Figure 3.19, B) (p=0.0272, non-linear regression) but not PVAT-denuded arteries (Figure 3.19, A) (NS, nonlinear regression) The Contribution of Nitric Oxide to Endothelium-Dependent Relaxation with Ageing The contribution of NO to endothelium-dependent ACh-induced dilation was assessed using NOS inhibition in mesenteric arteries with and without PVAT from rats aged 3, 12 and 24 m.o. (Figure 3.20). Inhibition of NOS with 5 x 10-5 mol/l L-NNA had no effect on acetylcholine-induced dilation in either the presence or absence of PVAT in mesenteric arteries from 3 or 12 m.o. rats. At 24 m.o., however, NOS inhibition significantly reduced endothelium-dilation in the absence of PVAT (p 0.05, 1-way ANOVA). 95

96 n=6 n=6 n=5 n=5 n=8 n=8 n=6 n=6 n=5 n=5 n=8 n=8 n=6 n=6 n=5 n=5 n=8 n=8 n=6 n=6 n=5 n=5 n=8 n=8 n=6 n=6 n=5 n=5 n=8 n=8 n=6 n=6 n=5 n=5 n=8 A * B C C Figure 3.18: Effects of ageing on endothelium-dependent and -independent relaxation of isolated mesenteric small arteries. Endothelium-dependent relaxation elicited from arteries pre-constricted with either 3 x 10-5 mol/l phenylephrine (A) or 3 x 10-6 mol/l U46619 (B) tended to decline in the presence of PVAT, but remained unchanged in the absence of PVAT. The presence of PVAT had no effect on endothelium-dependent relaxation at 3 or 12 m.o. (A), however at 24 m.o., the presence of PVAT significantly reduced endotheliumdependent relaxation (A) (p 0.05, 1-way ANOVA). Endothelium-independent relaxation of small arteries pre-constricted with 3 x 10-6 U46619 (C) did not change with ageing (NS, 1-way ANOVA). 96

97 % Dilation to 10µM ACh % Dilation to 10µM ACh A B R = R = Figure 3.19: Correlation between ageing and endothelial function of isolated mesenteric small arteries. There was no correlation between ageing and endothelial function (defined as the dilation to 1 x 10-5 mol/l acetylcholine as a percentage of the pre-constriction generated to 3 x mol/l phenylephrine) in the absence of PVAT (A). However there was a significant negative correlation between ageing and decreased endothelial function in the presence of PVAT (B) (p=0.0272). All data analysed using non-linear regression, R=Pearson correlation coefficient, n=10 for all groups. 97

98 -PVAT +PVAT NS NS ** NS NS NS Figure 3.20: Effect of ageing on the contribution of NO to endothelium-dependent relaxation of isolated mesenteric small arteries. In -PVAT small arteries (A), inhibition of nitric oxide synthase with 5 x 10-5 mol/l L-NNA had no effect on Ach-induced relaxation at 3 and 12 m.o., (NS) but significantly reduced relaxation induced by 1 x 10-5 mol/l acetylcholine at 24 months (p 0.01). In +PVAT small arteries (B), L-NNA also tended to reduce acetylcholine-induced relaxation at 24 months, although this was not statistically significant (NS). All data analysed using 1-way ANOVA. 98

99 In summary, both the endothelium-dependent and -independent relaxations of small mesenteric arteries in the absence of PVAT remained unchanged. However there was a decline in the endothelium-dependent relaxation of small mesenteric arteries in the presence of PVAT at 24 m.o. Ageing also appeared to alter the contribution of NO to endothelium-dependent relaxation, as NO inhibition had a significant effect on ACh-induced relaxation at 24 m.o. but not 3 m.o The Effect of Ageing on the Expression and Phosphorylation of enos in Mesenteric Arteries Total expression of enos was significantly increased in mesenteric small artery samples (including the endothelium) from 12 m.o. rats, before reducing at 24 m.o. back to levels comparable to 3 m.o. (Figure 3.21, A) (1-way ANOVA). The total amount of phosphorylated enos appeared slightly increased at all time points, but differences were not statistically significant (Figure 3.21, B) (NS, 1- way ANOVA). Both total and phosphorylated enos were non-significantly increased in mesenteric arteries from 24 m.o. rats compared with 3 m.o. rats (Figure 3.21, A & B) (NS, 1-way ANOVA). However the proportional increase in phosphorylated enos was smaller than the increase in total enos, resulting in the tendency towards a reduced phosphorylated:total ratio of enos at 24 m.o. compared with 3 m.o. (Figure 3.21, C) (NS, 1-way ANOVA). 99

100 * * 150kDa kDa 75kDa kDa 100kDa 75kDa Figure 3.21: Expression and phosphorylation of enos in mesenteric small arteries. Expression of enos in mesenteric arteries (A) was significantly increased at 12 m.o. compared with 3 m.o but was unchanged at both 18 and 24 m.o. despite appearing increased. Total levels of phosphorylated enos (B) appeared slightly increased at all age groups, but differences were not statistically significant. The ratio of phosphorylated to total enos (C) appeared to be greatly reduced at both 12 and 24 m.o. compared with 3 m.o., however again this was not statistically significant. (All data analysed using 1-way ANOVA; n=3 for all groups). 100

101 3.15 The Effect of Pharmacological AMPK Activation in Young Mesenteric Arteries The effect of pharmacological activation of AMPK with A on vascular function was then investigated, as AMPK is a key kinase upstream of enos activation and furthermore is a downstream target of numerous medications in current clinical use 243. PVAT-denuded and PVAT-intact small arteries from 3 m.o. rats were incubated with the specific AMPK activator A769662, and an increase in AMPK phosphorylation at Thr 172 in both PVAT and arteries was observed (Figure 3.22). In PVAT-denuded arteries, incubation with A significantly reduced contractility to both U46619 (Figure 3.23, A) (p 0.01, 2-way ANOVA) and phenylephrine (Figure 3.23, C) (p=0.0210, 2-way ANOVA), and contractions were similar to those produced in PVAT-intact small artery segments (Figure 3.23, E & F). In the presence of PVAT, AMPK activation with A had no effect on contractility to either U46619 or phenylephrine (Figure 3.23, B & D) The Role of Nitric Oxide in AMPK Mediated Reductions in Contractility In PVAT-denuded small arteries, inhibition of NO synthesis with L-NNA prevented A mediated reductions in contractility in response to both U46619 and phenylephrine, resulting in concentration-dependent contractions comparable to those in control PVAT-denuded small arteries (Figure 3.24 A & B), suggesting that A mediated reductions in contractility in the absence of PVAT are dependent on the production of NO. In PVAT-intact arteries, inhibition of NO synthesis had no effect on responses to A in small arteries exposed to increasing concentrations of either U46619 (Figure 3.24 C) or phenylephrine (Figure 3.24, D). 101

102 A Control +A B Control +A Figure 3.22: The effects of A incubation on AMPK Thr 172 phosphorylation. Incubation with 1 x 10-5 mol/l A resulted in an increase in the phosphorylated:total ratio of AMPK in both mesenteric small arteries (A) and PVAT (B). 102

103 -PVAT +PVAT -PVAT +PVAT Figure 3.23: Effect of AMPK activation on contractility of isolated mesenteric small arteries at 3 months. Activation of AMPK with 1 x 10-5 mol/l A significantly reduced contractility in PVAT small arteries in response to increasing concentrations of U46619 (A) (p 0.01) and phenylephrine (C) (p=0.0210), but had no effect in +PVAT small arteries in response to either U46619 (B) or phenylephrine (D). The contractility of PVAT small arteries in the presence of A was similar to the contractility of +PVAT small arteries exposed to U46619 (E) (-PVAT vs. +PVAT p 0.001) or phenylephrine (F) (-PVAT vs. +PVAT p=0.0397). All data analysed using 2-way ANOVA. 103

104 3.17 The E ffect of Pharmacological AMPK Activation in Aged Mesenteric Arteries To investigate age-related alterations in the AMPK/eNOS functional pathway that may potentially explain changes in vascular function at 24 m.o. or indeed restore said changes, the effects of the specific AMPK activator A were explored in small mesenteric arteries from 24 m.o. rats. In contrast to the effects in small arteries of 3 m.o. rats, A incubation had no effect on contractility in PVAT-denuded small arteries from 24 m.o. rats (Figure 3.25, A & C). However, in the presence of PVAT, small arteries of 24 m.o. rats incubated with the specific AMPK activator A showed significantly reduced contractility over a wide range of concentrations of both U46619 and phenylephrine (Figure 3.25, B & D) (p 0.01 and p 0.001, respectively, 2-way ANOVA) The Role of Nitric Oxide in AMPK-Mediated Effects on Contractility of Isolated Small Mesenteric Arteries of 24 Month Old Wistar Rats Inhibition of NO synthase with L-NNA prevented the A induced reduction in contractility seen in PVAT-intact small arteries in response to increasing concentrations of U46619 (Figure 3.26, A). The AMPK-induced reduction in contractility in response to increasing concentrations of phenylephrine, however, was only partially prevented by incubation with L-NNA (Figure 3.26, B). 104

105 A -PVAT B -PVAT Figure 3.24: Effect of NOS inhibition on (n=5) the anti-contractile effect of AMPK activation in isolated mesenteric small arteries. Inhibition of nitric oxide synthase with 5 x 10-5 mol/l L-NNA prevented the A mediated reduction in contractility to both U46619 (A) (p 0.001) and phenylephrine (B) (-A vs. C +PVAT D +PVAT +A p=0.0210, +A vs. +A L-NNA p=0.0273) seen in the absence of PVAT. In the presence of PVAT, NOS inhibition with L-NNA in the presence of A has no effect on small artery contractility to either U46619 (C) or phenylephrine (D). All data analysed using 2-way ANOVA. 105

106 A -PVAT B +PVAT Figure 3.25: Effect of AMPK activation ** ** ** on isolated mesenteric small artery contractility at 24 months old. Activation of AMPK with 1 x 10-5 mol/l A significantly reduced contractility in +PVAT small arteries in response to increasing concentrations of U46619 (B) (-A C -PVAT D +PVAT vs. +A p 0.01) or phenylephrine (D) (-A vs. +A769662p 0.001) but had no effect in -PVAT small arteries in *** ** *** *** response to either U46619 (A) or phenylephrine (C). All data analysed by 2-106

107 A +PVAT B +PVAT * Figure 3.26: Effect of NOS inhibition on A mediated effects on isolated mesenteric small artery contractility at 24 months. Inhibition of NOS with 5 x 10-5 mol/l L-NNA prevented the A mediated reduction in contractility to U46619 in PVAT-intact small arteries (A). In response to phenylephrine (B), NOS inhibition only partially reversed the A mediated reduction in contractility in +PVAT small arteries (p=0.1557, 2-way ANOVA). 107

108 3.19 The Effect of Ageing on the Expression and Phosphorylation of AMPK in Mesenteric Arteries and PVAT The expression and phosphorylation of AMPK, a kinase upstream of enos activation, were also determined using Western blot on PVAT and mesenteric small artery samples. Total expression of AMPK was significantly reduced in the mesenteric arteries of 24 m.o. rats compared with 3 m.o. (Figure 3.27, A) (p 0.05, 1-way ANOVA). Although appearing increased at 12 and 18 m.o. compared to 3 m.o., there were no significant differences in the levels of phosphorylated AMPK between any age groups (Figure 3.27, B). The reduction in total AMPK with no change in phosphorylated AMPK resulted in a non-significant increase in the phosphorylated:total ratio of AMPK (Figure 3.27, C). AMPK expression in PVAT was significantly increased at 24 m.o. compared with 3 m.o. (Figure 3.28, A) (p 0.05, 1- way ANOVA), however the total amount of phosphorylated AMPK was significantly decreased (Figure 3.28, B) (p 0.05, 1-way ANOVA), resulting in a reduced phosphorylated:total ratio of AMPK in PVAT from 24 m.o. rats compared with 3 m.o. rats The Effect of Ageing on O-GlcNac Modification of Proteins Due to the increased blood glucose and significantly increased serum insulin seen at 24 m.o. compared with 3 m.o., and the altered phosphorylation of enos and AMPK in PVAT, the level of O-GlcNac modification of proteins was investigated in PVAT samples from rats aged both 3 and 24 m.o. There was around a 5-fold increase in the O-GlcNac modification of proteins in the PVAT of 24 m.o. rats compared with 3 m.o. rats (Figure 3.29). There were noteworthy bands of O-GlcNac modification at approximately 60 and 140 kda, which may indicate O-GlcNacylation of both AMPK and enos. 108

109 A Total AMPK * 100kDa kDa 63 50kDa B Total Phosphorylated AMPK kDa kDa 63 50kDa C Phosphorylated:Total AMPK Figure 3.27: Expression and phosphorylation of AMPK in mesenteric small arteries. Expression of AMPK in mesenteric arteries was unchanged at 12 and 18 m.o. and significantly reduced at 24 m.o. compared with 3 m.o. (A) (p=0.0144, unpaired t test). Total levels of phosphorylated AMPK appeared increased at 12 m.o. and 18 m.o. compared with 3 m.o., but differences were not statistically significant. At 24 m.o., phosphorylated AMPK was decreased compared with 18 m.o. (p=0.0300, unpaired t- test) and unchanged compared with 3 m.o.. The phosphorylated:total ratio (C) appeared to be increased in all age groups compared with 3 m.o., but differences were not statistically significant. (All data analysed using 1-way ANOVA; n=3 for all groups). 109

110 Total AMPK A * 75kDa kDa B Total Phosphorylated AMPK * 100kDa 63 75kDa * 50kDa C Phosphorylated:Total AMPK Figure 3.28: Expression and phosphorylation of AMPK in PVAT. Expression of AMPK was significantly higher at 24 m.o. compared with 3 m.o. (A), and also appeared increased at both 12 and 18 m.o. Total levels of phosphorylated AMPK were unchanged at 12 and 18 m.o. but significantly reduced at 24 m.o. (B). The ratio of phosphorylated/total AMPK (C) appeared increased at both 12 and 18 m.o., but changes were not statistically significant; phosphorylated:total ratio remained unchanged at 24 m.o. (All data analysed using 1-way ANOVA; n=3 for all groups). 110

111 Age Age months 24 months Figure 3.29: O-GlcNac modification of PVAT. Proteins from the PVAT of 24 m.o. rats show increased O-GlcNac modification of proteins compared with those from 3 m.o. PVAT (n=2 in both groups). 111

112 4 Chapter 4: Discussion 4.1 Key Findings This investigation has highlighted several previously unidentified changes that occur within the vasculature in this rat model of ageing, as well as furthering our understanding of some of the mechanisms underlying PVAT function in young animals. PVAT surrounding mesenteric small arteries was shown to express enos and furthermore was identified as the source of the NO previously shown to be instrumental in the PVAT anti-contractile effect in the presence of U The involvement of PVAT-derived prostaglandins in the PVAT anti-contractile effect in the presence of U46619 was also identified. Inhibition of the PVAT anticontractile effect in the presence of phenylephrine was shown to rely on neither NO nor prostaglandins, the latter being shown to potentially contribute to phenylephrine-induced contraction. This study demonstrated that the contractility of mesenteric small arteries to U4619 was reduced with ageing, and that the anti-contractile effect of PVAT in the presence of U46619 was lost with ageing. The level of contractility to phenylephrine, as well as the anti-contractile effect of PVAT in the presence of phenylephrine were both shown to be maintained with ageing, indicating differing mechanisms of contraction generation and stimulation of anti-contractile factors released from the PVAT between the two agonists. There was a preservation of endothelium-dependent relaxation in PVATdenuded vessels, but the presence of PVAT was shown to have a deleterious effect on endothelium-dependent relaxation at 24 months. 112

113 Furthermore age-related changes in the expression and activation of enos and AMPK were identified that may provide some explanation for the alterations in function described above. 4.2 Vascular Contractions and Relaxation The presence of PVAT is well-documented to exert an anti-contractile paracrine effect that occurs in response to numerous vasoconstrictors, including noradrenaline 150,151 and serotonin 130 as well as the ones employed in the present study, U ,145 and phenylephrine 130,244. The α1-adrenoceptor agonist phenylephrine and the thromboxane A2 mimetic U46619 have both been shown to cause robust concentration-dependent contractions in isolated rat mesenteric small arteries. Phenylephrine binding to α1-adrenoceptors activates the receptor s Gq subunit, which ultimately results in increased [Ca 2+ ]I and VSMC contraction 245. Binding of U46619 to TP receptors also stimulates this pathway through Gq activation, but works predominantly through activation of the G12-13 subunit and instead increases the Ca 2+ -sensitivity of the VSMC contractile machinery 32. Therefore phenylephrine-induced contraction is more reliant on Ca 2+ release and subsequent increases in [Ca 2+ ]i than U46619, which instead increases Ca 2+ sensitivity. This in turn affects the way in which these constriction methods are most susceptible to opposition. Phenylephrine-mediated constriction relying upon Ca 2+ entry through voltagegated Ca 2+ channels makes it more sensitive to hyperpolarisation. Alternatively, factors released from the endothelium or PVAT that reduced basal [Ca 2+ ]i would have a greater effect on U46619-mediated contraction, as changes in [Ca 2+ ]i would be less effective in generating constrictions in the context of reduced [Ca 2+ ]i. Given the different mechanisms of action, both agonists were used in 113

114 order to elucidate some of the downstream signalling pathways involved in vascular contractility in the presence and absence of PVAT. In the present study, PVAT exerted an anti-contractile effect over a range of concentrations of U46619, including a reduced maximum constriction, suggesting that U46619 stimulates the release of a PVAT-derived relaxing factor. Previous data regarding the effect of PVAT on U46619-induced constriction is limited, although PVAT has previously been shown to exhibit an anti-contractile effect to U46619 in the rat thoracic aorta 144,145 and female rat small mesenteric arteries 246, but not the male rat superior mesenteric artery 247. An anti-contractile effect of PVAT in the presence of phenylephrine was also seen across a range of concentrations, although maximum constrictions in the presence and absence of PVAT are similar. The anti-contractile effect of PVAT on phenylephrine-induced contractions has been previously demonstrated in numerous vascular beds, including the rat aorta and the mesenteric arteries of both male 146 and female 147 rats. An alternative adrenergic agonist, noradrenaline, has previously been shown to elicit a pronounced PVAT anti-contractile effect 151, which is likely due to downstream effects of binding to β3-adrenoceptors 32,198 (which are present on perivascular adipocytes 248 ) in addition to contractile VSMC α1- and α2-adrenoceptors 249,250. Phenylephrine, however, selectively binds only VSMC α1-adrenoceptors 249. The different mechanisms of action of phenylephrine and U46619 (increased Ca 2+ influx and Ca 2+ sensitisation respectively) could provide insight into the different PVAT anti-contractile effects in opposition of their relative constrictions. 114

115 Unlike U46619 or phenylephrine, high extracellular K + causes vasoconstriction through a receptor-independent membrane depolarisation that subsequently activates L-Type Ca 2+ channels 251 resulting in increased [Ca 2+ ]i and vasoconstriction. It has previously been shown that the anti-contractile effect of PVAT can be abolished by the presence of high extracellular K + or blockade of KCa channels 130,252, suggesting that the downstream anti-contractile pathway involves efflux of K + out of VSMCs, most likely inducing VSMC hyperpolarisation. In the present study PVAT had no effect on constrictions elicited by high extracellular K +, which supports this hypothesis. High extracellular K + would counteract any PVAT-induced K + efflux and moreover trigger Ca 2+ entry that would also overcome any effects of a PVAT anti-contractile factor on [Ca 2+ ]i. 4.3 NO-Mediated Opposition of Contraction NO has previously been shown to be involved in the PVAT anti-contractile effect in isolated mesenteric arteries from Wistar rats in response to noradrenaline 150. In the present study, inhibition of NO synthesis in PVATdenuded mesenteric arteries had no effect on constrictions generated by either U46619 or phenylephrine. This suggests that, if the endothelium is having an anticontractile effect, NO is not the main vasodilator responsible. This result agrees with the findings of several studies, that show no effect of NOS inhibition on contractility to either U or phenylephrine in mesenteric small arteries 254. The lack of effect of NOS inhibition in phenylephrine induced constrictions does, however, contradict the findings of Briones et al. (2005) 255, who showed a significant increase in mesenteric small artery contractility to phenylephrine following NOS inhibition with L-NAME, albeit in Sprague-Dawley rats and using 115

116 a physiological solution containing higher calcium and glucose than the one used in the present study. NO is not the only endothelial regulator of vascular contractility; as mentioned previously the endothelium can also cause VSMC hyperpolarisation through the release of a transferable hyperpolarising factor and/or direct transfer of hyperpolarisation through myoendothelial gap junctions, known collectively as the endothelium-dependent hyperpolarising factor (EDHF) 97. This endotheliumdependent hyperpolarisation is also induced by ACh binding to muscarinic receptors 256 which increases [Ca 2+ ]i in endothelial cells, inducing a hyperpolarisation that then spreads to the VSMCs. The ability of ACh to induce endothelial cell hyperpolarisation thus means that EDHF is likely to be responsible for the robust ACh-induced relaxations seen in the presence of NOS inhibition in the present study. The number of myoendothelial gap junctions, and thus the contribution of EDHF to endothelium-dependent relaxation is known to increase as vessel diameter decreases 107. Moreover, ACh also stimulates the production of the vasodilator PGI2 from endothelial cells 257. Binding of PGI2 to the IP receptor in VSMCs increases intracellular camp 258, activating protein kinase A which, like protein kinase G, also phosphorylates MLCP, in addition to opening KATP channels 259 on VSMCs. Ultimately, these results indicate that endotheliumdependent, ACh-induced relaxations in the 3 m.o. animals are not totally dependent on NO release. In the presence of PVAT, NOS inhibition had divergent effects on vessel contractilities to U46619 and phenylephrine. In tissues contracted with U46619, NOS inhibition not only reversed the anti-contractile effect of PVAT but actually potentiated constriction to a level significantly higher than in PVAT-denuded small 116

117 arteries in the presence of the NOS inhibitor. In vessels constricted with phenylephrine, however, NOS inhibition had no effect on contractility in PVATintact small arteries. Taken together these findings suggest that TP receptor and α1-adrenoceptor activation have differing effects in stimulating the NO-mediated anti-contractile capacity of PVAT. Potentially, U46619 activation of PVAT TP receptors increases NO production from the PVAT, whereas phenylephrine activation of PVAT α1-adrenoceptors does not. This hypothesis is corroborated by the increased phosphorylated:total ratio of enos as determined by Western blot in the present study, seen in PVAT incubated with U46619, and lack of effect with phenylephrine. Moreover a previous study actually demonstrated a significant decrease in NO production in isolated aortic endothelial cells stimulated with both 1 x 10-7 and 1 x 10-6 mol/l phenylephrine 260. This differing effect on enos phosphorylation and thus NO production is likely the result of the differing signal transduction pathways downstream of agonist-receptor binding, described above. Similar to phenylephrine-induced constriction, inhibition of NO production had no effect on K + -mediated constriction either in the presence or absence of PVAT, suggesting that K + -mediated Ca 2+ entry into VSMC is sufficient to overcome any functional antagonism by any PVAT or endothelium-derived NO. Of note is the increased contractility to U46619 of PVAT-intact small arteries incubated with L-NNA compared with PVAT-denuded small arteries. This increase could be the result of a significant PVAT pro-contractile effect, previously counteracted by the overwhelming release of NO. PVAT has previously been shown to release a plethora of pro-contractile adipokines and other transferable factors, including angiotensin II 261, chemerin 262 and superoxide 155 as well as 117

118 vasoconstrictor prostaglandins 263. Further studies using specific inhibitors for each of these candidates may elucidate the mechanism behind this procontractile effect and provide a potential therapeutic target for reducing vessel contractility and thus reducing blood pressure. 4.4 PVAT is the Primary Source of Anti-Contractile NO It has previously been shown that the PVAT anti-contractile effect is the result of the release of a transferable factor, and not simply a barrier effect 252,264. This has generally been demonstrated using a solution transfer technique, in which the bath solution from a stimulated PVAT-intact vessel was transferred to the bath of a stimulated PVAT-denuded vessel which elicits a significant relaxation,74,75. The transferable nature of the PVAT anti-contractile effect in the present model was alternatively confirmed by the application of sections of exogenous PVAT to PVAT-denuded small arteries. However it is always worth acknowledging that a barrier effect, also known as diffusion hindrance), may possibly contribute to the reduced actions of vasoconstrictors/vasodilators in the presence of PVAT 150, as would agonist absorption by lipids present within perivascular adipocytes. It is well established that inhibition of NO synthesis in PVAT-intact vessels abolishes the PVAT anti-contractile effect in response to a variety of agonists and in a variety of tissues 130,151,154,252, but it was previously unknown whether this NO is sourced from the PVAT or the endothelium. The former is suggested as a source of NO given previous studies showing the expression of enos in cultured adipocytes 152,153 and periadventitial adipose tissue surrounding the saphenous vein 265. However, to our knowledge, the expression of enos in the PVAT 118

119 surrounding small resistance arteries has not previously been shown. In this present study, the expression of enos in both mesenteric small arteries and PVAT taken from around the mesenteric small arteries was confirmed by Western blotting. This robustly demonstrates the ability of the PVAT to produce NO. Subsequently, the application of exogenous PVAT allowed the determination of the source of NO essential for the U46619-mediated anticontractile effect by exclusively inhibiting NOS within the PVAT before addition back into the bath of a PVAT-denuded vessel. Inhibition of NOS in either exogenous PVAT or in PVAT-intact arteries did not have any effect on contractility to phenylephrine. In response to U46619, however, NOS inhibition in the PVAT alone mimicked the effect seen with NOS inhibition in a PVAT-intact vessel. This provides evidence that the NO responsible for the U46619-mediated anticontractile effect is derived from PVAT, most likely perivascular adipocytes, and is not the result of PVAT-mediated release of endothelium-derived NO. 4.5 The Effect of COX-Derived Prostaglandins on Vascular Contractility In addition to NO, endogenous COX-derived prostaglandins can also influence contractility through a variety of actions within the vasculature including either vasoconstriction or vasodilation (reviewed by Félétou et al., ). Previous studies have shown an almost complete prevention of the contraction elicited by U46619 in rat mesenteric small arteries pre-incubated with indomethacin 94,267. Furthermore, it has also been shown that U46619 activation of the TP receptor causes significant production of PGE2 within the mesenteric bed 94. This PGE2 can elicit significant vasoconstriction through activation of 119

120 VSMC EP3, FP and TP receptors 268 ; a contractile pathway that would indeed be abrogated under COX inhibition. Indeed the present study also showed that indomethacin incubation resulted in significantly reduced mesenteric small artery contractions to U46619 in the absence of PVAT, in addition to a similar reduction in contractility in the presence of PVAT, potentially suggesting a net contractile influence of artery/endothelium and PVAT prostaglandins stimulated by TP receptor activation with U In response to phenylephrine, COX inhibition again significantly reduced contractility in the presence of PVAT, however it had no effect on contractility in PVAT-denuded vessels. A similar lack of effect of COX inhibition has also been shown in mesenteric vessels constricted with noradrenaline 267, however it does again contradict the findings of the study by Briones et al. (2005) 255 who show a significant decrease in the contractility of PVAT-denuded vessels across a range of phenylephrine concentrations in the presence of indomethacin. It is worth noting however that in Brione s study, responses with and without indomethacin were generated from the same preparation, and that responses were expressed at a percentage of maximum KCl-induced contraction, which may diminish the effect seen. Several papers have shown increased α-adrenoceptor mediated contractility in the presence of exogenous PGE In the presence of PVAT, a known source of PGE2 82, contractility to phenylephrine could also therefore be potentiated. The reduction in contractility to phenylephrine in PVAT-intact vessels during COX inhibition further suggests that PVAT is indeed exerting a significant 120

121 anti-contractile effect during phenylephrine constriction, but PGE2-mediated amelioration of phenylephrine-induced contraction may be diminishing this effect. Small artery contractility to U46619 in the presence of PVAT pre-incubated with indomethacin was similar to that of PVAT-denuded arteries. This suggests that the anti-contractile effect of PVAT usually seen in the presence of U46619 involves COX-derived prostaglandins released from the PVAT, as COX inhibition has prevented this reduction in contraction. Furthermore it also suggests that the major source of contractile prostaglandins within the vasculature is the endothelium. As previously discussed, U46619-mediated activation of TP receptors is known to result in PGE2 production and increases in vessel contractility, however PGE2 is also known to activate enos 86 through binding to the EP4 receptor, which is expressed in adipose tissue 272. Following PGE2 binding to EP4, there is a subsequent dephosphorylation of enos at its inhibitory Thr 495 site, increasing production of NO which subsequently diffuses to VSMCs resulting in an accumulation of cgmp in VSMCs 86. This suggests a direct increase in NO bioavailability as a result of PGE2 production from COX enzymes present in the PVAT that is stimulated, potentially, by TP receptor activation. Further studies using a specific EP4 receptor antagonist, such as ER , would elucidate whether PGE2 is involved in the anti-contractile effects of PVAT in response to U The failure of L-NNA to increase small artery contractility to U46619 during COX inhibition further suggests that the pro-contractile factor previously discussed is likely COX-derived, and could indeed be PGE2 itself. 121

122 In the presence of indomethacin-incubated exogenous PVAT, PVATdenuded vessels showed increased contractility compared to control PVAT-intact vessels. Indomethacin incubation of a PVAT-intact vessel, however, decreased contractility compared to a control PVAT-intact vessel. This difference (i.e. that the addition of COX inhibition within the vessel itself significantly reduces contractility), suggests that the pro-contractile effects of vessel-derived prostaglandins outweigh the relaxant effects of PVAT prostaglandins, leading to the net effect of a reduction in contractility in PVAT-intact vessels pre-incubated with indomethacin. Inhibition of COX activity had no effect on phenylephrine-mediated contractility of PVAT-denuded vessels, but significantly reduced small artery contractility to phenylephrine in PVAT-intact vessels. PVAT-denuded vessel contractility to phenylephrine was also significantly reduced in the presence of exogenous PVAT pre-incubated with indomethacin. This again alludes to a net pro-contractile effect of COX-derived products in PVAT-intact small artery preparations. However in contrast to effects seen in U46619-contracted vessels, it appears that PVAT-derived prostaglandins produced in the presence of phenylephrine are solely pro-contractile, rather than additionally providing anticontractile effects. This could be the result of different downstream effects of TP and α1- adrenoceptor activation; potentially that PVAT α1-adrenoceptor and TP receptor activation both stimulate the production of vasoconstricting prostaglandins, but TP receptor activation additionally stimulates counteracting anti-contractile factors, such as vasorelaxant prostaglandins, or NO. As previously mentioned, U44169 activation of TP receptors may cause PGE2-induced NO-mediated 122

123 reductions in contractility which may be the basis of the anti-contractile effect of PVAT during U46619 constriction. For phenylephrine, the lack of NO phosphorylation in PVAT in response to phenylephrine stimulation could be an indication of a lack of stimulated PGE2 production, perhaps suggesting the production of an alternative vasoconstrictor prostaglandin. Alternatively, potentially, a phenylephrine-mediated decrease in NO production 260 as mentioned previously, may counter enos activation, PGE2 stimulated or otherwise, leaving pro-contractile actions to dominate the net effect of COX activation. Further studies using specific prostaglandin receptor blockers may help to identify the specific prostaglandins produced and evaluate their relative contributions to phenylephrine-mediated vasoconstriction. One important aspect to consider under COX inhibition is the possible shift of the arachidonic acid pathway from production of prostaglandins to the production of epoxyeicosatrienoic acids, which not only possess vasodilatory properties 274, but have also been shown to act as endogenous inhibitors of TP receptors 275. This latter effect could potentially explain the dramatic reduction in contractility to U46619 seen during COX inhibition, and potentially the reduction seen in ageing in this study. 4.6 Summary of Findings from Young Animals The present studies identified some important findings regarding the mechanisms involved in PVAT anti-contractility in small mesenteric arteries from 123

124 young animals. A summary of these conclusions with reference to the figure demonstrating the finding is provided below in Figure 4.1. A B Figure 4.1: Summary of contractile/anti-contractile mechanisms in small arteries from young animals. In the presence of U46619 (A), Cyclooxygenase (COX) within the PVAT produces prostaglandins with a net anti-contractile effect, whereas COX in PVAT-denuded arteries produces prostaglandins with a net pro-contractile effect. In the presence of U46619, NO synthase (NOS) in PVAT produces NO (NO) which has an anti-contractile effect, whereas the activity of NOS in the mesenteric small artery has no effect on contractility. In the presence of phenylephrine (B), COX within the PVAT produces prostaglandins that have a net pro-contractile effect, whereas COX in the small arteries has no effect. NOS in both PVAT and small arteries has no effect on contractility. 124

125 4.7 The Effect of Ageing on Cardiometabolic Parameters The effects of ageing on several cardiometabolic parameters were measured in rats aged 3, 12, 18 and 24 m.o. Body weight increased significantly between 3 and 12 m.o. as rats reached maturity and then remained relatively stable between 12 and 24 months, a similar pattern seen in other studies using ageing Wistar rats 276. Furthermore the weights of rats at 12 and 24 m.o. were also comparable to other studies using Wistar rats at these time points 276. While age-related increases in systolic blood pressure are well documented in humans, the effect of ageing on systolic blood pressure in rodents is less clear-cut. Studies vary in their results, showing increases 277,278, decreases 279 and no changes 280 in rat systolic blood pressure with ageing. In the present study systolic blood pressure was constant up to 18 months in male Wistar rats, and significantly increased at 24 m.o.. These ages are equivalent to circa 45 and 60 years respectively in humans, age groups between which the incidence of hypertension approximately doubles 7. The non-invasive tail cuff method used to measure blood pressure in this study is sometimes considered less accurate than invasive techniques such as radiotelemetry 281, which involves this insertion of a probe into the aorta 282. However many studies have been published that support the validity of the technique 283, including at least two that obtained parallel readings using both tail cuff and telemetry methods 284,285, with one study by Feng et al. 285 concluding that the tail-cuff method provides accurate blood pressure measurements over the physiological range of blood pressure. Fasting blood glucose appeared increased at 12, 18 and 24 m.o. rats compared with 3 m.o. Although this increase was only significant at 18 m.o., 125

126 fasting blood glucose was significantly positively correlated with age overall. This rise in blood glucose was accompanied by a trend of increasing plasma insulin with ageing. One study by Larkin et al. (2009) 286 in male Fischer 344 rats showed increased plasma insulin at 18 and 28 months compared with 8 months, similar to the increases shown here. However, that study also reported no changes in fasting plasma glucose with ageing. This could, however, be the result of different measurement technique, strain differences or differences in the diets provided during ageing, although body weights were comparable to those presented in this study. HOMA-IR is a well-established indicator of insulin resistance first defined in In this study ageing was significantly positively correlated with increased HOMA-IR, suggesting that insulin resistance is a feature of increasing age in this rodent model, similar to age-related insulin resistance seen in humans 288. As obesity is known to be a significant risk factor for hypertension, any confounding effect of body weight was investigated using correlation analysis. There was a lack of correlation between body weight and systolic blood pressure in rats aged 12 m.o. onwards (once adult body weight was reached) (R=0.232), meaning that changes in blood pressure are likely the result of ageing alone and not simply increases in body weight. This is therefore comparable to age-related hypertension, which in humans is independent of obesity 289, which reciprocally can influence blood pressure independent of ageing 290. In obesity, increases in PVAT adipocyte size are well documented 138,242 and are associated with inflammation and oxidative stress as a result of hypoxia 151, altered adipokine secretion 291 and insulin resistance 292. Indeed 126

127 several studies have shown a prevention/reversal of increased adipocyte size with drugs that target oxidative stress and insulin resistance As similar substrates (overt hyperglycaemia and probable insulin resistance) are seen in this model of ageing, adipocyte size was measured in rats from all age groups. Adipocyte size increased steadily from 3 to 18 months. Although adipocyte size decreased from 18 to 24 m.o., it remained significantly larger than at 3 m.o.. This is consistent with age-related changes in epididymal and perirenal fat depots described by Kirkland & Dobson (1997) 296, who showed an increase in adipocyte size in middle aged rats (aged months) which decreased in aged rats (aged months). The authors concluded that the ability of adipocytes to accumulate lipid declines with advanced age, which would correlate with the decrease in adipocyte size seen here between 18 and 24 m.o PVAT samples. Furthermore, changes within the PVAT local environment that occur with ageing can impair adipogenesis 186 which too could contribute to a reduction PVAT mass in the most advanced age group compared to early time points. Although the changes in adipocyte morphology and surrounding environment in ageing involve similar processes to those seen in obesity, such as hypoxia and oxidative stress 179, they are likely to occur through different mechanisms, as the two processing have distinct genetic fingerprints 178. With respect to all cardiometabolic measurements it is also worth pointing out that the rats used in the present study were externally sourced and purchased at their experimental weight. Since the mortality of male Wistar rats at 24 m.o. is approximately double that at 18 m.o. (mortality rates supplied by Charles River), there could be a bias at 24 m.o. towards those that have a healthier cardiometabolic phenotype (and consequently increased longevity), a possible 127

128 confounding factor that should be taken into consideration with all the data presented. 4.8 The Effect of Ageing on PVAT Anti-Contractile Effect At the start of this project, although a loss of PVAT anti-contractile effect and concomitant increases in blood pressure were well characterised in obesity and hypertension 151,172,297, the effects of ageing on the anti-contractile influence of PVAT on artery function were unknown. The PVAT anti-contractile effect in small arteries constricted with U46619 was maintained at 12 months (albeit at a markedly reduced level), but then lost at 24 months. This convergence of the contractilities of PVAT-intact and PVATdenuded vessels to U46619 is the result of significantly increased contractility of PVAT-intact vessels at 24 months compared with 12 months, whilst PVATdenuded contractility remained the same. This increase in PVAT-intact vessel contractility suggests changes in the PVAT itself; either a reduction in the production of PVAT-derived relaxant factors and/or increase in PVAT-derived contractile factors stimulated by TP receptor activation between 12 and 24 months. In the present study, the anti-contractile effect of PVAT in arteries contracted with phenylephrine was maintained with ageing. As discussed previously, the anti-contractile effects of PVAT differ in the presence of U46619 and phenylephrine; for example both NO and prostaglandins being instrumental in the former but not the latter. The preservation in PVAT anti-contractile effect with ageing suggests that the mechanisms involved in the anti-contractile effect 128

129 of PVAT in the presence of phenylephrine are perhaps unaffected, but those in the presence of U46619 are altered with ageing. 4.9 The Effect of Ageing on Vascular Contractility The contractility of PVAT-denuded small arteries to U46619 was significantly reduced at both 12 and 24 months compared with 3 months This reduced contractility to U46619 seen with ageing corresponds with other studies, such as Kang et al. (2007) 42, who demonstrated a dramatic reduction in U46619-induced constriction in coronary arterioles from 24 m.o. versus 3 m.o. male Fisher 344 rats and Félétou et al. (1994), who demonstrated a reduced responsiveness to U46619 of male Sprague Dawley rat abdominal aortas at 12 m.o. compared with those from 3 m.o. rats 43,298. The reduced response to U46619 appears to be tissue specific, as U46619-mediated constriction is preserved in the aged rat thoracic aorta 44,45, but not the abdominal aorta (as mentioned above) 43. In a human study, PVATdenuded mesenteric arteries exhibited a lack of correlation between ageing and changes in contractility to U , however factors such as medication-use, underlying pathology and inflammation, and increased vessel diameter may explain these differences. Furthermore, the youngest patients used were approximately 20 years of age, equivalent to 8 m.o. in rats, by which time there may already have been significant reductions in U46619-mediated contractility, as seen at 12 m.o. in this study. There are many potential reasons for the reduced contractility to U46619 seen with ageing. TP receptor activation causes VSMC contraction primarily through activation of the Rho kinase pathway, increasing the calcium sensitivity 129

130 of the VSMC contractile machinery 32. Modifications of the Rho pathway are unlikely to explain the reduced contractility, since both the expression and activity of RhoA have previously been shown to be upregulated in both the aorta and basilar arteries of aged Sprague Dawley rats 300. Downregulation of TP receptor expression is also unlikely to be the cause of the reductions in contractility to U46619 as it has been shown to remain constant, or even increase, with ageing in the vasculature of male rodents 42,301. Quantification of TP receptor expression via Western blot would confirm whether the same maintenance with ageing occurs specifically in the PVAT, providing insight into the effects of ageing on TP receptor mediated anti-contractile effects. As discussed earlier, TP activation by U46619 causes the release of PGE2 94 from the endothelium or VSMCs. COX enzyme expression has been widely shown to be upregulated with ageing and, moreover, the PGE2 EP4 receptor, which is known to induce vasorelaxation, has also been shown to have increased expression on VSMCs with ageing, while vasoconstriction EP receptor expression remains constant 301. This results in a shift in the predominantly expressed VSMC prostaglandin receptor from the vasoconstricting EP3 receptor at approximately 6 m.o. to the vasorelaxant EP4 receptor at approximately 18 m.o If such changes occurred as early as 12 months, this could explain the reduced contractility at 12 and 24 m.o.. In these age groups, PGE2 produced in the endothelium following COX activation by TP receptors would then greatly oppose constriction, due to the increased proportion of relaxatory EP4 receptors on the VSMCs. Alternatively, the TP receptor is known to be susceptible to desensitisation 305 which can occur via PGE2 binding to the EP1 receptor 306. As 130

131 there is increased basal PGE2 release from the intact mesenteric bed at 24 m.o. compared with 3 m.o. 34, this could therefore result in VSMC TP receptor desensitisation. TP receptor desensitisation can also occur through DP receptor activation by PGD2 307, synthesis of which is known to be increased in the vascular endothelium of aged Wistar Kyoto rats 301. These hypotheses remain speculative, however, and further investigation into specific age-related changes in TP receptor sensitivity, prostaglandin receptor expression and prostaglandin synthesis in this model would be needed to fully elucidate the mechanisms involved. Small artery contractility to phenylephrine was preserved at 24 months compared with 3 months, although contractility was unexpectedly decreased in arteries from the 12 m.o. group. The reason for this reduction is unclear; it could be a genuine physiological change, within either the endothelium or smooth muscle, or it could be a result of the potential predisposition towards healthy 24 month old rats. Overall, however, ageing was not significantly correlated with any changes in small artery contractility to phenylephrine, either in the presence or absence of PVAT. The preservation of contractility between arteries at 3 and 24 m.o. correspond with the findings of other studies investigating α1-adrenergic receptormediated constrictions with ageing. For instance contractility to phenylephrine is preserved within skeletal muscle small arteries from both aged rats 308 and in accelerated senescence mouse model of ageing 129. Furthermore, phenylephrineinduced IP3 accumulation in mesenteric small arteries has been shown to be preserved in m.o. rats 309, indicating that vasoconstrictor pathways downstream of α1-adrenceptors, at least as far as IP 3, are unaffected by age. 131

132 This preservation of contractility has also been shown to occur with other adrenergic agonists. For example in whole mesenteric bed preparations there is no change in pressor response to noradrenaline at 24 months compared with 3 months 34, nor is there a change in noradrenaline contractility in the rat aorta 35 or mesenteric arteries 36 of aged rats. It is worth noting that the maintenance of phenylephrine-induced constriction may be gender-specific, as one study has shown a significant leftward shift in the concentration response curve to phenylephrine in 26 m.o. female Fisher 344 rats compared with their 3 m.o. counterparts, although maximum responses were unaffected 310. A study using human mesenteric arteries also showed no correlation between ageing and small artery contractility to noradrenaline 299, however there is the caveat of increased vessel diameter (1-3mm in human arteries versus 200µm in rat arteries), medication use and underlying pathology/inflammation within the cohort which make direct comparisons difficult. Small artery constrictions elicited by high extracellular K + remained unchanged at 24 m.o. compared with 3 m.o. in both the presence and absence of PVAT. A study on PVAT-denuded mesenteric arteries from Fisher 344 rats aged 3 and 26 m.o., also showed no effect of ageing on K + -elicited constrictions, in both vessels from either male or female rats 310. It appears that this maintenance of K + -elicited constriction with ageing may occur in a variety of tissues, as it has also been reported in basilar arteries 311, skeletal muscle arteries 312, coronary arteries 127 and the aorta 280. This is further supported by the preserved L-Type Ca 2+ channel density and therefore functional expression, previously shown in VSMCs with ageing

133 4.10 The Effect of Ageing on enos Expression, Activation and Effects on Vascular Contractility The effect of NOS inhibition, shown by this study to abolish the PVAT anticontractile effect in small arteries from 3 m.o. rats constricted with U46619, was investigated in the arteries of 24 m.o. rats. NOS inhibition had no effect in either the presence or absence of PVAT, suggesting the presence of PVAT NOproduction dysfunction. In young rats, NOS inhibition in PVAT-intact arteries increased contractility to a level significantly greater than PVAT-denuded arteries, suggesting the probable U46619-stimulated secretion of a contractile factor(s) from PVAT. This increase in contractility was no longer produced by NOS inhibition of 24 m.o. PVAT-intact vessels, suggesting a reduced action of a U46619-stimulated contractile factor with ageing, either through reduced production or downstream activity. The cause of the apparent age-related decline in PVAT-derived NO was investigated using Western blot to determine the expression and activation of enos in the PVAT of rats aged 3, 12, 18 and 24 m.o. Expression of enos in PVAT appeared to increase with age, although increases did not reach statistical significance, even at 24 months. This lack of significance is most likely due to the low replicate numbers, which if increased to 5 or 6 could show a true increase in expression. Total levels of phosphorylated (i.e. active) enos in PVAT, however, remained constant with ageing. Increases in enos expression, as suggested here, often occur alongside seemingly paradoxical decreases in NO bioavailability 314,315 and have been demonstrated in the aortic endothelium of aged rats 45,301. Such increases in expression can occur as the result of transcriptional and posttranscriptional 133

134 effects of hydrogen peroxide, a product of NAD(P)H oxidase (NOX) enzymes 315. H2O2 is a hallmark of oxidative stress, which is known to occur in adipose tissue with ageing 179. Such oxidative stress also occurs in the vasculature with ageing, where it results in increased enos uncoupling 316, a state in which enos activity shifts from the production of beneficial NO to deleterious reactive oxygen species. A similar oxidative stress-mediated increase in enos uncoupling could occur within the PVAT with ageing, and could explain the lack of PVAT anti-contractile effect at 24 months. Uncoupling of enos could also explain why there is an apparent reduction in NO bioavailability despite no significant change in the level of phosphorylated enos within the PVAT. Other potential explanations could be increased production of the endogenous NOS inhibitor asymmetric dimethylarginine 317, known to occur with ageing 318, or increased arginase activity, which is known to occur in response to oxidative stress 319 and would compete with enos for available L-arginine. Indeed one study demonstrated an age-related increase in vascular arginase activity, inhibition of which restored enos coupling and vascular NO production 119. Further investigations would be needed to define the mechanisms behind the apparent reduction in NO bioavailability seen in this model of ageing. Moreover only one phosphorylation site (Ser 1177 ) was investigated, and it would be beneficial to observe the effects of ageing on alternative phosphorylation sites, such as the activating site Ser 617, Ser and the inhibitory sites Thr In addition to phosphorylation at various sites, enos is also susceptible to other post-translational modifications including acylation, nitrosylation, glycosylation, glutathionylation and acetylation (reviewed by Heiss & Dirsch, ), Additional studies using a NOS activity assay to determine the level of NOS 134

135 activity 323, as opposed to crude observations of phosphorylation status would provide insight into the actual activity of enos within the PVAT The Effect of Ageing on Vascular Prostaglandins Time constraints and aged animal availability mean that investigations into the role of COX-derived prostaglandins in the U and phenylephrinemediated contractility of PVAT-intact and PVAT-denuded vessels in aged animals are unfortunately yet to be performed. There are numerous previously documented age-related changes in prostaglandin synthesis and activity within the PVAT-denuded vasculature. There are differential changes in the synthesis of specific prostaglandins within the vasculature; for example the production of vasodilatory PGI2 from rat aortic endothelial cells is decreased with ageing 324, whereas expression of the enzyme responsible for synthesising vasoconstricting TxA2 is upregulated 301. Several studies have also shown increases in vascular COX expression and changes in the proportional expression of the various prostaglandin receptor subtypes on both VSM and endothelial cells 301. There are currently no studies, however, investigating the effect of ageing on adipose-derived prostaglandins. This leaves the field ripe for further investigation into the effect of ageing on this specific facet of PVAT function, especially given the contribution of prostaglandins to the PVAT anti-contractile effect in young animals identified by the present study. By inhibiting COX-derived prostaglandin production in the arteries and/or the PVAT, and by differentially targeting COX-1 and COX-2, it may elucidate the involvement of prostaglandins 135

136 in vascular contractility with ageing and provide insight into the mechanism involved in age-related changes in mesenteric small artery and PVAT function The Effect of Ageing and PVAT on Endothelial Function ACh binds to muscarinic receptors on endothelial cells causing the release of several endogenous vasodilators (including NO 325, H2S 326 and prostaglandins 257 ) which ultimately robustly oppose vessel constriction induced by numerous vasoconstrictors 138,171,172. In the present study, ageing had no significant effect on ACh-induced relaxation in PVAT-denuded arteries. Despite the apparent paradigm of agerelated endothelial dysfunction 327,328, several studies demonstrate maintained endothelium-dependent vasodilation in PVAT-denuded arteries with ageing. For example ageing has no effect on ACh-induced relaxation of coronary arteries from 24 m.o. Fisher 344 rats 127, carotid arteries from lean Zucker rats 128 or mesenteric arteries from an accelerated senescence mouse model of ageing 129. In PVAT-intact vessels there was a non-significant tendency for ageing to reduce endothelium-dependent relaxation. This was supported by a significant negative correlation between increasing age and ACh-induced relaxation in PVAT-intact, but not PVAT-denuded, arteries. In young animals, the presence of PVAT has no effect on endotheliumdependent relaxation, an indication of the inability of PVAT-derived relaxant factors to potentiate ACh-induced endothelium-dependent relaxation. At 24 m.o., however, the presence of PVAT had a significant deleterious effect on endothelium-dependent relaxation. The mechanism behind this negative influence is unknown. It could potentially be the result of increased reactive 136

137 oxygen species that could quench ACh-induced NO release or an increase in PVAT-derived contractile factors that ACh-induced relaxation pathways are unable to functionally antagonise fully. PVAT-derived factors in turn could also stimulate the release of an endothelium-derived contractile factor. Indeed ageing has previously been associated with increases in endothelium-dependent contractions 329,330. One such endothelium-derived contractile factor, endothelin- 1, is known to be upregulated in ageing 124 and its production can be stimulated by PVAT-derived factors such as TGF-β 331, TNF-α 332 and IL This secretion of a contractile factor from PVAT is, however, perhaps unlikely given the maintenance of the phenylephrine-induced PVAT anti-contractile effect and the previously discussed potential loss of a U46619-stimulated PVAT contractile factor. Regardless of cause, this deleterious effect of PVAT on endothelial function has been demonstrated numerous times in the setting of obesity 132,133,138, but has previously been unidentified in ageing and may provide a potential therapeutic target for the treatment of in vivo dysfunction in endothelium-dependent vasodilation. Endothelium-dependent relaxation results from the release of relaxant factors, such as NO, as well as endothelium-dependent hyperpolarisation. The contribution of endothelium-derived hyperpolarisation to endothelium-dependent relaxation has been shown to decrease with both increasing vessel diameter 334,335 and ageing 327,336,337. For example one study by Mantelli et al. (1995) 338 showed no effect of NOS inhibition on ACh-induced relaxation of PVATdenuded mesenteric arteries at 2 m.o. but a significant decrease at 18 m.o. from approximately 70% to approximately 30% relaxation 338. In agreement with this, 137

138 this study showed that NOS inhibition had no effect on ACh-induced endotheliumdependent dilation in 3 or 12 m.o. animals, but attenuated ACh-induced relaxation at 24 m.o. (although this difference was not significant in the presence of PVAT). The reduction in ACh-induced relaxation seen with NOS inhibition in the 24 m.o. age group of this study (77% to 18%) is comparable to the reduction reported by Mantelli et al. mentioned above 338. These data provide some evidence for a decrease in endothelium-derived hyperpolarisation with ageing, with a concomitant increase in NO-mediated endothelium-dependent relaxation, as suggested by Fujii et al However, unlike in the study of Fujii et al., AChinduced relaxation was maintained with ageing in this study. The age-related increase in the contribution of NO to endotheliumdependent vasodilation shown in this study could be the result of altered enos functionality in the mesenteric arteries of aged rats. As such, Western blotting was used to determine the expression and activation of enos in PVAT-denuded arteries from 3, 12, 18 and 24 m.o. rats. Expression of enos was significantly increased in the mesenteric small arteries of 12 m.o. rats before steadily decreasing again from 12 m.o. to 24 m.o, where expression was not significantly different from 3 m.o. A similar pattern, but to a lesser extent, is observed in the levels of phosphorylated enos. Comparing the youngest and oldest groups, this maintenance of enos expression and phosphorylation is consistent with a preservation of endothelial-function with ageing. The increases in enos expression in PVAT with ageing (described above) that are potentially the result of enos dysfunction and oxidative stress are not seen in the mesenteric small arteries. This could suggest a lack of enos uncoupling in the vasculature, again correlating with preserved endothelium-dependent relaxation, which is in fact 138

139 increasingly reliant on NO with age. Increases in small artery expression of enos have previously been shown with ageing 304, where they occur with concomitant dysfunction in endothelium-dependent relaxation, thus supporting the idea that enos dysfunction occurs in tandem with increases in enos expression, as shown in PVAT in the present study. To examine endothelium-independent vasodilation, or essentially the ability of the VSMCs to relax, the NO donor sodium nitroprusside was used, which is readily metabolised to NO, a key endogenous vasodilator, within the vasculature 340. NO activates guanylyl cyclase in the smooth muscle, increasing intracellular cgmp and causing protein kinase G-mediated phosphorylation of myosin light chain phosphatase (MLCP) and vasorelaxation. Similarly to ACh-induced relaxation, in young animals the presence of PVAT has no effect on endothelium-independent SNP-mediated relaxation. This suggests that the robust relaxations induced by SNP are incapable of potentiation by PVAT-derived relaxing factors. Age also had no effect on SNP-induced relaxation, indicating a preservation of endothelium-independent relaxation, i.e. a preservation of NO sensitivity of VSMCs with ageing. Some studies have, in contrast, shown age-dependent decreases in SNP-elicited vasodilation which correlate with decreased soluble guanylyl cyclase expression 341,342. However others 343, such as Briones et al. (2005) 255, report no change in soluble guanylyl cyclase expression and subsequently no change in SNP-elicited vasodilation with ageing. Taken together, these findings suggest that preservation of SNP-induced dilation is consistent with preserved VSMC soluble guanylyl cyclase expression. Measurement of the expression of soluble guanylyl cyclase with ageing in this 139

140 model may confirm this proposed relationship and provide explanation of the preserved endothelium-independent relaxation The Effect of Ageing on AMPK Expression, Activation and Effects on Vascular Contractility In the present study, the phosphorylated:total ratio of enos in PVAT was reduced with ageing, indicative of reduced activity of upstream activators of enos. One important upstream activator of enos is the threonine/serine kinase, AMPK, known to phosphorylate enos in both the endothelium 65 and adipocytes 217. As such, the expression and activation of AMPK in PVAT from 3, 12, 18 and 24 m.o. rats were also investigated. Expression of AMPK in PVAT was increased at 24 months compared with 3 months, as has been shown to occur in other tissues, such as rat skeletal muscle 344 (although some studies have shown no changes in AMPK expression in skeletal muscle with ageing 233,345,346 ). However, consistent with the reduced phosphorylated:total ratio of enos, there was indeed a reduction in the total level of phosphorylated (active) AMPK in the PVAT of 24 m.o. rats compared with 3 m.o. rats, despite the significant increase in expression. Decreases in AMPK phosphorylation have been shown to occur with ageing, with the majority of studies again investigating activity in skeletal muscle 233,346. This decreased level of activation, as demonstrated by the decreased phosphorylated:total ratio, may also result from the reduced activity of upstream activators such as LKB1, CaMKKβ and increased [Ca 2+ ]i. Little is known about the effects of ageing on LKB1 or CaMKKβ expression, and thus further experiments to determine their expression and level of activation in the PVAT of 140

141 aged rats may elucidate the reasons for the reduced AMPK activation seen with ageing. The potential dysfunction in PVAT NO production and loss of the anticontractile capacity of PVAT in the presence of U46619 prompted a preliminary investigation into the potential for restoring ex vivo PVAT enos function at 24 months, through activation of its upstream kinase AMPK. AMPK was chosen due to its proposed activation by numerous therapeutic agents already in clinical use, such as statins 243, metformin 347 and resveratrol 348. The thienopyridine A was chosen as an experimental exogenous activator of AMPK which mimics both effects of AMP (allosteric activation and inhibition of dephosphorylation of Thr 172 ) 210. In mesenteric small arteries from 3 m.o. rats, AMPK activation with A significantly reduced PVAT-denuded artery contractility to a level similar to that of a vessel with intact PVAT. However, in the presence of PVAT, A had no effect on contractility. These data suggest that AMPK activation within the vasculature/endothelium can mimic the effects of the presence of PVAT, and perhaps further suggest that the PVAT anti-contractile effect may be similarly mediated, through AMPK activation. Further investigations using a specific AMPK inhibitor would be needed to confirm this latter point, although current commercially available AMPK inhibitors are unfortunately relatively unspecific. Indeed it has been shown that TP receptor activation leads to significant AMPK Thr 172 phosphorylation 349, which correlates with the involvement of NO in the U46619-mediated anti-contractile response. This is further supported by the complete prevention of AMPK-mediated reductions in contractility in the presence of the NOS inhibitor L-NNA, shown in the present study. Previously, reductions 141

142 in contractility seen in the presence of simvastatin, shown to be AMPK-mediated, have also been revealed to be completely reliant on NO production 253. Phenylephrine can also phosphorylate AMPK at Thr 172, although transiently 350 and to a much lesser degree than U , perhaps explaining the lack of enos phosphorylation stimulated by phenylephrine. Furthermore, in rat neonatal ventricular myocytes at least, phenylephrine has also been shown to phosphorylate AMPK at its inhibitory Ser 485/491 site 352, suggesting that it may actually impair AMPK and thus enos activity. This may potentially explain the (non-significant) reduction in small artery enos phosphorylation following phenylephrine incubation shown in the present study. Although minimal, the effect of phenylephrine on AMPK activation through Thr 172 phosphorylation can be exaggerated in the presence of an exogenous activator of AMPK 351. This suggests that in the presence of phenylephrine, A may be activating AMPK and increasing NO production, counteracting the contraction induced by phenylephrine. This is supported by the ability of AMPK to reduce contractility to phenylephrine, potentially by stimulating the NO pathways traditionally downstream of U46619 stimulation. The complete prevention of AMPK-mediated reductions in phenylephrine contractility by NOS inhibition lend weight to this hypothesis. In mesenteric small arteries from 24 m.o. rats, AMPK activation significantly reduced contractility both to U46619 and phenylephrine in the presence, but not the absence, of PVAT. As shown previously, the U mediated anti-contractile effect of PVAT was lost at 24 m.o., where there was also putative enos uncoupling and decreased NO bioavailability within the PVAT. Here exogenous activation of AMPK is therefore potentially reducing the 142

143 contractility of PVAT-intact vessels by restoring the anti-contractile capacity of PVAT seen in the mesenteric arteries of younger rats. Indeed it has been shown that AMPK activation can, in addition to directly phosphorylating enos, actually reverse enos uncoupling, as activated AMPK is known to supress the actions of NAD(P)H oxidases (mechanisms reviewed by Song & Zou (2012) 353 ). AMPK activation may also provide further benefits by reversing the potentially H2O2- mediated enos dysfunction in aged rats PVAT. Recent research has certainly uncovered the potential for AMPK-mediated restoration of PVAT and endothelial function. For example, one study 354 demonstrated that fructose feeding induced a PVAT phenotype that had a detrimental effect on aortic endothelium-dependent vasodilation (similar to the effect of ageing on PVAT function seen in this study). The development of this deleterious PVAT phenotype was shown to be prevented with either metformin or resveratrol treatment in vivo, agents both thought to work through activation of AMPK 354. Other studies have shown direct effects of pharmacological AMPK activation on enos function 355, such as one investigation by Wu et al. (2012) 356 that demonstrated an increase in enos phosphorylation, NO release and upregulation of mitochondrial uncoupling protein-2 (UCP-2) in cultured endothelial cells as a result of AMPK activation. The involvement of NO in the A mediated reductions in contractility to U46619 was confirmed by its complete prevention in the presence of NOS inhibition. However NOS inhibition only partially prevented A mediated reductions in contractility to phenylephrine, suggesting alternative α1 adrenoceptor-mediated effects of AMPK activation in aged rats. AMPK is known to have direct effects on the contractile machinery of VSMCs by phosphorylating and thus desensitising myosin light chain kinase 207 in addition to potentially 143

144 inducing VSMC hyperpolarisation through indirect activation of BKCa channels on VSMC membranes 357,358. However, these additional effects are not revealed by NOS inhibition in the PVAT-denuded arteries of 3 m.o. rats, or of 24 m.o. rats exposed to U Further investigation into the downstream effects of AMPK activation in young and old mesenteric arteries exposed to U46619 and phenylephrine may provide insight into the differing contributions of NO to AMPKmediated reductions in contractility. Since the endothelium-dependent relaxation of PVAT-denuded arteries was not diminished in the 24 m.o. rats, it may have been predicted that A incubation within PVAT-denuded small arteries would still reduce contractility to U46619 and phenylephrine, as it did at 3 m.o.. However, in the present study, A incubation had no effect on the contractility of PVAT-denuded vessels exposed to either U46619 or phenylephrine. To explore potential causes of this reduced responsiveness to A769662, the expression and phosphorylation of AMPK in both mesenteric arteries and PVAT was investigated using Western blotting. There was a significant reduction in the expression of AMPK in the mesenteric small arteries of 24 m.o. rats compared with 3 m.o. rats, which could result in a lack of A mediated effects, as there is less AMPK available for activation. Furthermore, as AMPK is upstream of enos phosphorylation, a reduction in AMPK expression could also explain the slight (non-significant) decrease in the phosphorylated:total ratio of enos in the mesenteric arteries of 24 m.o. rats compared with 3 m.o.. However the total amount of phosphorylated (active) AMPK was unchanged at 24 m.o. compared with 3 m.o., suggesting that activation of enos by AMPK would be preserved with ageing. Further 144

145 investigations are needed to clarify the direct effects of AMPK phosphorylation on enos phosphorylation in both young and old small arteries and PVAT. In the PVAT there was a significant reduction in phosphorylated AMPK, despite a significant increase in expression, suggesting a decrease in endogenous activation of AMPK. The reductions in contractility achieved in PVAT-intact small arteries of 24 m.o. rats following A incubation could therefore potentially be due to exogenous increase in AMPK activity, potentially to the endogenous level seen in young animals. Taken all together it therefore appears possible to restore age-related PVAT dysfunction, hitherto unknown to exist, through pharmacological activation of AMPK ex vivo. One benefit of AMPK activation as a target for therapeutic treatment is the availability of agents known to activate AMPK that are already in clinical use; as mentioned above, statins 243, metformin 347 and resveratrol 348 have all been shown to activate AMPK. Furthermore, exercise is also known to be a potent activator of AMPK 359 as well as an efficient restorative measure combatting endothelial dysfunction 121. By employing these already wellestablished therapies, or by developing new, vascular-specific treatments, it may therefore be possible to target PVAT dysfunction and improve or even reverse age-related vascular dysfunction and therefore, potentially, age-related hypertension The Effect of Ageing on O-GlcNacylation AMPK and enos are proteins that are both susceptible to alternative modifications in addition to activation via phosphorylation. Deleterious O-GlcNac modification of proteins occurs in conditions of high glucose or, in cells expressing 145

146 the Glut4 transporter, high insulin. This O-GlcNacylation has recently been shown to inhibit the activity of enos 360 and AMPK 197 caused significant O-GlcNacylation in PVAT, thus potentially explaining the alterations in enos and AMPK activity. Results showed that levels of O-GlcNacylation were approximately 5 times higher in PVAT from aged animals, with significant bands present at the approximate molecular weights of both AMPK and enos. These preliminary results warrant further investigation into the levels of O-GlcNacylation of proteins in other tissues, in particular the vascular endothelium and smooth muscle which could provide further insight into the effects of hyperglycaemia and hyperinsulinaemia on vascular function in aged animals. Moreover, co-immunoprecipitation of enos and/or AMPK with O-GlcNacylation would give an indication of the level of modification of these specific proteins, which may explain age-related changes in their function Conclusions This study investigated the impact of ageing on vascular function in male Wistar rats, especially in the context of hypertension, which was observed in the most advanced age group. Overall, this study has revealed significant decreases in both vasoconstriction and PVAT anti-contractile effect in response to a thromboxane A2 agonist. These alterations may be the result of significant agerelated PVAT dysfunction, in particular the loss of the thromboxane-induced PVAT anti-contractile effect. Furthermore, age-related changes in PVAT function result in it having a potentially detrimental influence on endothelium-dependent vasodilation. Alterations in enos and AMPK expression, activation and functionality that occur with ageing may explain this PVAT dysfunction and preliminary experiments have suggested a role of O-GlcNacylation, as a result of 146

147 hyperglycaemia and/or hyperinsulinaemia, in its pathogenesis. This previously unidentified PVAT dysfunction may prove a potentially valuable therapeutic target for combatting alterations in vascular function shown to occur with ageing. In a similar fashion to its known effects on endothelial dysfunction, in vitro activation of AMPK in the vasculature can seemingly reverse this PVAT dysfunction, restoring its anti-contractile capacity. Thus the use of AMPK-activators already available in clinical practice may be a viable strategy for pharmacological treatments aimed at improving age-related declines in vascular function in the clinical setting. Many of the mechanisms involved in age-related dysfunction (such as changes in cardiometabolic profile, the loss of a PVAT anti-contractile effect and alterations in endothelial function) are very similar to those involved in obesity and metabolic syndrome. These similarities not only allow us to glean valuable information on disease pathogenesis from studies done in obesity, but furthermore suggest that the vascular dysfunction occurring naturally with increasing age can be accelerated by obesity and potentially other life-style factors Limitations and Future Work The small n-numbers used in this study, due to limited animal availability, make it relatively difficult to draw wider conclusions on the implications of the work. The limited availability of aged animals also meant that some key experiments, such as the effect of ageing on prostaglandin activity within the vasculature were unfortunately not performed. 147

148 Some key future experiments to give the present study greater impact and context could include: Identification of the purported PVAT pro-contractile factor and the effects of ageing on expression of the factor (and its receptor) Identification of the specific prostaglandins involved in the PVAT anticontractile effect in the presence of U46619 Investigations into the downstream effects of AMPK activation with regards enos activity and how AMPK activity is altered with ageing Determination of the effects of ageing on Ca 2+ sensitivity of the VSMC contractile machinery Determination of the effects of ageing on prostaglandin synthesis and receptor expression and sensitivity Determination of the effects of ageing on AMPK s upstream activators, LKB1 and CaMKKβ Investigations on vascular contractility with ageing in the presence of a specific AMPK inhibitor Determination of the effects of ageing on the O-GlcNacylation of specific proteins within the vasculature, namely enos and AMPK 148

149 Figure 4.2: Mechanisms contributing to age-related hypertension. Ageing was previously known to cause changes to vascular smooth muscle and the endothelium which in turn contribute to changes in vascular function, such as vascular stiffening, inflammation and oxidative stress, all of which can contribute to the pathogenesis of hypertension. This study has provided some evidence for additional dysfunction within the perivascular adipose tissue that, through potential effects on endothelial function, oxidative stress and vascular contractility, may also contribute to the pathogenesis of age-related hypertension. 149