.363, ‘. .. .a s 3.5 .5 .3. . lirmw, . A, F. . a. “db Xian! I. (T... . . . :- e. .93? filiu .4. A... 1...? ..m..... u 3.7 «(3&5 at?“ u T. ... . 1...,» a?%§%§m : , Egg §+s%§:§1 . gag. . ‘ .....¢. “if?“ LIBRARY 30:,» Michigan State | 0 University This is to certify that the dissertation entitled Dyslipidemia and Diabetic Retinopathy: Effect of n6 and n3 Polyunsaturated Fatty Acids (PUFA) on Inflammation in human Retinal Endothelial Cells presented by Weiqin Chen has been accepted towards fulfillment of the requirements for the Ph. D degree in Microbiology and Molecular Genetics was Mafia Maj£ rofessor’s Signature 9’ / Q /c>.§" / I Date MSU is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2/05 c:/ClRCIDateDue.lndd-p.15 DYSLIPIDEMIA AND DIABETIC RETINOPATHY: EFFECT OF N6 AND N3 POLY-UNSATURATED FATTY ACIDS (PUFA) ON INFLAMMATION IN HUMAN RETINAL ENDOTHELIAL CELLS BY WEIQIN CHEN A DISSERTATION Submitted to Michigan State University In partial fulfillment of the requirements For the degree of DOCTORAL OF PHILOSOPHY Department of Microbiology and Molecular Genetics 2005 ABSTRACT DYSLIPIDEMIA AND DIABETIC RETINOPATHY: EFFECT OF N6 AND N3 PUFA ON INFLAMMATION IN HUMAN RETINAL ENDOTHELIAL CELLS BY WEIQIN CHEN Early diabetic retinopathy (DR) has been recognized as a low-grade chronic inflammatory disease. The mechanism(s) leading to inflammatory conditions in the diabetic retina are not well understood, but likely involve diabetic hyperglycemia and dyslipidemia. The effects of hyperglycemia have been studied in detail; while the role of dyslipidemia in the development of DR has received less attention. Diabetes induces a decrease in the major n3-PUFA, docosahexaenoic acid (DHA22;6,.3) in the plasma and the retina with a shift toward a higher n6/n3 PUFA. The increase in n6/n3 PUFA ratio can profoundly affect the inflammatory state in the retinal endothelium due to the proinflammatory role of n6-PUFA and the anti-inflammatory effect of n3-PUFA. We first demonstrated that n6-PUFAs have a profound proinflammatory effect in human retinal vascular endothelial cells (hRVE) by inducing intercellular adhesion molecule (ICAM)—1 and vascular cell adhesion molecule (VCAM)-1 expression and leukocyte adhesion to hRVE. The induction was through lipoxygenase (LOX) pathway as only LOX, but not COX or MOX inhibition blocked n6-PUFAs response in hRVE cells. To assess the anti-inflammatory effects of n3-PUFA, the signaling of major inflammatory cytokines upregulated in diabetic eyes, TNFa, IL-lB and VEGF165, was first characterized. All three cytokines stimulated CAMS expression in hRVE through NFIcB activation. DHA22;6,,3 dramatically decreased the cytokines induced CAM expression in hRVE cells. A hypothesis that DHA22;6,,3 exerts the anti-inflammatory effects through direct activation of nuclear receptors PPARs was first addressed. Activation of PPARot downregulated cytokines induced CAMS expression. However, saturated 16:0, n6-PUFAs and DHA22;6,,3 could all activate PPARot similarly. Moreover, DHAzmM could not only prevent nucleus NFKB binding, but also inhibit IicBa phosphorylation and degradation implying that DHA22;6,.3 works upstream of IKBa phosphorylation and not at the nuclear receptors level. The hypothesis that DHAZZW treatment results in modification of membrane lipids that affects signal transduction in specific membrane microdomains, caveolae/lipid rafts was next addressed. Biochemical fiactionation coupled with mass spectrometry was employed to characterize proteins of caveolae/lipid rafts resulting in an identification of about 70 proteins involved in many crucial endothelial cellular functions. The integrity of caveolae/lipid rafts together with its exclusive residents, the Src-family kinases (SFK) Fyn and c-Yes were required in cytokine induced inflammatory signaling. DHA22;6n3 treatment led to a significant displacement of Fyn and c-Yes from caveolae/lipids rafts. The mechanisms of selective displacement of SF Ks were further investigated. Analyses of Lipids from caveolae/lipid rafts indicated a significant incorporation of DHA22;6n3 into its phospholipids, causing an increase in the unsaturation index and cholesterol depletion from caveolae/lipid rafts. nanoESI-MS/MS further confirmed DHA22;6n3 incoporation into the major phospholipids known to localize in the caveolae/lipid rafts. In summary, our study is the first to show that diabetes induced decrease in DHAmm with a concomitant increase in the n6 to n3 PUFA ratio promotes basal and cytokine induced adhesion molecules expression and leukocyte adhesion in the retina. ACKNOWLEDGEMENTS I would like to greatly thank my committee members Dr. Walter J Esselman, Dr. Richard C Schwartz, Dr. Kathleen Anne Gallo, and Dr. Susan E Conrad for their valuable guidance and support for all these years. Especially I deeply appreciate my mentor Dr. Esselman, who no matter under what circumstances always provides me with the most generous and warmest support and encouragement throughout my graduate career. Also I would like to dedicate my special thanks to my co-mentors I have had during different stages of my graduate study: Dr. Kathleen Anne Gallo and Dr. Julia V Busik. The former guides me through the early stage of my research with her sharpest mind; the latter, with all her wisdom, hard work, enthusiasm and friendship, leads me through my strenuous research journey; teaches me all her special techniques hand by hand, and makes her great effort to educate me to be a good and happy scientist. A lot of thanks also go to Dr. Richard C Schwartz for being always there for me to discuss the philosophy of doing science and the upsets of my results; and also for his unconditional opening of his lab for me to apply any resources. And more, Dr. Donald B Jump in Department of Physiology, who contributes a lot of insightful suggestions, heated discussions, technical supports and resources, and well-maintained facilities for me to freely carry out my research; And Dr. Shiela Ferguson-Miller in Department of Biochemistry, Dr. Gavin Edmund Reid in Department of Chemistry and their wonderfiil student Xi Zhang for their great collaboration in lipids analyses by mass spectroscopy. iv I’m also very thankful for the friendship of all my “lab mates” in Dr. Jump Lab, Dr. Gallo lab and Dr. Olson lab during my graduate training, especially J inghua Xu, Yun Wang, Barbara Christian, Daniela Botolin in Dr. Jump lab; Diana, Zi Ye in Dr. Olson lab for their valuable discussions and co-operation and particularly their unlimited, unselfish “Love Project” to shine the spirit on my research and my life. At last, no words can express my appreciation to my parents, my dear brother and his family, for them to be always with me during this special and important stage of my life... TABLE OF CONTENTS Page List of Tables ....................................................................................... xii List of Figures ...................................................................................... xiii Key to Abbreviations ............................................................................... xvi I. Literature review ............................................................................................................ 1 1. Etiology of diabetic retinopathy ................................................................................. l 2. Inflammation and diabetic retinopathy ....................................................................... 4 3. Hyperglycemia and diabetic retinopathy .................................................................... 9 4. Dyslipidemia and diabetic retinopathy ..................................................................... 13 4.1. Diabetic dyslipidemia ....................................................................................... 13 4.2. Insulin and lipid metabolism ............................................................................. 14 4.3. Dyslipidemia and diabetic retinopathy ............................................................. 19 5. Role of PUFA in inflammation ................................................................................. 20 5.1. n6-PUFA and inflammation .............................................................................. 20 5.2. n3 PUFA and inflammation .............................................................................. 23 5.2.1. n3-PUFA and oxidized lipids ..................................................................... 23 5.2.2. n3-PUFA and PPARs ................................................................................. 27 5.2.3. n3 PUFA and Caveolae/lipid rafts ............................................................. 31 6. Retinal endothelial cell and fatty acid profile ........................................................... 37 7. Objective of Thesis ................................................................................................... 39 vi 8. References ................................................................................................................. 40 II. Dyslipidemia, but not hyperglycemia, induces inflammatory adhesion molecules in human retinal vascular endothelial cells ........................................................................... 57 1. Abstract ..................................................................................................................... 57 2. Introduction ............................................................................................................... 59 3. Materials and Methods ............................................................................................. 62 3.1 Reagents and supplies ......................................................................................... 62 3.2. Cell culture and fatty acid treatments ................................................................ 62 3.3. Electrophoresis and immunoblotting ................................................................. 63 3.4. Leukocyte adhesion assay .................................................................................. 64 3.5. Statistical Analysis ............................................................................................. 65 4. Results ....................................................................................................................... 66 4.1. Polyunsaturated n6 fatty acids induce inflammatory adhesion molecule expression in hRVE cells .......................................................................................... 66 4.2. Hyperglycemia does not affect CAM expression in hRVE cells ....................... 67 4.3. Induction of inflammatory adhesion molecules by cytokines and PMA in hRVE cells ........................................................................................................................... 67 4.4. HUVEC do not respond to fatty acid treatment ................................................. 67 4.5. Inhibition of fatty acid oxidation suppresses fatty acid-induced adhesion molecule expression in hRVE cells .......................................................................... 68 4.6. Leukocyte adhesion correlates with fatty acid induction of inflammatory CAMS ................................................................................................................................... 69 vii 5. Discussion ................................................................................................................. 77 6. References ................................................................................................................. 82 III. Anti-Inflammatory Effect of Docosahexaenoic Acid (DHA22,(,,,3) and Peroxisome Proliferator—activated Receptors (PPARs) on Cytokine Induced Adhesion Molecules Expression in Human Retinal Vascular Endothelial Cells ............................................... 87 1. Abstract .................................................................................................................... 87 2. Introduction ............................................................................................................... 89 3. Materials and Methods .............................................................................................. 93 3.1. Reagents ............................................................................................................. 93 3.2. Cell culture ......................................................................................................... 93 3.5. Real time RT-PCR ............................................................................................. 95 3.6. Transfection of hRVE using lipofectamine 2000 .............................................. 96 4. Results ....................................................................................................................... 97 4.1. TNFCL, IL-1 [3 and VEGF 165 induce adhesion molecules expression in hRVE cells ........................................................................................................................... 97 4.2. DHA22;5n3 inhibits TNFOI, IL-IB and VEGF.” induced CAM expression ...... 97 4.3. NFKB is an important transcription factor regulating adhesion molecules expression in hRVE .................................................................................................. 98 4.4. DHA22;6..3 pretreatment inhibits cytokine induced NFIcB binding to the VCAM-l promoter .................................................................................................... 99 4.5. DHA22;6,,3 pretreatment inhibits IIcBot phosphorylation and degradation, an immediate event upstream of NFKB nuclear translocation ..................................... 100 viii 4.6. Expression pattern of PPAR isoforms in hRVE by RT-PCR ........................ 100 4.7. PPARoI specific agonists partially inhibit cytokine induced CAMS expression ................................................................................................................................. 101 4.8. DHA22;6,,3 activates PPARoI in hRVE ............................................................ 101 5. Discussion ............................................................................................................... 116 6. References .............................................................................................................. 120 IV. Proteomic analyses of detergent resistant caveolae/lipid rafts from cultured human retinal vascular endothelial cells (hRVE) ....................................................................... 126 1. Abstract ................................................................................................................... 126 2. Introduction ............................................................................................................. 128 3. Materials and Methods ............................................................................................ 131 3.1. Reagents and antibodies ................................................................................... 131 3.2. Cell culture ....................................................................................................... 131 3.3. Electrophoresis and immunoblotting ............................................................... 131 3.4. Isolation of lipid rafis/caveolin-rich membrane domains ................................ 132 3.5 . Protein in-gel digestion .................................................................................... 133 3.6. Protein in-solution digestion ............................................................................ 133 3.7. LC/MS/MS ....................................................................................................... 134 4. Results ..................................................................................................................... 135 4.1. Isolation of caveolae/lipid rafts from hRVE .................................................... 135 4.2. Characterization of caveolae/lipid rafts components by in gel digestion and LC/MS/MS .............................................................................................................. 135 ix 4.3. Identification of caveolae/lipid rafts components by in solution digestion and LC/MS/MS .............................................................................................................. 136 4.4. Identification of less abundant proteins in hRVE by western blot .................. 137 5. Discussion ............................................................................................................... 149 6. References ............................................................................................................... 156 V. Inhibition of retinal endothelial cell inflammatory response: Modification of caveolae/lipid rafts by Docasahexanoic acid (DHA22;(,,,3) treatment .............................. 160 1. Abstract ................................................................................................................... 160 2. Introduction ............................................................................................................. 162 3. Materials and Methods ............................................................................................ 165 3.1. Reagents and antibodies .................................................................................. 165 3.2. Cell culture and fatty acid treatment ............................................................... 165 3.3. SDS-PAGE and western blot .......................................................................... 166 3.4. Subcellular fractionation ................................................................................. 166 3.5. Fatty Acid Metabolism ................................................................................... 168 3.6. Fatty Acid and Cholesterol Analysis .............................................................. 168 3.7. Mass spectrometry of phospholipids .............................................................. 169 4. Results ..................................................................................................................... 172 4.1. Caveolae/lipid rafts are involved in VEGF165 and TNFa induced CAM expression ............................................................................................................... 172 4.2. DHA22;6.,3 displaces Src family kinase from caveolae / lipid rafts in endothelial cells ......................................................................................................................... 1 73 4.3. DHA22;(,..3 treatment alters fatty acyl compositions of phospholipids residing in caveolae/lipid rafts .................................................................................................. 174 4.4. DHA22;6,.3 enrichment causes cholesterol depletion in caveolae/lipid rafts.... 177 5. Discussion ............................................................................................................... 190 6. References ............................................................................................................... 197 xi LIST OF TABLES Table Page Chapter I 1. Natural and synthetic peroxisome proliferator-activated receptor (PPAR) ligands ....................................................................................... 30 Chapter IV 1. Proteins identified by in-gel digestion and LC/MS/MS ............................. 141 2. Proteins identified by in-solution digestion and LC/MS/MS ....................... 147 Chapter V 1. Fatty acid composition of phospholipids from caveolae/lipid rafts and general plasma membranes of hRVE cells treated with control (BSA), lipid control (16:0) and n3-PUFA (22:6n3) ................................................................... 187 2. Identification of the most abundant signals in the ESI mass Spectra of total membrane lipid extracts of hRVE cells treated with BSA, palmitate 16:0 and DHA22;(,,.3, as given in Fig. 6 A and B .................................................. 188 xii LIST OF FIGURES Page Chapter I Fig. l. Canonical pathways of NFkB activation by inflammatory cytokines. ................... 8 Fig. 2. Potential mechanisms by which hyperglycemia activates four pathways of hyperglycemic damage. ............................................................................................ 12 Fig. 3. Classification of fatty acids. ................................................................................. 17 Fig. 4. Effect of insulin on unsaturated fatty acid synthesis. ........................................... 18 Qhagter 11 Fig. 1. Induction of endothelial cell adhesion molecules by free fatty acids in hRVE cells. ................................................................................................................................... 71 Fig. 2. Evaluation of cell adhesion molecule expression after treatment of hRVE cells with hyperglycemic conditions, cytokines and PMA. .............................................. 72 Fig. 3. Fatty acid treatment of HUVEC cells fails to induce endothelial cell adhesion molecule expression. ................................................................................................. 73 Fig. 4. Induction of cell adhesion molecule expression by fatty acids is inhibited by LOX, but not COX and MOX inhibitors. ........................................................................... 74 Fig. 5. Increased adhesion of leukocytes after induction of cell adhesion molecule expression by fatty acids. .......................................................................................... 76 xiii Chapter 111 Fig. 1. Induction of cell adhesion molecules after treatment of hRVE cells with cytokine TNFCI, IL-lB and VEGF165. .................................................................................. 104 Fig. 2. n3-PUFA pretreatment specifically downregulates the induction of VCAM-1 by proinflammatory cytokines. .................................................................................... 106 Fig. 3. Inflammatory cytokines activate NFIcB signaling to induce adhesion molecules expression in hRVE cells. ....................................................................................... 108 Fig. 4. DHA22z6n3 inhibits VEGF165 and IL-1 [3 induced NFIcB signaling ................. 109 Fig. 5. Inhibition of IL-1 [3 induced IIcBa phosphorylation and degradation by DHA22z6n3 pretreatment in hRVE. ....................................................................... 110 Fig. 6. Relative mRNA abundance and protein expression of PPAR isoforms in hRVE. ................................................................................................................................. 112 Fig. 7. PPARa agonists inhibit VEGF165 induced CAMS expression. ........................ 113 Fig. 8. Effect of PPAR ligands on TNFa induced VCAM-1 expression in hRVE cells. ................................................................................................................................. 114 Fig. 9. Activation of PPARoI by WY14,643 and exogenous free fatty acids in hRVE cells. ................................................................................................................................. 115 Chapter I V Fig. 1. Characterization of caveolae/lipid rafts in hRVE cells. ..................................... 138 Fig. 2. SyproB Blue staining of caveolae/lipid rafts proteins (1) and soluble (S) proteins isolated from hRVE cells after SDS-PAGE ............................................................ 139 Fig. 3. Localization of Src family kinases in hRVE. ..................................................... 140 xiv Chapter V Fig. 1. Methyl-B-cycloxydextrin (MCD) pretreatment disrupts TNFa induced NFKB signaling in hRVE cells. ......................................................................................... 178 Fig. 2. Src family kinases are involved in VEGF165, TNFCI induced VCAM-1 expression. .............................................................................................................. 179 Fig. 3. Specific displacement of Src family kinase F yn and c-yes from caveolae/lipid rafts by DHA treatment in hRVE ............................................................................ 181 Fig. 4. Incorporation of 14C-22z6n3 into different lipid complexes in hRVE cells. ..... 182 Fig. 5. DHA22:6n3 alters fatty acyl compositions of phospholipids in caveolae/lipid rafts and bulk membranes. .............................................................................................. 185 Fig. 6. nano-ESI-MS analyses of total plasma membranes phospholipids fiom hRVE treated with different fatty acids. ............................................................................ 187 Fig. 7. DHA22:6n3 enrichment causes cholesterol depletion in caveolae/lipid rafts. 189 XV Key to Abbreviations AGE BREC CAM CE COX DAG DHA DPA DR DRM eNOS EC EEA EPA ER ERK ESI FFA Flk-l GAPDH arachidonic acid (20:4n6) advanced glycation end product bovine retinal endothelial cells cell adhesion molecule cholesterol ester cycloxygenase diacylglycerol docosahexaenoic acid (22:6n3) docosapentaenoic acid (22:5n3) diabetic retinopathy detergent resistant membrane endothelial nitric oxide synthase endothelial cell early endosomal antigen eicosapentaenoic acid (20:5n3) endoplasmic reticulum extracellular signal-regulated kinase electrospray ionization free fatty acid Fms like kinase receptor glyceraldehyde-B-phosphate dehydrogenase high density lipoprotein xvi HETE hydroxyeicosatetraenoic acid HMG-CoA hydroxymethylglutaryl coenzyrne A HODE hydroxyoctadecadienoic acid HPETE hydroperoxyeicosatetraenoic acid hRVE human retinal endothelial cells HSL hormone-sensitive lipase IKBoI inhibitor of NFKB alpha IKK IKB kinase ICAM-l intercellular cell adhesion molecule -1 IL interleukin LA linoleic acid (18:2(0-6) LAT linker for activation of T cell LC-PUFA long-chain polyunsaturated fatty acid LDL low density lipoprotein LOX lipoxygenase LPL lipoprotein lipase LT leukotriene MMP matrix metalloprotease MCD methyl-B-cyclodextrin ME macular edema MS mass spectrometry MUFA monounsaturated fatty acid NDGA nordihydroguaiaretic acid xvii NEFA PAI PC PDR PE PG PI PKC PPAR PS PUFA PVR RAGE ROS RTK SFK SM SRA STZ TG TGF nonesterified fatty acid nuclear-factor kappa B plasminogen activator inhibitor phosphocholine proliferative diabetic retinopathy phosphatidylethanolamine prostaglandin phosphoinositide protein kinase C peroxisome proliferator-activated receptor phosphotidylserine polyunsaturated fatty acid proliferative vitreous retinopathy receptor for advanced glycation end product reactive oxygen species receptor for tyrosine kinase Src family kinase Sphingomyelin saturated fatty acid streptozotocin triglyceride transforming growth factor tissue factor xviii uPA SMC VCAM- 1 VEGF VLDL tumor necrosis factor urokinase-type plasminogen activator smooth muscle cell vascular cell adhesion molecule-1 vascular endothelial grth factor very low density lipoprotein xix I. Literature review 1. Etiology of diabetic retinopathy Diabetic retinopathy (DR) is a microvascular complication of diabetes and a leading cause of blindness in adults [1, 2]. It occurs when diabetes damages capillaries inside the retina, the light-sensitive tissue at the back of the eye. Diabetic retinopathy, in the way of proliferative diabetic retinopathy (PDR) and macular edema (MB), is the commonest cause of new cases of legal blindness in Europe and in North America in the age group 20 to 70-74 years[3]. Approximately 5.3 million out of 16 million people in the US living with diabetes have some form of DR. About 24,000 people are blinded each year by the disease. Both Type 1 and Type 2 diabetics are at risk of developing diabetic retinopathy. PDR may be present in half of those who have had Type 1 diabetes for 15 years with approximately 10% for those with Type 2 diabetes sustaining the same duration of disease[4]. The earliest stage of DR is diagnosed as mild nonproliferative retinopathy characterized by vascular basement membrane thickening, microaneurysms (out- pouchings of capillaries) with small areas of balloon-like swelling in the retinal capillaries; AS the disease progresses, selective loss of intramural pericyte attachment to the capillaries occurs which leads to acellular or nonfunctional capillaries causing the dot and blot hemorrhages (tiny hemorrhages in the retina itself) and exudates (retinal deposits occurring as a result of leaky vessels). These symptoms, often present without any visual compromise, are also called background DR and happen at any time with the onset of diabetes. As more capillaries are blocked, severe nonproliferative retinopathy happens. 1 The blockage of capillaries compromises the blood circulation, causing downstream ischemia as manifested by an increase in the Size and number of intraretinal hemorrhages. The down stream hypoxia thus induces synthesis of an endothelial-cell-specific angiogenic factor, vascular endothelial growth factor (VEGF). The increase of VEGF acts as a local hormone to induce more new blood vessel synthesis (neovascularization), a halhnark of proliferative retinopathy. PDR is the most advanced stage of the disease carrying the greatest risk of visual loss. The proliferating new vessels could develop along the retina and break into the surface of the clear, vitreous gel that fills the inside of the eye and eventually lead to serious retinal detachment. The newly synthesized blood vessels are abnormal and fragile, tending to leak blood into the center of the eye, blurring vision or even causing blindness. Also, fluid can leak into the center of the macula, the part of the eye where Sharp, straight-ahead vision occurs. The fluid makes the macula swell thus blurring vision. This is called macular edema. Macular edema can occur at any stage of diabetic retinopathy. About half of the people with proliferative retinopathy also have clinically Significant macular edema[5]. Despite the extensive research, only a few therapies are currently available to DR patients. Scatter laser photocoagulation has been recommended for treatment of advanced PDR. The risk of visual loss is reduced by more than 50% for patients with macular edema who undergo focal laser photocoagulation. It was estimated that timely detection and photocoagulation treatment could prevent 95% of severe vision loss in patients with diabetes[4]. More recently, the Diabetes Control and Complications Trial has demonstrated the efficacy and cost effectiveness of glycemic and blood pressure control in reducing the incidence and progression of DR[6]. Phase II and III clinical studies involving anti-vascular endothelial growth factor, protein kinase C (PKC) inhibitors and antioxidants for the management of DR and diabetic macular edema (DME) are underway (See reviews [5, 7]). The inhibition of these biochemical pathways holds the promise of intervention for DR at earlier non-Sight-threatening stages. Even though careful screening, good control of blood glucose, and laser photocoagulation can help mitigate the effects of DR, patients suffering blindness from diabetes are still arising, with approximately 5,800 new cases reported annually. The identification of risk factors and determinants for early onset of retinopathy becomes crucial for our understanding of disease mechanisms thus developing strategies to prevent the disease at the early stage. 2. Inflammation and diabetic retinopath y More than 40 years ago, aspirin was shown to decrease the severity of diabetic retinopathy in humans[8], which provides a link between inflammation and DR. Only recently, additional data suggested that early-stage diabetic retinopathy is a low-grade chronic inflammatory condition[9-1 1]. In experimental diabetes, the earliest event of leukocyte adhesion to the retinal vasculature results in early blood retinal barrier breakdown, capillary non-perfusion and endothelial cell injury and death. Support for this view is provided by the finding that a marked increase in leukocyte density and retinal vascular ICAM-1 and P-Selectin immunoreactivity was found in human eyes with diabetic retinopathy[12]. Lymphocyte activation, an increased serum L-selectin level and increased lymphocytes adhesion to the endothelium in DR patients were observed significantly higher than control normal and diabetic (with no DR) patients[13]. In a canine model aspirin prevented certain classic histopathological features of DR, including formation of acellular capillaries; retinal hemorrhage; and an indicator of cell degeneration, capillary sudanophilia[14]. In a rat model of diabetic retinopathy, nonsteroidal anti-inflammatory agents such as aspirin, meloxicam and etemacept prevent early diabetic retinopathy development by decreasing the endogenous level of retinal proinflammatory cytokine TNFa thus suppressing diabetic retinal ICAM-l expression, leukocyte adhesion and blood—retinal breakdown[10]. Furthermore, in CD18-/- and ICAM-l-/- mice a marked reduction of STZ diabetes-induced blood-retinal barrier breakdown, pericyte and endothelial cell loss and formation of acellular capillaries were observed [1 5]. Several inflammatory pathways are activated at the early stage of diabetic retinopathy. A pro-inflammatory cytokine TNFa is found in the extracellular matrix, endothelium and vessel walls of eyes of patients with proliferative diabetic retinopathy[16] and is elevated in the vitreous from human eyes with this complication[17-20]. Moreover, inhibition of TNFCI signaling with a TNFCI receptor/PC construct reduced leukocyte adhesion and suppressed blood-retinal barrier breakdown in STZ diabetic rats[10]. The VEGF family of growth factors and their receptors have been well studied for their function in regulating endothelial proliferation and migration, vascular permeability and tube formation (reviewed in [21]). They are widely involved in controlling pathological angiogenesis and increased vascular permeability in important diseases such as cancer[22]. VEGF also has been strongly implicated in the pathogenesis of both background and proliferative diabetic retinopathy[23-27]. Increased intraocular VEGF levels as well as VEGF receptor 1 and 2 were detected in rat and human diabetic retina[23-26, 28-32]. In addition to its well known mitogenic and angiogenic activity, VEGF was recently recognized as a proinflammatory cytokine[33, 34]. As such, VEGF induces ICAM-l expression on endothelial cells[33]. Specific inhibition of VEGF activity inhibited ICAM-l expression, leukocyte adhesion, blood-retinal barrier breakdown and neovascularization in STZ diabetic rats[33]. Another principal inflammatory cytokine IL-l B recently has also been related to the progress of diabetic retinopathy. The level of IL-1 B was increased by more than twofold in retina of rats with two months of diabetes[35]. lnj action of IL-lB into vitreous of rat eyes induced microvascular apoptosis and acellular capillaries[35]. High glucose could increase IL-lB secretion in bovine retinal endothelial cells (BREC), while application of ILRa Significantly decreases IL-lB induced EC injuries[36]. These data suggests a possible role of IL-lB and its receptors mediated Signaling pathways in inducing EC injury that contributes to the development of diabetic retinopathy. NFIcB is a family of transcription factors prerequisite for many inflammatory genes expression such as adhesion molecules, cycloxygenase 2 (COX2), tissue factors (TF), matrix metalloprotease (MMP) and inflammatory cytokines[37, 38]. It was first described as B-lymphocyte-specific nuclear protein, essential for transcription of irnmunoglobulin kappa (K) light chains. There are five NFKB subunits-(RelA (p65), RelB, c-rel, p50 and p52) forming homo- and heterodimers and characterized by the conserved ‘rel homology’ domain in mammalian cells. In resting cells, NFIcB is sequestered in the cytoplasm with members of the inhibitor of NFIcB (IIcB) family which consist of IchoI, IKBB, 1K8 and Bcl3[39]. Many different stimuli including inflammatory cytokines, LPS, oxidized LDL and microbial agents have the potential to activate the NFIcB pathway through different signaling components and cascades depending on the cell type and stimuli. However, all will lead to the phosphorylation and activation of [ICE kinases (IKK) complex, a central component to the NFIcB cascade. The activation of IKKS triggers the phosphorylation of IKB at N-terminal serines which facilitates its recognition by ubiquitin kinase complex thus degradation by 26s proteosome. This finally leads to the liberation of NFKB from the inactive complex and transport to the nucleus for DNA binding (reviewed in [40], [41]). Both IL-lB and TNFa, and VEGF165 could activate the canonical NFKB pathway as Fig. 1. However, depending on different cell types, different signaling components are probably utilized. Activation of NFIcB (p65 and p50) has been well documented in diabetes especially in the retinal vasculature of diabetic patients and animals[10, 42, 43]. NFKB was shown to be activated in the pericytes of rat and human retina affected by diabetic retinopathy[42]. Also, NFIcB was activated in the retinas of advanced glycation end- products (AGEs)-treated rats[44] and in bovine retinal endothelial cells treated with high glucose[43]. The activation of NFKB by inflammatory cytokines in human retinal endothelial cells has not been thoroughly characterized. While there is a developing consensus on the role of inflammation in retinopathy, individual molecular steps leading to retinal adhesion molecules expression and to diabetic retinopathy are not well resolved. Two major metabolic disorders that are likely to play role in diabetes-induced inflammation in retina are hyperglycemia and dyslipidemia. Considerable progress in understanding of hyperglycemia-induced disease has been made over the past decade, while the role of diabetic dyslipidemia in the development of microvascular complications has received much less attention and the link between diabetic metabolic disorders and retinopathy still eludes us. NFIcB-dependent genes -Adhesion molecules -Cytokines and chemokines -Mo‘trix metalloproteinoses ~Tissue factor Fig. 1. Canonical pathways of NFKB activation by inflammatory cytokines. 3. Hyperglycemia and diabetic retinopath y Both Type 1 and Type 2 patients can develop DR. Proliferative DR typically develops with Type 1 diabetes, whereas nonproliferative retinopathy with maculae edema is more common in Type 2 diabetes. However, the microvascular alterations in both conditions have the same pathophysiological basis in that the occurrence and progression of retinopathy has been largely correlated with the degree of hyperglycemia. The molecular mechanisms of hyperglycemia induced retinal endothelial dysfunction have attracted a lot of attention. Several pathways have been implicated, including increased flux of glucose through the polyl pathway; increased advanced glycation end product formation and receptor activation[45]; activation of PKC isoforms[46]; and increased hexoamine pathway flux[47] (Fig. 2). Hyperglycemia induces increases in polyl pathway flux and acts through aldose reductase that converts glucose to sorbitol. Sorbitol then is further oxidized to fi'uctose by sorbitol dehydrogenase. This process can either produce sorbitol as an osmotic stress activator or change the cellular NADH/NAD+ or NADPH level that could potentially affect the redox status inside the cell and thus affect cellular functions. However, sorbitol levels in diabetic vessels are far too low to cause osmotic stress. Studies using aldose reductase inhibitors failed to prevent retinopathy and the thickening of the basal membrane of the retina in dogs in vivo[48, 49], although it prevented diabetic nephropathy. The nonenzymatic glycation and oxidation of proteins produce intracellular and extracellular AGES (Fig. 2). Other than causing abnormal protein function, cell-matrix or 9 matrix-matrix interactions, AGES could affect endothelial cells function mainly through binding to one of specific AGE receptors (RAGE) to produce reactive oxygen Species, causing the activation of NFKB[50] and thus pathological changes in gene expression. A wide range of evidence has implicated AGES, mainly through RAGE, in priming proinflammatory mechanisms in endothelial cells by inducing proinflammatory molecules expression such as adhesion molecules (VCAM-l, ICAM-l etc) and tissue factor expression[51—53]. Increased amounts of AGES are found in diabetic retina vessels concomitant with the higher level of RAGE[45, 54]. Furthermore, polymorphisms of RAGE genes have been associated with diabetic retinopathy[55]. Hyperglycemia also induces the de novo synthesis of lipid second messenger Diacylglycerol (DAG) which activates PKC isoforms[56]. Activation of PKC has been associated with many vascular abnormalities in retina, renal and cardiovascular diseases[56]. Among PKC isoforms, PKCB and 5 are preferentially activated in the vasculatures of diabetic animals, such as in the retina and glomeruli[56, 57]. Activation of PKC has a number of pathogenic consequences by affecting expression of endothelial nitric oxide synthase (eNOS)[58], VEGF, transforming growth factor (TGF-B)[59] and plasminogen activator inhibitor-1 (PAI-l)[60], and by activating NF-KB[61] and membrane associated NAD(P)H oxidases. Shunting of excess intracellular glucose into the hexosarnine pathway might also cause several manifestations of diabetic complications. The increase in the hexoamine pathway flux causes the modification of important transcription activators such as Spl to be modified by n-acetylglucosamine (GlcNAc), a product of hexoarnine pathway. This 10 modification of Spl could increase the transcription of tumor growth factor TGF-ot, TGF- B1 and PAI-1[62], resulting in many changes in both gene expression and protein function, which together contribute to the pathogenesis of diabetic retinopathy. Other than Spl, O-glycosylation of NFIcB components was also reported to regulate NFKB binding activities thus influence the expression of NFIcB dependent genes such as VCAM-1[63]. Blocking the hexoamine pathway could prevent experimental DR[47]. It has been suggested that hyperglycemia activates all these mechanisms by a single underlying process: overproduction of superoxide by the mitochondrial electron transport chain with subsequent inhibition of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) activity (reviewed in[64]). This leads to the diversion of upstream metabolites from glycolysis pathways into pathways of glucose overutilization such as the increased flux of dihydroxyacetone phosphate (DHAP) to DAG, an activator of PKC; and of triose phosphates to methylglyoxal, the main intracellular AGE precursor. Also, increased flux of fructose-6-phosphate to UDP-N-acetylglucosamine increases modification of proteins by 0-linked N-acetylglucosamine (GlcNAc) and increased glucose flux through the polyol pathway consumes NADPH and depletes glutathione (GSH) (Fig. 2). The central role of reactive oxygen species (ROS) in hyperglycemia mediated pathological effects are confirmed by the evidence that normalization of mitochondrial ROS by an inhibitor of electron transporter chain complex 2, by an uncoupler of oxidative phosphorylation, by uncoupling protein 1 and by manganese superoxide dismutase[65] could block the above pathways of hyperglycemic damage in viva. However, this hypothesis still needs to be firrther tested. 11 1' III NA PH NADP” + DH l Glurose Ll Sorbitollqm l Fructose Polyl pathway 1Glucose—6-P GFAT l l — - u—p lF ru ct o 6 P W GlucosamIne-6-P UDP—GlcNAC Gin Glu Hexoamine pathway NAD 0* if / T DHAP Ta-glycerol-lL’] DAG —-] PKC [Glyceraldehyde-3-P PrOtein kinase C Pathway A TMethylglyoxal —> ] AGES AGE Qathwgy Fig. 2. Potential mechanisms by which hyperglycemia activates four pathways of hyperglycemic damage. 12 4. Dyslipidemia and diabetic retinopath y Although a multitude of pathogenic mechanisms have been proposed, the underlying dysfunctional biochemical and molecular pathways that lead to initiation and progression of DR remain largely unresolved. Hyperglycemia is at a very late stage in the sequence of events leading flom insulin resistance to flank diabetes; whereas lipoprotein abnormalities are manifested during the largely asymptomatic diabetic prodrome and may contribute substantially to the increased risk of macrovascular and microvascular diseases. The potential role of dyslipidemia in the development of microvascular complications has received much less attention compared to hyperglycemia and will be the focus of this dissertation. 4.1. Dlabetic dyslipidemia Dyslipidemia is a major metabolic syndrome prevalent in both Type 1 and Type 2 diabetes. An imbalance in the coordinated complex regulation of fatty acid uptake, metabolism, release by adipocytes, and clearance flom circulation causes changes in serum triglycerides, lipoproteins and flee fatty acids profiles. Indeed, Type 2 diabetes is flequently associated with elevation of blood levels of LDL-C, triglycerides and flee fatty acids along with decrease in HDL cholesterol[66-72]. In Type 1 diabetes the overall cholesterol, triglycerides and nonesterified fatty acid levels do not Significantly differ from the control values[73-75]; however, there is a substantial change in fatty acid profile of these pools. Changes in serum lipids and lipoproteins have been observed in Type 1 diabetes mainly with an increased level of linoleic acid (LA) and alpha linoleic acid (or- LA)[73, 76]. The total n3-polyunsaturated fatty acids (PUFA) especially 13 docosahexaenoic acid (DHA22;6,.3) is decreased in both plasma and retina of diabetic children and human diabetic eyes[73, 77]. 4.2. Insulin and lipid metabolism Fatty acids are structurally classified according to the number of carbons, double bonds, and proximity of the first double bond to the methyl terminal of the fatty acyl chain. For example, fatty acids of n6 family contain a double bond at the sixth carbon flom the methyl end while n3 family contains a double bond at the third carbon. The nomenclature and the chemical structures of the major classes of fatty acids such as saturated (SRA), mono-unsaturated (MUF A) and PUFA are indicated as Fig. 3. AS such, DHA is represented as 22:6n3, indicating carbon chain length of 22 with 6 double bonds and the first unsaturated bond is inserted at carbon 3. To understand the effects of diabetes on the plasma and tissue fatty acid compositions, two metabolic routes have to be considered: de novo lipogenesis and polyunsaturated fatty acids remodeling pathways, or the Sprecher pathway[78] (Fig. 4). Saturated (SRA), mono- (MUFA) and PUFA are synthesized flom dietary precursors (glucose, 16:0, 18:1n9, 18:2n6, 18:3n3 & n6, 20:5n3) through a series of desaturation (AS-desaturase [A5D], A6-desaturase [A6D] or A9-desaturase [A9D]) and elongation (Elovl-Z, -5 & -6) reactions. The A6 and A5 desaturations are considered to represent the rate-limiting steps in fatty acid metabolism, and insulin is the most potent activator of the desaturase enzymes, as well as the induction of A5, A6, and A9 desaturases[67, 69, 74-76, 79-83]. Thus, subnorrnal availability of insulin in the liver will result in reduced activity of the desaturase enzymes and, consequently, will lead to accumulation of the substrates l4 and depletion of the products. The overall effect of such metabolic perturbations would lead to a shift in fatty acid profile towards higher saturation index and Shorter fatty acid chains. Indeed, tissue lipid profiles of people with diabetes are characterized by lower than normal concentrations of long chain PUFAs (LCPUFA) and higher than normal concentrations of monounsaturated and saturated fatty acids. Insulin inhibits hormone-sensitive lipase (HSL) and activates lipoprotein lipase (LPL) [81, 84]. HSL is an intracellular neutral lipase capable of hydrolyzing triacyl, diacyl and monoacyl glycerols and cholesteryl esters as well as other lipid and water soluble substrates in many tissues[85]. Responsive to many hormones such as catecholarnines, ACTH and glucagon, HSL is responsible for release of flee fatty acids flom adipose tissue thus providing the major source of energy for most tissues. Altered expression and activity of HSL in different cell types may be associated with obesity and Type 2 diabetes [85, 86]. Lipoprotein lipase (LPL) is synthesized in adipose and muscle and transported to the surface of endothelial cells where it hydrolyzes the core of triglyceride-rich lipoproteins (chylomicrons and very low density lipoprotein) into flee fatty acids and monoacylglycerol, facilitating the removal of triglyceride-rich lipoproteins flom the bloodstream. In the liver, activation of LPL by insulin stimulates conversion of fatty acids to triglycerides, followed by secretion as very low-density lipoprotein (V LDL)[87]. Thus, insulin resistance in Type 2 diabetes and low portal insulin levels in Type 1 diabetes resulting in an increased activity of HSL and a decreased activity of LPL, are expected to have a profound effect on plasma lipid levels, lipid profile and fatty acid composition. Indeed, diabetes is also characterized by increased lipolysis and altered lipogenesis that leads to increased concentration of nonesterified fatty acids (NEFA). 15 Patients both with Type 1 or Type 2 manifest decreased adipose LPL activity, and this is accompanied by a higher rate of HDL-cholesterol catabolism thus an increase in plasma triglycerides [79, 88]. 16 Palmitic Acid (16:0) HOOC - Mono Unsaturated Fatty Aciasawvm /\/\/\/\.=/\/\/\/\CH3 Olcic Acid (18:1.n‘)) HOOC ° Poly Unsaturated Fatty _ coo" Acids (PUFA) Linnlcic Acid (1822mm _ CH3 (n6-PUFA) Docosahexaenoic Acid _ _ (DHA, 22:5, n3) °°°" (n3-PUFA) _ _ _ 3 Fig. 3. Classification of fatty acids. Classification iS based on the carbon lengths, number of double bonds and the Site of the first double bond flom the methyl end. Fatty acids can be classified into saturated, monounsaturated and polyunsaturated (n6 and n3) fatty acids. 17 De Novo Lipogenesis(DNL) & 18:] Synthesis DNL MUFA “WI PM U1 18::2—>202 Elovl6 M £10le “* ' PUFA Sprecher Pathway Plant f 9’ R N6: ->l8: 39203»- ->22 :4->24. -:4—>24 s—s 22: 5 :Mi ElovlS £10le £10le A’s/L pp-Ox N118: 3-> l8: 4—> 20.4 12245» 24:6—>22:6 4 J.’ Plant Fish Fish Fig. 4. Effect of insulin on unsaturated fatty acid synthesis. Dietary 16:0, 18:1n9, 18:2n6 and 18:3n3 are converted to long chain unsaturated fatty acids in vivo by a series of desaturation (AS-desaturase [A5D], A6-desaturase [A6D] or A9-desaturase [A9D]) and elongation (Elovl-2, -5 & -6) reactions. Fatty acids that accumulate in most animal and human tissues are in solid boxes. Dietary 18:2n6 and 18:3n3 are obtained flom plants; 20:5n3 and 22:6n3 are rich in fish meals. There is no interconversion between n3, n6 and n9 fatty acids in the animals. Insulin controls A5-, A6-, and A9-desaturases. The activity and expression of these desaturases is low in Type 1 diabetes, as indicated by X8. 18 4.3. Dyslipidemia and diabetic retinopathy Clinical data Show that dyslipidemia could be a critical factor in the development of diabetic retinopathy. Retinal hard exudates are the component of diabetic retinopathy most likely to be related to plasma lipoproteins, because the exudates are lipid rich. The Early Treatment Diabetic Retinopathy Study (ETDRS) has demonstrated that higher levels of total triglycerides, total cholesterol, and LDL cholesterol are strongly associated with an increased risk of development of hard exudates in the macula and visual loss[89-91]. In a recent clinical treatment of diabetic patients with simvastatin, a hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase inhibitor, a significant retardation of the progression of diabetic retinopathy was documented[92, 93]. Another clinical trial with atorvastatin in patients with Type 2 diabetes with dyslipidemia demonstrated a reduction in the severity of hard exudates and subfoveal lipid migration in clinically significant macular edema, suggesting that lipid- lowering therapy could be an important adjunct in the management of clinically significant macular edema[94, 95]. The association of retinopathy with cardiovascular disease and elevated plasma LDL cholesterol[96] was also suggested flom another population-based study. The newly conducted Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study (DCCT/EDIC) cohort study revealed new associations between retinopathy status and the subclasses of lipoproteins. A strong inverse association between the severity of DR and average VLDL particle Sizes was defined with gender differences in Type 1 diabetes[97]. l9 5. Role of PUFA in inflammation The irnmunomodulatory properties of lipids were first reported more than half century ago. Cells of the immune system can use the lipids as both intracellular and extracellular signal mediators. It has been reported that lipid mediators can have proinflammatory properties or anti-inflammatory properties depending on the cell target or lipid derivatives[98] (Fig. 5A, 5B). The fatty acid composition of inflammatory and immune cells is sensitive to change according to the fatty acid composition of the diet. In particular, the proportion of different types of PUFA in these cells is readily changed, and this provides a link between dietary PUFA intake, inflammation, and immunity. 5.1. n6-PUFA and inflammation Unsaturated fatty acids are substrates for oxygenases such as Cyclooxygenases (COX)[99], Lipoxygenases (LOX or LO)[100], and cytochrome P450 Monoxygenases (MOX)[101, 102], and also for nonenzymatic oxidation. The COX pathways including COX-1 and COX-2 enzymes catalyze the first step in the biosynthesis of prostaglandins (PGS) by converting arachidonic acid (AA) to PGH2. PGH2 is further converted into 2- series of PGs and eicosanoids such as PGE2, PGD2, PGFZa, PG12 (prostacyclin) and thromboxanes (TXBz)[98, 103]. COX-2 and its potent inflammatory products. like PGE2 and thromboxanes have been implicated in the pathogenesis of several inflammatory diseases including diabetic vascular diseases such as atherosclerosis ([99] and reviewed in [104]). Recent research indicates that high glucose as well as ligands for RAGE could upregulate the expression and activity of COX-2 primarily through NFIcB activation under diabetic conditions[105, 106], 20 LOXS are a diverse family of nonheme ferroproteins that catalyze the hydroperoxidation of polyunsaturated fatty acids both region- and stereospecifically. Thus far, six LOXS have been identified in humans: platelet-type 12-LOX, 12(R)- LOX[107], lS-LOX-l, 15-LOX-2, e-LOX-3[108], and 5-LOX[100, 109]. Lipoxygenases products, such as the hydroperoxyeicosatetraenoic acids (HPETE), hydroxyeicosatetraenoic acids (HETE) and their metabolites the leukotrienes, play roles in many of the steps involved in diabetes, inflammation, arthrosclerosis, especially in modulating cell-cell interactions[98, 110]. Under hyperglycemia conditions, endothelial cells have been Shown to increase the monocyte-endothelial interactions through 12-LOX pathway by generating 12 (S)-HETE flom arachodonic acid[110]. Products of linoleic acid metabolism by leukocyte-type 12-LOX may also play roles in mediating inflammatory processes. 9- and 13-hydroxy-linoleic acid possess chemotactic activity for bovine and human polymorphonuclear leukocytes[1 l 1]. Significant quantities of 12- HETE were produced in db/db mice in vivo which corresponded to the increased monocyte-endothelium adhesion[112]. Disruption of 12/ 15 LOX mRNA in these animals by a catalytic ribozyrne blocked the monocyte adhesion[113]. Furthermore, 12- LOX activity and expression were highly increased in HLHG (hyperlipidemia- hyperglycemic) pigs in a diabetic pig model[92]. Other LOXS and their bioactive lipids also have profound proinflammatory effects. Emerging data implicate 5-LOX and its products, especially the 4 series leukotriences LTB4 and LTE4, as major players in the cardiovascular diseases due to their proinflammatory activities[98, 110]. The role of LOXS and their bioactive lipid mediators in the pathogenesis of diabetic retinopathy is not well characterized. The only evidence is flom the epiretinal membrane tissue of 21 patients with PVR and PDR where significant amount of 15-HETE (IS-LOX product of AA20:4.,6) were detected[114]. The mechanisms of the biological activities ascribed to AA20;4..6 and LA13;2,.6 based COX and LOX metabolites including prostaglandins, individual HETES and HPETES (flom AA20;4,.6), HODES and HPODES (flom LA18;2n6) etc are not well addressed. Evidence suggests these products can act as discrete signaling molecules through G-protein mediated Signaling pathways. lZ-HETE has been Shown to activate extracellular-Signal-regulated kinase (ERK)[115], PI3K[116], PKC and Src kinases[1l7]. The activation of PKC is linked to 15(s)-HPETE and 12(s)-HETE induced surface expression of cell adhesion molecules (CAM)s in human umbilical vein endothelial cells (HUVEC) associated with an increased binding activity of the transcription factor, NF- KB[118]. Recently, 12(s)-HETE alone has been shown to be able to direct the translocation and activation of PKCS in lens epithelial cells[119]. The regulation of protein kinase C by mono-HETES has also been suggested through a specific guanine nucleotide-binding protein linked receptor mediated hydrolysis of inositol phospholipids[120], a receptor mediated signaling pathway similar to other eicosanoid receptors identified for prostaglandin E, prostacyclin, thromboxane A2 and leukotriene D4. Overall, the oxidized lipids of the lipoxygenase and cycloxygenase pathways of M201“ and LAM“ metabolism have potent growth, vasoactive, chemotactic, oxidative and proinflammatory properties in vascular smooth muscle cells, monocytes and endothelial cells (Fig. 5A). Cellular and animal models have Shown that these enzymes 22 were induced in diabetic complications suggesting the role of these pathways and their lipid products in the pathogenesis of diabetic vascular diseases[121]. However, the biological effect of fatty acid oxidation also depends critically on particular fatty acid precursors[122]. 5.2. n3 PUFA and inflammation The anti-inflammatory effects of long-chain omega-3 polyunsaturated fatty acids (n-3 LCPUFAS) flom fish oil were amongst the earliest identified biological actions of these fatty acids. Animal studies have shown that dietary fish oil results in altered lymphocyte function and suppressed production of proinflammatory cytokines by macrophages[123]. Supplementation of the diet of healthy human volunteers with fish oil-derived n3- PUFAS results in decreased monocyte and neutrophil chemotaxis and decreased production of proinflammatory cytokines[l24]. Fish oil feeding has been Shown to ameliorate the symptoms in some animal models of autoimmune disease (reviewed in [125]. Clinical studies have reported that fish oil supplementation has beneficial effects in treating asthma[126], atherosclerosis, cancer, and cardiovascular disorders (for recent reviews, see [127, 128]), supporting the idea that the n3-PUFAS in fish oil is anti- inflarnmatory and immunomodulatory. 5.2.1. n3-PUFA and oxidized lipids The first established link between the n3-PUFA and anti—inflammatory role is through inhibition of n6cPUFA derived inflammatory eicosanoids. This could be 23 achieved through the incorporation of n3-PUFA into membrane phospholipids of immune cells which leads to the displacement of the inflammatory n6-PUFA substrates (AA20,4,,6) available for catalysis. Also n3-PUFA could inhibit the activation of phospholipase A2 (PLA2), the enzyme necessary to release the substrate for metabolism[124, 129]. Generally, eicosapentaenoic acid (EPAzo;5n3) and DHA22,6,,3 are poor substrates for oxygenases. EPA20;5,.3, or DHA2236n3 (by reconversion to EPA20;5n3), can to some extent be metabolized by COX and 5-LOX, but the derived 3-Series proglandins and 5-series leukatrienes have an order of magnitude lower biological activity compared to n6-PUFA and mainly play an anti-inflammatory function (Fig. 5B) (reviewed in[122]). Recent studies have provided a novel mechanism(s) for the therapeutic benefits of n-3 dietary supplementation important in inflammation, neoplasia, and vascular diseases. A novel group of mediators -- termed E-series resolvins -- formed flom EPA20;5,,3 by COX-2 in the presence of aspirin and DHA22:6n3-derived mediators termed D-series resolvins, docosatrienes and neuroprotectins, also produced by COX-2, have been identified flom resolving inflammatory exudates[l30] and DHA22:6n3 enriched tissues such as brain and retina[ 131] (Fig. 5B). Aspirin acetylates COX-2, enabling the synthesis of those bioactive 17R-hydroxy-containing di- and tri-hydroxy-docosanoids/epoxyanoids termed resolvins flom DHA22,6,,3 and EPA20,5n3 via epoxide-containing intermediates. They are proved to be potent anti-inflammatory mediators with pico- to nanomolar range efficacy to reduce both leukocytes infiltration in vivo and block cytokine production flom glial cells. These results indicate that DHA22;6n3 or EPA20;5.,3 are the precursors to potent protective mediators generated via enzymatic oxygenations to novel docosatrienes and 24 17S series resolvins that each regulate events of interest in inflammation and resolution[l32]. 25 A- * AA20:4ne l LA18z2n6 C | . C - 50 Irc‘gxi’iyifgiizejl IL'F’Wenasel I (Epoygygpgnase) I / l \ \ ’ . . Epoxides 2 senes Prostaglandins _ _ _ _ F3602, PGE2. PGI2, PGF2a+ 5 LO’ 8 [‘0’ 12 LO’ 15 L0 HETES BEE- Vasodilation 1 Ion transport Vasodilation, Vasoconstnctron. “WWW" Inflammation ' C9" 9M Anti-inflammato effects Platelet aggregation InflammatIon Matrix deposition TY Migration Inflammation Adhesion B. DHA 22:6n3 |"— ------- . EPA 20:5n3 | \\\ / \ ‘4 cycloxygenase [Lipoxygenases] cox-2 + * ’ I , I l D 3-sen'es Prostaglandins _ . . [ (PGES. TXA3) 5 series Leukatrrenesl E-series resolvins (LTBS) \ x x s s \ Anti-inflammatory effects l D-series resolvins Fig. 5. Syntheses and functions of eicosanoids from PUFAS of n6 (A) and n3 (B). Lipid mediators can have proinflammatory properties or anti-inflmmatory properties depending on the cell target or lipid derivatives. 26 5.2.2. n3-PUFA and PPARS Several lipid mediators, mainly PUF A or their derivatives can function as ligands for Peroxisome Proliferator-Activated Receptor (PPAR) family[133, 134]. PPARS are members of nuclear hormone receptor super family transducing environmental, nutritional and inflammatory signals into cell at the level of gene transcription (reviewed in [134]). Three PPAR isoforms - PPARoI, PPARB/8 and PPARy have been identified with a unique quantitative pattern of tissue distribution and differential functions in regulating glucose and lipid metabolism[l34]. The activation of PPARS by various types of fatty acid and their derivatives has some degree of isoform specificity. Endogenous NEFA such as a-LA13;3,,3, ‘Y-LA133n5, M20406 and EPA20,5.13 are weak activators of PPARy[135]. PPARoI can be activated by similar PUFA but also some medium-chain saturated and mono-unsaturated fatty acids (e.g. palmitic (Cl6:0) and oleic (C18zl) acids). DHA22;6n3 has been shown to be a potent PPARoI activator[136]. 13-HODE, 9-HODE, generated flom LA13;2n6 flom endogenous or component of oxidized-LDL, via 12/ 15 lipoxygenases, are natural ligands for both PPARa and PPARy[137], while AAZMM, metabolites flom 5 and 8 lipoxygenases such as LTB4 and 8(s)HETE can selectively activate PPARa. The cycloxygenase metabolites of AAzoano, mainly prostaglandins such as 15dPG12 have great potential in activating PPARy as well as the 15-lipoxygenase metabolitelS-HETE[137, 138]. Various types of eicosanoids, including prostaglandin Al (PGAl), prostaglandin D2 (PGD2) and possible the natural prostacyclins might be the endogenous agonists for PPARB/8[139] (Fig. 6). 27 The composition of the intracellular non-esterified fatty acids (NEFA) pool and their metabolites is an important determinant in the control of PPAR activity. Several pharmacological compounds have been synthesized to differentially activate PPAR isoforms, as shown in Fig. 6. Fibrates, a class of drugs in the treatment of dyslipidemia, and WY14,643, are synthetic ligands for PPARoI, whereas the antidiabetic glitazoncs are high-affinity ligands for PPARy. Synthetic ligands for PPARB/S have recently also been identified, including the prostacyclin analog carbaprostacyclin and GW501516. Recent evidence has indicated an important role of PPARS in the control of various types of inflammatory response (reviewed in [140]). The first role of PPARS in inflammation came with the evidence that PPARa-null mice demonstrated a prolonged response to inflammation stimuli[l4l]. Later on, specific agonists of PPARoI such as fenofibrate or WY14,643 have been demonstrated to be able to reduce cytokine induced TF expression in human monocytes and macrophages[l42] and IL6 production in human aortic smooth muscle cells[143]; thrombin-induced endothelin-1(ET-1) production[l44] and VCAM-1 expression in endothelial cells[145-l47] etc. Not only PPARa, but also PPARy ligands have been reported to suppress the LPS induced T cell active CXC chemokines production in human EC[147]. Modulation of vascular inflammation in HUVEC and in vivo[l48, 149] by PPARy activators were documented with observation of suppressing proinflammatory adhesion molecules expression. However, the anti- inflarnmatory role of PPARy is still under debate Since the effects are most pronounced with the least selective and active PPARy agonist, lSd-PGJ2[150]. The role of PPARB/8 28 in he Cl Frel; of PPARS exprsss dorm tit. .\ TCSfltt‘. that Ziplti Mt“ in the control of inflammatory responses has not yet been fully investigated due to the lack of isoforrn-specific agonists. Recently treatment of endothelial cells with new PPARB/8 agonist GW510156 inhibits the stimulus induced upregulation of VCAM-1 expression[lSl] and a crucial role of PPARB/B in keratinocyte inflammation was also documented[152]. The involvement of PPARS in the control of inflammation and inflammatory gene expression is mainly through the transrepression. Agonist targeted PPARcI can effectively antagonize the NFIcB and AP-l signaling pathways by physically interacting with NFIcB (p65) through its Rel homology domain and with the amino-terminal c-Jun respectively, thereby resulting in a functional cross-inhibition of both transcriptional activators[144, 153]. Also, ligand-activated PPARoI was Shown to upregulate the expression of mRNA and protein of the Inhibitor of K3 (IKBOI), which prevents NFKB translocation into the nucleus[154]. 29 hill: 1. ligands. Table 1. Natural and synthetic peroxisome proliferator—activated receptor (PPAR) ligands. PPARa PPAR/9’6 PPARy NEFA (SRA, MUFA and PUFA eg. LA, AA, EPA, NEFA ( PUFAS) NEFA ( PUFAS) DHA) Naturally 9—HODE Ei°°san°ids 15-HETE occurring FA- (PGAl. PGDZ) derived 13-HODE prostacyclin 15-dPGJ2 molecules BS-HETE 9-l-lODE LTB4 13-HODE Fibrates Glitazones Pharmacological (Fenoflbrates, Cargawfggtgfclin (Rosiglitazone, compounds benzofibrates) L g 6504 1 y Pioglitazone. WY14,643 trioglitazone) 30 5.2.3. 5.23.1. dare-2: fanatic z glycol Slim rim menb This ll I ' no tII Cate 5.2.3. n3 PUFA and Caveolae/lipid rafts 5.2.3.1. Caveolae/lipid rafts It has long been recognized that incorporation of n3-PUFA into membrane lipids decreases the microviscosity of membranes in general that may affect the mobility and function of membrane proteins. One specialized membrane microdomain whose structure and function could possibly be affected by n3-PUFA is called lipid rafts. Lipid rafts are enriched with cholesterol and sphingolipids, such as Sphingomyelin and glycolipids. They also contain polar lipids such as phospholipids that mainly consist of saturated fatty acyl residues. This makes them spontaneously aggregate to form liquid- ordered membrane regions facilitating their isolation as nonionic detergent—resistant membrane domains (DRM). Caveolae were first identified as 50-500nm flask-shaped invaginations in the plasma membrane nearly fifty years ago[155]. They share some similarities with lipid rafts as caveolae are also DRMS enriched with cholesterol and sphingolipids. Unlike lipid rafts, caveolae has a distinct invaginated form that can be easily detected by electron microscope. This is due to the presence of the scaffolding/regulatory protein caveolin. Caveolae are most numerous in well differentiated cell types such as adipocytes, myocytes, and fibroblasts especially endothelial cells[156]; while some cell types such as lymphocytes and neurons only contain lipid rafts devoid of caveolin. Other structures of caveolae independent of plasma membrane are also present including detached plasmalemmal and tubular-vesicular[156]. The density gradient centrifugation following Triton X-100 extraction at 4°C, the most flequently used method, does not result in pure 31 :2;le ratio will tin the "w MM. acyl use mat caveolae, but also coisolation of lipid rafts flom all cellular membranes. Several other methods have been newly devised to study caveolae and lipid rafts separately, including irnmuno-isolation[157, 158], double-label immunoelectroscopy and photonic force microscopy etc; each with its own advantages and disadvantages (Reviewed in [159]). Caveolae/lipid rafts consist of dynamic assemblies of cholesterol and sphingolipids in the exoplasmic leaflet of the bilayer. The presence of saturated hydrocarbon chains allows for cholesterol to be tightly intercalated, similar to the biophysical liquid-ordered state in model membranes. The characterization of the inner leaflet lipid composition is still incomplete, but they are probably rich in phospholipids such as phosphatidylethanolamine (PE) and phosphatidylcholine (PC) with saturated fatty acyl chains and cholesterol[160]. The membrane surrounding the caveolae/lipid raft is fluid since it contains more unsaturated phospholipids[160]. The importance of cholesterol to the structure and function of caveolae/lipid rafts was demonstrated by the use of sterol binding agent such as filipin, nystatin or cholesterol depletion agent such as methyl-B-cyclodextrin (MCD) (reviewed in [159] ). A major property of caveolae/lipid rafts is the differential inclusion and exclusion of proteins. Caveolae/lipid rafts are highly enriched with GPI-anchored proteins and doubly acylated proteins such as Src—farnily kinases (SFK) or the alpha subunits of heterotrimeric G proteins, etc. GPI anchored proteins have affinity with the exoplasmic leaflet of the lipid rafis[157, 161]. The double acylation of some intracellular proteins happens with the irreversible cotranslational modification of myristoylations at the N- terminal together with the s-acylation on cysteines usually with pahnitoylation. The high 32 paling with hi Some mesh molt {If-115C: crept III-din itIolI the ca than in In dim packing order of saturated myristol and pahnitoyl moieties facilitates the interactions with high-ordered cytoplasmic leaflets of lipid rafts or caveolae membrane subdomains. Some transmembrane proteins are targeted to caveolae/lipid rafts by yet unclear mechanisms with possible involvement of amino acids in the transmembrane domains near the exoplasmic leaflet[162]. 5.2.3.2. Functions of caveolae/lipid rafts in endothelial cells The function of caveolae and lipid rafts in endothelial cells was first suggested to involve transmembrane transport, endocytosis and exocytosis. In endothelial cells, the transcytosis of albumin, insulin and native or modified LDL are reported through receptors localized in caveolae, since this specialized process could be perturbed by sterol binding agent such as filipin or cholesterol depletion such as MCD[163, 164]. The involvement of caveolae in transcytosis suggests a role in regulating vascular permeability. The selective permeability of endothelial cells is especially important in the case of infectious diseases, atherosclerosis, and blood-brain and blood-retinal barriers functioning. Indeed, in the vascular systems of caveolin-l deficient mice, abnormalities in permeability and contractile functions were observed[165, 166]; and endothelial cells derived from mice lacking caveolae membranes have defects in transcytosis[l67]. Caveolae-mediated endocytosis constitutes an alternative endocytic pathway to clathrin-coated pits. The endothelial caveolae have the molecular transport machinery for vesicle budding, docking and fusion including VAMP, NSF, SNAP, annexins and GTPases[l68]. Caveolae carry select cargo, distinct flom clathrin coated pit mediated endocytosis, and mediate ultimate delivery to the Golgi and endoplasmic reticulum 33 .A£RI{169. 888 5 angel 1 called tax nportam caveolae itctlyll complex receptor I there the (ER)[169, 170] or lysosomes[l7l], such as the uptake of folate via the folate receptor. Viruses such as SV4O are internalized by caveolae to early endosomal antigen 1 (EEAl) -negative, Tflt—negative, fluid-phase marker-negative compartments called caveosomes; followed by subsequent transport to the ER[172]. Recently, an important role for caveolae has been well documented in cholesterol homeostasis. The caveolae structural integral membrane protein caveolin-l binds to cholesterol directly[l7l] and participates in shuffling of the flee cholesterol through ER, the Golgi complex and the cell surface[l73]. The HDL receptor SR-BI[174] and oxidized—LDL receptor CD36 have all been shown to be highly concentrated in caveolae[174, 175] where they mediate cholesterol homeostasis in the cell. The most important role of caveolae/lipid rafts at the cell surface is their function in signal transduction. It is well established that caveolae and lipid rafts both can act as signaling platforms to recruit and compartrnentalize the specific signaling receptors, adaptors, scaffolds and enzymes. Receptor tyrosine kinases such as FceRI receptor, T cell receptor, B cell receptor are all localized or recruited to the lipid rafts to initiate the Signaling upon ligand binding[159, 176, 177]. Several classes of Signaling proteins including receptor tyrosine kinases (RT Ks), Src farniy kinases, G proteins and GTPases have been identified by biochemical flactionation and immunohistochemistry, mainly flom macrovascular endothelial cells or tissues rich in endothelium such as lung[164]. Growth factor receptors such as PDGFR, EGFR, Flk-l/KDR together with receptors for EDG-l, uPAR, and G protein coupled receptors including bradykinin 2 Receptor (B2R) and endothelin receptor (ETA) are all reported to be partially localized in endothelial ceveolae[l64]. Caveolin-l, the most characterized caveolin among three caveolin 34 llllCOl 166]. pics; The l HIOUC lune deal . llllClI phos; ! . illllal isoforms in mammalian cells, interacts with and modulates the functions of many signaling proteins in caveolae, including endothelial isoform of nitric oxide synthase (eNOS), Src family tyrosine kinases and Ga proteins. Caveolae and caveolin-l coordinately regulate VEGF mediated endothelial proliferation, angiogenesis and permeability that requires NO produced by eNOS. Binding of caveolin-1 to eNOS inhibits eNOS activity as shown in Cav-l-/- mice a constitutive activation of eNOS[l66, 178]. Recent report of Cav-1-/- mice showed an absence of caveolae together with uncontrolled endothelial cell proliferation and increased microvascular permeability[165, 166]. Moreover, Si-RNA knock down of caveolin-1 in BAEC perturbs the Sphingosine 1 phosphate (Slp) and VEGF mediated Akt and Rac activation[179]. Recent advances in protein identification technology involving proteomics have permitted the identification of tens to thousands of proteins in the lipid microdomains. The protein components in lipid rafts flom Jurkat T cells[l80], neutrophils[181] and monocytes [182] or in caveolae/lipid rafts flom Cos cells[183], Vero cells[184] and human endothelial cells (HUVEC)[185] have all been characterized by mass spectroscopy following the isolation of caveolae/lipid rafts as triton-insoluble, low density membrane flactions. However, the protein components of caveolae/lipid rafts in microvascular endothelial cells flom retina have not been addressed yet. 5.2.3.3. n3-PUFA modification of caveolae/lipid rafts Since n3-PUFA is readily attached to the sn-2 position of membrane phospholipids, incorporation of PUFA into membrane phospholipids will increase the fluidity of the membranes in general, which may influence the localization and function 35 recrui produ locali Cairo Cil’eo litre I of membrane proteins. Inhibitory effects of n3-PUFA on T cells are reported to be primarily due to eicosanoid-independent effects of PUFA in vitro[186]. It has been proposed that n3-PUFA inhibits T cell activation by interfering with lipid raft signaling at the T cell receptor level[l87]. Enrichment of T cells with n3-PUFA leads to changes in lipid composition of the lipid rafts which result in the displacement of Src family kinases Lck and adaptor protein LAT flom the cytoplasmic leaflet of lipid rafts thus inhibiting T cell activation[188, 189]. N3-PUFA are incorporated into lipids of the cytoplasmic and exoplasmic leaflets of lipid rafts[187]. Moreover, dietary n3-PUFA treatment in vivo leads to a Significant decrease in the Sphingomyelin (SM) content of mouse T cell lipid rafts which may not only alters the exoplasmic membrane leaflets but also cytoplasmic leaflets[190]. The latest in viva mice feeding study of Fan, YY demonstrated that dietary fish oil or highly purified DHA22:6n3 are incorporated into the splenic T cells lipid rafts and soluble membrane phospholipids[190]. This results in a 30% decrease in rafts Sphingomyelin content. N3-PUFA feeding attenuates the antibody induced PKCB recruitment to the TCR complex thus the downstream NFIcB, AP-l activation and IL-2 production[191]. N3-PUFA also has been reported to alter lipid composition and protein localization of caveolae in mouse colon[192]. A 46% decrease in cholesterol level of caveolae was observed in n3-PUFA feeding compared with n6-PUFA. Also the caveolin-1 localization and the subdomain distribution of H-Ras and eNOS in caveolae were negatively modulated by n3-PUFA feeding[193]. 36 6. Re in h of the he: dISIrll can or in (ii Rein. iflli‘t “ ‘ '9 1th h. .1 iron unsa‘. cult; EPA "M u... 11: illp‘. mi The 6. Retinal endothelial cell and fatty acid profile Two tissues, retina and brain, are unique among other peripheral tissues due to a tight barrier that separates them flom circulation and a unique fatty acid profile with one of the highest levels of LCPUFA (especially 20:4n6 and 22:6n3) in the body[194-198]. Liver has been shown to be a key site for biosynthesis of LCPUFAS, where they are distributed into plasma lipids and lipoproteins for transport and tissue uptake. LCPUFAS can be transported through choriocapilaries to the retinal-pigmented epithelial cells (RPE) (an outer blood-retinal barrier) to support the needs of the photoreceptors in the retina. Retinal vascular endothelial cells form another inner blood-retinal barrier, a metabolically active interface responsible for fatty acid uptake flom blood and delivery to the retina. In addition to uptake and delivery, retinal endothelium can also actively remodel fatty acids through elongation and desaturation providing the retina with long chain highly unsaturated PUFA such as DHA22;6,.3 ([199] and our preliminary results). Indeed, in cultured primary bovine retinal endothelial cells, DHA22;6n3 and AA20;4,,6 each represents approximately 8% and 10% of total fatty acids and the retroconversion of DHA22:6n3 to EPA20;5..3 is negligible, indicating a specificity of fatty acid composition and metabolism different flom endothelial cells isolated flom other vascular tissues ([198] and our unpublished data). Although healthy retinas are relatively insensitive to changes in blood fatty acid profile (our preliminary data), in diabetic conditions, if hormonal control of retinal desaturases is altered, this could lead to a prominent change in retinal fatty acid profile. The retina selectively accumulates DHA22,6,,3, but not a-LA13;3n3 and EPA20;5,.3, the 37 pmdilt strum Own it: a graft: tabcI fond retina pail: who hit I precursors of DHA22;6,,3. Thus, in diabetes, a shift in plasma fatty acid profile flom products to substrates would favor an increased accumulation of LA13;2n6 without accumulation in its n3 counterpart, a-LA133n3 , and a decrease in AAzoamr, and DHA22;6n3. Overall fatty acid composition is expected to shift toward more n3 deficient, n6 rich state. The experimental data on the anticipated diabetes-induced changes in retinal fatty acid profiles are very sparse. Retinal fatty acid profiles in diabetic subjects[77] and alloxan diabetic rats maintained for 116 days without insulin[200] were reported. These studies found increased level of LA13;2n6 and reduced levels of AA20;4,,6 and DHA22:6n3 in diabetic retina[200]. However, this study was not followed up later on and the time course of profile change and correlation between the fatty acid profile and development of retinopathy is still not known. Overall, n3-PUFA deficiency in the retina was shown to have detrimental effect on the visual acuity[201-203]. 38 7. Oil diabetic mdlabl Isof pa the cf: diabetc Isolate: 7. Objective of Thesis The early steps that lead to the inflammatory state in the pathogenesis of diabetic retinopathy are not completely understood. The disturbance of lipid metabolism in diabetes with an increase in LA18;2,,6 and decrease in long chain AA20;4,,6 and DHA22;6n3 is of particular importance in the retina. The purpose of this dissertation is to investigate the effect of diabetic dyslipidemia especially the n6-PUFA/n3-PUFA ratio change in diabetes an inflammatory Signaling using primary microvascular endothelial cells isolated flom human retina. Chapter H describes the proinflammatory aspect of n6-PUFA (LA18;2n6 and AA20;4,.6). Chapter 3 analyzes the anti-inflammatory role of n3-PUFA especially (DHA22;6n3) and its possible target PPARS on cytokine induced inflammatory signaling. Chapter IV details the characterization of protein components in a Specialized highly lipid-ordered microdomain called caveolae/lipid rafts. The final chapter describes investigations into the anti-inflammatory mechanism of DHA22:6n3 by its potential role in the modification of lipid composition in the caveolae/lipid rafts. Results of this work will add to our understanding of the role of dyslipidemia in the development of diabetic retinopathy and facilitate new treatments and preventive strategies. 39 8. References 10. ll. 12. l3. 14. Frank, R.N., Diabetic retinopathy. N Engl J Med, 2004. 350(1): p. 48-58. Stefansson, E., Diabetic retinopathy. N Engl J Med, 2004. 350(24): p. 2525-6; author reply 2525-6. Aiello, L.P., et al., Diabetic retinopathy. Diabetes Care, 1998. 21(1): p. 143-56. Ferris, F.L., 3rd, M.D. Davis, and L.M. Aiello, Treatment of diabetic retinopathy. N Engl J Med, 1999. 341(9): p. 667-78. Bandello, F., et al., Diabetic macular edema: classification, medical and laser therapy. Semin Ophthalmol, 2003. 18(4): p. 251-8. Retinopathy and nephropathy in patients with type I diabetes four years after a trial of intensive therapy. The Diabetes Control and Complications T rial/Epidemiology of Diabetes Interventions and Complications Research Group. N Engl J Med, 2000. 342(6): p. 381-9. Comer, GM. and TA. Ciulla, Pharmacatherapyfar diabetic retinopathy. Curr Opin Ophthalmol, 2004. 15(6): p. 508-18. Powell, ED. and RA. Field, Diabetic Retinopathy and Rheumatoid Arthritis. Lancet, 1964. 41: p. 17-8. Schroder, S., W. Palinski, and G.W. Schmid-Schonbein, Activated monocytes and granulocytes, capillary nonperfusian, and neovascularization in diabetic retinopathy. Am J Pathol, 1991. 139(1): p. 81-100. Joussen, A.M., et al., Nansteraidal anti-inflammatory drugs prevent early diabetic retinopathy via TNF-alpha suppression. Faseb J, 2002. 16(3): p. 438-40. Adamis, A.P., Is diabetic retinopathy an inflammatory disease? Br J Ophthalmol, 2002. 86(4): p. 363-5. McLeod, D.S., et al., Enhanced expression of intracellular adhesion molecule-1 and P-selectin in the diabetic human retina and choroid. Am J Pathol, 1995. 147(3): p. 642-53. MacKinnon, J .R., R.M. Knott, and J .V. Forrester, Altered L-selectin expression in lymphocytes and increased adhesion to endothelium in patients with diabetic retinopathy. Br J Ophthalmol, 2004. 88(9): p. 1137-41. Kern, TS. and R.L. Engerman, Pharmacological inhibition of diabetic retinopathy: aminaguanidine and aspirin. Diabetes, 2001. 50(7): p. 1636-42. 40 ‘11 hi 3.“! (J 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Joussen, A.M., et al., Long-term leukocyte-mediated vascular pathology in the diabetic retina. Investigative Ophthalmology & Visual Science, 2003. 44: p. U221-U221. Limb, G.A., et al., Distribution of INF alpha and its reactive vascular adhesion molecules in fibravascular membranes of proliferative diabetic retinopathy. Br J Ophthalmol, 1996. 80(2): p. 168-73. Limb, G.A., ct al., Evidence for control of tumour necrosis factor-alpha (INF - alpha) activity by INF receptors in patients with proliferative diabetic retinopathy. Clin Exp Irnmunol, 1999. 115(3): p. 409-14. Limb, G.A., et al., Platelet expression af tumour necrosis factor-alpha (INF - alpha), INF receptors and intercellular adhesion molecule-1 (ICAM-1) in patients with proliferative diabetic retinopathy. Clin Exp Immunol, 1999. 118(2): p. 213-8. Spranger, J ., et al., [ INF -alpha level in the vitreous body. Increase in neovascular eye diseases and proliferative diabetic retinopathy]. Med Klin (Munich), 1995. 90(3): p. 1347. Doganay, S., et al., Comparison of serum N0, INF -alpha, IL-1 beta, sIL-ZR, IL-6 and IL-8 levels with grades of retinopathy in patients with diabetes mellitus. Eye, 2002. 16(2): p. 163-70. Matsumoto, T. and L. Claesson-Welsh, VEGF receptor signal transduction. Sci STKE, 2001. 2001(112): p. RE21. Brekken, RA. and RE. Thorpe, VEGF- VEGF receptor complexes as markers of tumor vascular endothelium. J Control Release, 2001. 74(1-3): p. 173-81. Adamis, A.P., et al., Increased vascular endothelial growth factor levels in the vitreous of eyes with proliferative diabetic retinopathy. Am J Ophthalmol, 1994. 118(4): p. 445-50. Murata, T., et al., The relation between expression of vascular endothelial growth factor and breakdown of the blood-retinal barrier in diabetic rat retinas. Lab Invest, 1996. 74(4): p. 819-25. Aiello, L.P., et al., Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc Natl Acad Sci U S A, 1995. 92(23): p. 10457-61. Qaum, T., et al., VEGF-initiated blood-retinal barrier breakdown in early diabetes. Invest Ophthalmol Vis Sci, 2001. 42(10): p. 2408-13. 41 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. Witmer, A.N., et al., Vascular endothelial growth factors and angiogenesis in eye disease. Prog Retin Eye Res, 2003. 22(1): p. 1-29. Ellis, E.A., et al., Increased H202, vascular endothelial growth factor and receptors in the retina of the BBZ/Wor diabetic rat. Free Radic Biol Med, 2000. 28(1): p. 91-101. Aiello, L.P., et al., Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med, 1994. 331(22): p. 1480-7. Amin, R.H., et al., Vascular endothelial growth factor is present in glial cells of the retina and optic nerve of human subjects with nonproliferative diabetic retinopathy. Invest Ophthalmol Vis Sci, 1997. 38(1): p. 36-47. Frank, R.N., et al., Basic fibroblast growth factor and vascular endothelial growth factor are present in epiretinal and choroidal neovascular membranes. Am J Ophthalmol, 1996. 122(3): p. 393-403. Obrosova, I.G., et al., Antioxidants attenuate early up regulation of retinal vascular endothelial growth factor in streptozotocin-diabetic rats. Diabetologia, 2001. 44(9): p. 1102-10. Joussen, A.M., et al., Retinal vascular endothelial growth factor induces intercellular adhesion molecule-1 and endothelial nitric oxide synthase expression and initiates early diabetic retinal leukocyte adhesion in viva. Am J Pathol, 2002. 160(2): p. 501-9. Usui, T., et al., VEGF(164(165)) as the Pathological Isoform: Diflerential Leukocyte and Endothelial Responses through VEGFRI and VEGFRZ. Invest Ophthalmol Vis Sci, 2004. 45(2): p. 368-374. Kowluru, RA. and S. Odenbach, Role of interleukin-1 beta in the development of retinopathy in rats: effect of antioxidants. Invest Ophthalmol Vis Sci, 2004. 45(11): p. 4161-6. Kowluru, RA. and S. Odenbach, Role of interleukin-1 beta in the pathogenesis of diabetic retinopathy. Br J Ophthalmol, 2004. 88(10): p. 1343-7. Rothwarf, D.M. and M. Karin, Ihe NF -kappa B activation pathway: a paradigm in information transfer from membrane to nucleus. Sci STKE, 1999. 1999(5): p. REl . Karin, M. and A. Lin, NF -kappaB at the crossroads of life and death. Nat Irnmunol, 2002. 3(3): p. 221-7. 42 39. 40. 41. 42. 43. 45. 46. 47. 48. 49. 50. 51. Karin, M. and Y. Ben-Neriah, Phosphorylation meets ubiquitination: the control of NF -[kappa]B activity. Annu Rev Irnmunol, 2000. 18: p. 621-63. Li, X. and GR. Stark, NF kappaB-dependent signaling pathways. Exp Hematol, 2002. 30(4): p. 285-96. Monaco, C. and E. Paleolog, Nuclear factor kappaB: a potential therapeutic target in atherosclerosis and thrombosis. Cardiovasc Res, 2004. 61 (4): p. 671-82. Romeo, G., et al., Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes. Diabetes, 2002. 51(7): p. 2241-8. Kowluru, R.A., et al., Diabetes-induced activation of nuclear transcriptional factor in the retina, and its inhibition by antioxidants. Free Radic Res, 2003. 37(11): p. 1169-80. Moore, T.C., et al., Ihe role of advanced glycation end products in retinal microvascular Ieukostasis. Invest Ophthalmol Vis Sci, 2003. 44(10): p. 4457-64. Stitt, A.W., et al., Advanced glycation end products (AGES) co-localize with AGE receptors in the retinal vasculature of diabetic and of A GE-infused rats. Am J Pathol, 1997. 150(2): p. 523-31. Curtis, T.M. and C.N. Scholfield, The role of lipids and protein kinase Cs in the pathogenesis of diabetic retinopathy. Diabetes Metab Res Rev, 2004. 20(1): p. 28-43. Hammes, H.P., et al., Benfotiamine blocks three major pathways of hyperglycemic damage and prevents experimental diabetic retinopathy. Nat Med, 2003. 9(3): p. 294-9. Engerman, R.L. and T.S. Kern, Aldose reductase inhibition fails to prevent retinopathy in diabetic and galactosemic dogs. Diabetes, 1993. 42(6): p. 820-5. Engerman, R.L., T.S. Kern, and ME. Larson, Nerve conduction and aldose reductase inhibition during 5 years of diabetes or galactosaemia in dogs. Diabetologia, 1994. 37(2): p. 141-4. Bierhaus, A., et al., Diabetes-associated sustained activation of the transcription factor nuclear factor-kappaB. Diabetes, 2001. 50(12): p. 2792-808. Basta, G., et al., Advanced glycation end products activate endothelium through signal-transduction receptor RAGE: a mechanism for amplification of inflammatory responses. Circulation, 2002. 105(7): p. 816-22. 43 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. Schmidt, A.M., et al., The multiligand receptor RAGE as a progression factor amplifying immune and inflammatory responses. J Clin Invest, 2001. 108(7): p. 949-55. Hofmann, M.A., et al., RAGE mediates a novel proinflammatory axis: a central cell surface receptor for SI 00/calgranulin polypeptides. Cell, 1999. 97(7): p. 889- 901. Soulis, T., et al., Advanced glycation end products and their receptors co-localise in rat organs susceptible to diabetic microvascular injury. Diabetologia, 1997. 40(6): p. 619-28. Hudson, B.I., et al., Eflects of novel polymorphisms in the RAGE gene on transcriptional regulation and their association with diabetic retinopathy. Diabetes, 2001. 50(6): p. 1505-11. Koya, D. and G.L. King, Protein kinase C activation and the development of diabetic complications. Diabetes, 1998. 47(6): p. 859-66. Ishii, H., et al., Amelioration of vascular dysfunctions in diabetic rats by an oral PKC beta inhibitor. Science, 1996. 272(5262): p. 728-31. Kuboki, K., et al., Regulation of endothelial constitutive nitric oxide synthase gene expression in endothelial cells and in vivo: a specific vascular action of insulin. Circulation, 2000. 101(6): p. 676-81. Koya, D., eta1., Characterization of protein kinase C beta isoform activation on the gene expression of transforming growth factor-beta, extracellular matrix components, and prostanoids in the glomeruli of diabetic rats. J Clin Invest, 1997 . 100(1): p. 115-26. Feener, E.P., et al., Role of protein Ia'nase C in glucose- and angiotensin II- induced plasminogen activator inhibitor expression. Contrib Nephrol, 1996. 118: p. 180-7. Pieper, G.M. and H. Riaz ul, Activation of nuclear factor-kappaB in cultured endothelial cells by increased glucose concentration: prevention by calphostin C. J Cardiovasc Pharmacol, 1997. 30(4): p. 528-32. Du, X.L., et al., Hyperglycemia-induced mitochondrial superoxide overproduction activates the hexosamine pathway and induces plasminogen activator inhibitor-1 expression by increasing Spl glycosylation. Proc Natl Acad Sci U S A, 2000. 97(22): p. 12222-6. 63. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. James, L.R., et al., Flux through the hexosamine pathway is a determinant of nuclear factor kappaB- dependent promoter activation. Diabetes, 2002. 51(4): p. 1 146-56. Brownlee, M., Biochemistry and molecular cell biology of diabetic complications. Nature, 2001. 414(6865): p. 813-20. Nishikawa, T., et al., Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage. Nature, 2000. 404(6779): p. 787-90. Laws, A., et al., Differences in insulin suppression of free fatty acid levels by gender and glucose tolerance status. Relation to plasma triglyceride and apolipoprotein B concentrations. Insulin Resistance Atherosclerosis Study (IRAS) Investigators. Arterioscler Thromb Vasc Biol, 1997. 17(1): p. 64-71. Poisson, J .P., Essential fatty acid metabolism in diabetes. Nutrition, 1989. 5(4): p. 263-6. Vessby, B., et al., Diet, nutrition and diabetes mellitus. Ups J Med Sci, 2000. 105(2): p. 151-60. Vessby, B., Dietary fat and insulin action in humans. Br J Nutr, 2000. 83 Suppl 1: p. S91-6. Zenobi, P.D., et al., Improvement of lipid profile in type 2 (non-insulin-dependent) diabetes mellitus by insulin-like growth factor I. Diabetologia, 1993. 36(5): p. 465-9. Lewis, G.F., et al., Disordered fat storage and mobilization in the pathogenesis of insulin resistance and type 2 diabetes. Endocr Rev, 2002. 23(2): p. 201-29. Baldeweg, 8.13., et al., Insulin resistance, lipid and fatty acid concentrations in 867 healthy Europeans. European Group for the Study of Insulin Resistance (EGIR). Eur J Clin Invest, 2000. 30(1): p. 45-52. Decsi, T., et al., Polyunsaturated fatty acids in plasma and erythrocyte membrane lipids of diabetic children. Prostaglandins Ieukot Essent Fatty Acids, 2002. 67(4): p. 203-10. Brown, J .13., RM. Lindsay, and RA. Riemersma, Linoleic acid metabolism in the spontaneously diabetic rat: delta 6-desaturase activity vs. product/precursor ratios. Lipids, 2000. 35(12): p. 1319-23. Brenner, R.R., A.M. Bernasconi, and HA. Garda, Effect of experimental diabetes on the fatty acid composition, molecular species of phosphatidyl-choline and 45 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. physical properties of hepatic microsomal membranes. Prostaglandins Leukot Essent Fatty Acids, 2000. 63(3): p. 167-76. Goldberg, RB. and D. Capuzzi, Lipid disorders in type I and type 2 diabetes. Clin Lab Med, 2001. 21(1): p. 147-72, vii. F utterman, S. and C. Kupfer, The fatty acid composition of the retinal vasculature of normal and diabetic human eyes. Invest Ophthalmol, 1968. 7(1): p. 105-8. Sprecher, H. and Q. Chen, Polyunsaturated fatty acid biosynthesis: a microsomal- peroxisomal process. Prostaglandins Leukot Essent Fatty Acids, 1999. 60(5-6): p. 3 1 7-21 . Goldberg, R.B., Lipid disorders in diabetes. Diabetes Care, 1981. 4(5): p. 561-72. Goldberg, 1.1., Clinical review 124: Diabetic dyslipidemia: causes and consequences. J Clin Endocrinol Metab, 2001. 86(3): p. 965-71. Weinstock, P.H., et al., Lipoprotein lipase controls fatty acid entry into adipose tissue, but fat mass is preserved by endogenous synthesis in mice deficient in adipose tissue lipoprotein lipase. Proc Natl Acad Sci U S A, 1997. 94(19): p. 10261-6. Zechner, R., The tissue-specific expression of lipoprotein lipase: implications for energy and lipoprotein metabolism. Curr Opin Lipidol, 1997. 8(2): p. 77-88. Brenner, R.R., Hormonal modulation of delta6 and delta5 desaturases: case of diabetes. Prostaglandins Leukot Essent Fatty Acids, 2003. 68(2): p. 151-62. Coppack, S.W., et al., Adipose tissue metabolism in obesity: lipase action in vivo before and after a mixed meal. Metabolism, 1992. 41(3): p. 264-72. Yeaman, S.J., Hormone-sensitive lipase-mew roles for an old enzyme. Biochem J, 2004. 379(Pt 1): p. 11-22. Kraemer, F .B. and W.J. Shen, Horrnone-sensitive lipase: control of intracellular tri-(di-)acylglycerol and cholesteryl ester hydrolysis. J Lipid Res, 2002. 43(10): p. 1585-94. Mead, J .R., S.A. Irvine, and DP. Ramji, Lipoprotein lipase: structure, function, regulation, and role in disease. J Mol Med, 2002. 80(12): p. 753-69. Frenais, R., et al., In vivo evidence for the role of lipoprotein lipase activity in the regulation of apolipoprotein AI metabolism: a kinetic study in control subjects and patients with type II diabetes mellitus. J Clin Endocrinol Metab, 2001. 86(5): p. 1962-7. 46 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. Davis, M.D., et al., Risk factors for high-risk proliferative diabetic retinopathy and severe visual loss: Early Treatment Diabetic Retinopathy Study Report #18. Invest Ophthalmol Vis Sci, 1998. 39(2): p. 233-52. Ferris, F.L., 3rd, E.Y. Chew, and B.J. Hoogwerf, Serum lipids and diabetic retinopathy. Early Treatment Diabetic Retinopathy Study Research Group. Diabetes Care, 1996. 19(11): p. 1291-3. Chew, E.Y., et al., Association of elevated serum lipid levels with retinal hard exudate in diabetic retinopathy. Early Treatment Diabetic Retinopathy Study (ETDRS) Report 22. Arch Ophthalmol, 1996. 114(9): p. 1079-84. Natarajan, R., et al., Role of lZ-lipoxygenase and oxidant stress in hyperglycaemia-induced acceleration of atherosclerosis in a diabetic pig model. Diabetologia, 2002. 45(1): p. 125-33. Sen, K., et al., Simvastatin retards progression of retinopathy in diabetic patients with hypercholesterolemia. Diabetes Res Clin Pract, 2002. 56(1): p. 1-11. Chowdhury, T.A., et al., The role of serum lipids in exudative diabetic maculopathy: is there a place for lipid lowering therapy? Eye, 2002. 16(6): p. 689-93. Gupta, A., et al., Lipid-lowering drug atorvastatin as an adjunct in the management of diabetic macular edema. Am J Ophthalmol, 2004. 137(4): p. 675- 82. Klein, R., et al., The relation of atherosclerotic cardiovascular disease to retinopathy in people with diabetes in the Cardiovascular Health Study. Br J Ophthalmol, 2002. 86(1): p. 84-90. Lyons, T.J., et al., Diabetic retinopathy and serum lipoprotein subclasses in the DCCT/EDIC cohort. Invest Ophthalmol Vis Sci, 2004. 45(3): p. 910-8. Funk, C.D., Prostaglandins and leukatrienes: advances in eicosanoid biology. Science, 2001. 294(5548): p. 1871-5. Hla, T., et al., Cyclooxygenase gene expression in inflammation and angiogenesis. Ann N Y Acad Sci, 1993. 696: p. 197-204. Brash, A.R., Lipoxygenases: occurrence, functions, catalysis, and acquisition of substrate. J Biol Chem, 1999. 274(34): p. 23679-82. 47 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. Capdevila, J .H., J .R. F alck, and RC. Harris, Cytochrome P450 and arachidonic acid bioactivation. Molecular and functional properties of the arachidonate monooxygenase. J Lipid Res, 2000. 41(2): p. 163-81. Zeldin, D.C., Epoxygenase pathways of arachidonic acid metabolism. J Biol Chem, 2001. 276(39): p. 36059-62. Dubois, R.N., et al., Cyclooxygenase in biology and disease. Faseb J, 1998. 12(12): p. 1063-73. Vane, J .R., Y.S. Bakhle, and RM. Botting, Cyclooxygenases 1 and 2. Annu Rev Pharmacol Toxicol, 1998. 38: p. 97-120. Shanmugam, N., I.T. Gaw Gonzalo, and R. Natarajan, Molecular mechanisms of high glucose-induced cyclooxygenase-2 expression in monocytes. Diabetes, 2004. 53(3): p. 795-802. Shanmugam, N., et al., Regulation of cyclooxygenase-Z expression in monocytes by ligation of the receptor for advanced glycation end products. J Biol Chem, 2003. 278(37): p. 34834-44. Boeglin, W.E., R.B. Kim, and AR. Brash, A [ZR-lipoxygenase in human skin: mechanistic evidence, molecular cloning, and expression. Proc Natl Acad Sci U S A, 1998. 95(12): p. 6744-9. Yu, 2., et al., The lipoxygenase gene ALOXE3 implicated in skin difi'erentiation encodes a hydroperoxide isomerase. Proc Natl Acad Sci U S A, 2003. 100(16): p. 9162-7. Zhang, Y.Y., et al., Expression of 5-lipoxygenase in pulmonary artery endothelial cells. Biochem J, 2002. 361(Pt 2): p. 267-76. Patricia, M.K., et al., Lipoxygenase products increase monocyte adhesion to human aortic endothelial cells. Arterioscler Thromb Vasc Biol, 1999. 19(11): p. 2615-22. Henricks, P.A., et al., 9- and I3-hydroxy-linoleic acid possess chemotactic activity for bovine and human polymorphonuclear leukocytes. Prostaglandins, 1991. 41(1): p. 21-7. Hatley, M.E., et al., Increased production of 12/15 lipoxygenase eicosanoids accelerates monocyte/endothelial interactions in diabetic db/db mice. J Biol Chem, 2003. 278(28): p. 25369-75. 48 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. Reilly, K.B., et al., 12/15-Lipoxygenase activity mediates inflammatory monocyte/endothelial interactions and atherosclerosis in vivo. J Biol Chem, 2004. 279(10): p. 9440-50. Augustin, A.J., et al., Detection of eicosanoids in epiretinal membranes of patients sufl'ering from proliferative vitreoretinal diseases. Br J Ophthalmol, 1997. 81(1): p. 58-60. Szekeres, C.K., et al., Eicosanoid activation of extracellular signal-regulated kinasel/Z in human epidermoid carcinoma cells. J Biol Chem, 2000. 275(49): p. 3 883 1-41 . Szekeres, C.K., et al., Eicosanoid I 2(S)-HE T E activates phosphatidylinositol 3- kinase. Biochem Biophys Res Commun, 2000. 275(2): p. 690-5. Szekeres, C.K., M. Trikha, and K.V. Honn, I 2(S)-HE T E, pleiotropicfitnctions, multiple signaling pathways. Adv Exp Med Biol, 2002. 507: p. 509-15. Sultana, C., et al., Lipoxygenase metabolites induced expression of adhesion molecules and transendothelial migration of monocyte-like HL-60 cells is linked to protein kinase C activation. J Cell Physiol, 1996. 167(3): p. 477-87. Zhou, J ., R.N. F ariss, and RS. Zelenka, Synergy of epidermal growth factor and I2(S)-hydroxyeicosatetraenoate on protein kinase C activation in lens epithelial cells. J Biol Chem, 2003. 278(7): p. 5388-98. Lin, B., et al., I2(S)-hydroxyeicosatetraenoic acid and 13(S)- hydroxyoctadecadienoic acid regulation of protein kinase C-alpha in melanoma cells: role of receptor-mediated hydrolysis of inositol phospholipids. Proc Natl Acad Sci U S A, 1995. 92(20): p. 9323-7. Nataraj an, R. and J .L. Nadler, Lipid inflammatory mediators in diabetic vascular disease. Arterioscler Thromb Vasc Biol, 2004. 24(9): p. 1542-8. Stulnig, T.M., Immunomodulation by polyunsaturated fatty acids: mechanisms and eflects. Int Arch Allergy Immunol, 2003. 132(4): p. 310-21. Trebble, T., et al., Inhibition of tumour necrosis factor-alpha and interleukin 6 production by mononuclear cells following dietary fish-oil supplementation in healthy men and response to antioxidant co—supplementation. Br J Nutr, 2003. 90(2): p. 405-12. Lee, T.H., et al., Eflect of dietary enrichment with eicosapentaenoic and docosahexaenoic acids on in vitro neutrophil and monocyte leukotriene generation and neutrophil function. N Engl J Med, 1985. 312(19): p. 1217-24. 49 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. Simopoulos, A.P., Omega-3 fatty acids in inflammation and autoimmune diseases. J Am Coll Nutr, 2002. 21(6): p. 495-505. Spector, S.L. and ME. Surette, Diet and asthma: has the role of dietary lipids been overlooked in the management of asthma? Ann Allergy Asthma Immunol, 2003. 90(4): p. 371-7; quiz 377-8, 421. Calder, P.C., N-3 polyunsaturated fatty acids and inflammation: from molecular biology to the clinic. Lipids, 2003. 38(4): p. 343-52. Calder, P.C., Polyunsaturated fatty acids, inflammation, and immunity. Lipids, 2001. 36(9): p. 1007-24. Calder, P.C., et al., Fatty acids and lymphocyte functions. Br J Nutr, 2002. 87 Suppl 1: p. S31-48. Serhan, C.N., et al., Novel functional sets of lipid-derived mediators with antiinflammatory actions generated from omega-3 fatty acids via cyclooxygenase 2-nonsteroidal antiinflammatory drugs and transcellular processing. J Exp Med, 2000. 192(8): p. 1197-204. Hong, S., et al., Novel docosatrienes and l 7S-resolvins generated from docosahexaenoic acid in murine brain, human blood, and glial cells. Autacoids in anti-inflammation. J Biol Chem, 2003. 278(17): p. 14677-87. Serhan, C .N., et al., Resolvins, docosatrienes, and neuroprotectins, novel omega- 3-derived mediators, and their aspirin-triggered endogenous epimers: an overview of their protective roles in catabasis. Prostaglandins Other Lipid Mediat, 2004. 73(3-4): p. 155-72. Pawar, A., et al., Fatty acid regulation of liver X receptors (L20?) and peroxisome proliferator-activated receptor alpha (PPARalpha) in HEK293 cells. J Biol Chem, 2002. 277(42): p. 39243-50. Desvergne, B. and W. Wahli, Peroxisome proliferator-activated receptors: nuclear control of metabolism. Endocr Rev, 1999. 20(5): p. 649-88. Xu, H.E., et al., Molecular recognition of fatty acids by peroxisome proliferator- activated receptors. Mol Cell, 1999. 3(3): p. 397-403. Diep, Q.N., R.M. Touyz, and EL. Schiffrin, Docosahexaenoic acid, a peroxisome proliferator-activated receptor-alpha ligand, induces apoptosis in vascular smooth muscle cells by stimulation of p38 mitogen-activated protein kinase. Hypertension, 2000. 36(5): p. 851-5. 50 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. Kliewer, S.A., et al., A prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor gamma and promotes adipocyte dtflerentiation. Cell, 1995. 83(5): p. 813-9. Shappell, S.B., et al., l5S-Hydroxyeicosatetraenoic acid activates peroxisome proliferator-activated receptor gamma and inhibits proliferation in PC 3 prostate carcinoma cells. Cancer Res, 2001. 61(2): p. 497-503. Forman, B.M., J. Chen, and KM. Evans, Hypolipidemic drugs, polyunsaturated fatty acids, and eicosanoids are ligands for peroxisome proliferator-activated receptors alpha and delta. Proc Natl Acad Sci U S A, 1997. 94(9): p. 4312-7. Daynes, RA. and DC. Jones, Emerging roles of PPARs in inflammation and immunity. Nat Rev Immunol, 2002. 2(10): p. 748-59. Devchand, P.R., et al., The PPARalpha-leukotriene B4 pathway to inflammation control. Nature, 1996. 384(6604): p. 39-43. Neve, B.P., et al., PPARalpha agonists inhibit tissue factor expression in human monocytes and macrophages. Circulation, 2001. 103(2): p. 207-12. Staels, B., et al., Activation of human aortic smooth-muscle cells is inhibited by PPARalpha but not by PPARgamma activators. Nature, 1998. 393(6687): p. 790- 3. Delerive, P., et al., Peroxisome proliferator-activated receptor alpha negatively regulates the vascular inflammatory gene response by negative cross-talk with transcription factors NF -kappaB and AP-I. J Biol Chem, 1999. 274(45): p. 32048-54. Jackson, S.M., et al., Peroxisome proliferator-activated receptor activators target human endothelial cells to inhibit leukocyte-endothelial cell interaction. Arterioscler Thromb Vasc Biol, 1999. 19(9): p. 2094-104. Marx, N., et al., PPARalpha activators inhibit cytokine-induced vascular cell adhesion molecule-1 expression in human endothelial cells. Circulation, 1999. 99(24): p. 3125-31. Marx, N., P. Libby, and J. Plutzky, Peroxisome proliferator-activated receptors (PPARs) and their role in the vessel wall: possible mediators of cardiovascular risk? J Cardiovasc Risk, 2001. 8(4): p. 203-10. Pasceri, V., J .T. Willerson, and ET. Yeh, Direct proinflammatory effect of C- reactive protein on human endothelial cells. Circulation, 2000. 102(18): p. 2165-8. 51 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. Pasceri, V., et al., Modulation of vascular inflammation in vitro and in vivo by peroxisome proliferator-activated receptor-gamma activators. Circulation, 2000. 101(3): p. 235-8. Chinetti, G., J .C . Fruchart, and B. Staels, Peroxisome proliferator—activated receptors (PPARs): nuclear receptors at the crossroads between lipid metabolism and inflammation. Inflamm Res, 2000. 49(10): p. 497-505. Rival, Y., et al., PPARalpha and PPARdelta activators inhibit cytokine-induced nuclear translocation of NF -kappaB and expression of VCAM-1 in EAhy926 endothelial cells. Eur J Pharmacol, 2002. 435(2-3): p. 143-51. Tan, N.S., et al., Critical roles of PPAR beta/delta in keratinocyte response to inflammation. Genes Dev, 2001. 15(24): p. 3263-77. Ruan, H., H.J. Pownall, and HF. Lodish, T roglitazone antagonizes tumor necrosis factor-alpha-induced reprogramming of adipocyte gene expression by inhibiting the transcriptional regulatory functions of NF -kappaB. J Biol Chem, 2003. 278(30): p. 28181-92. Delerive, P., et al., Induction of [kappa Balpha Expression as a Mechanism Contributing to the Anti-inflammatory Activities of Peroxisome Proliferator- activated Receptor-alpha Activators. J. Biol. Chem., 2000. 275(47): p. 36703- 36707. Yamada, B., The fine structure of the renal glomerulus of the mouse. J Histochem Cytochem, 1955. 3(4): p. 309. Razani, B., S.E. Woodman, and MP. Lisanti, Caveolae: from cell biology to animal physiology. Pharrnacol Rev, 2002. 54(3): p. 431-67. Schnitzer, J .E., et al., Separation of caveolae from associated microdomains of GPI-anchored proteins. Science, 1995. 269(5229): p. 1435-9. Oh, P. and J .E. Schnitzer, Immunoisolation of caveolae with high aflinity antibody binding to the oligomeric caveolin cage. Toward understanding the basis of purification. J Biol Chem, 1999. 274(33): p. 23144-54. Simons, K. and D. Toomre, Lipid rafts and signal transduction. Nat Rev Mol Cell Biol, 2000. 1(1): p. 31-9. Simons, K. and E. Ikonen, Functional rafts in cell membranes. Nature, 1997. 387(6633): p. 569-72. Vanna, R. and S. Mayor, GPI-anchored proteins are organized in submicron domains at the cell surface. Nature, 1998. 394(6695): p. 798-801 . 52 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 1 74. Seheiffele, P., M.G. Roth, and K. Simons, Interaction of influenza virus haemagglutinin with sphingolipid-cholesterol membrane domains via its transmembrane domain. Embo J, 1997. 16(18): p. 5501-8. Carver, LA. and J .E. Schnitzer, Caveolae: mining little caves for new cancer targets. Nat Rev Cancer, 2003. 3(8): p. 571-81. Frank, P.G., eta1., Caveolin, Caveolae, and Endothelial Cell Function. Arterioscler Thromb Vase Biol, 2003. 23(7): p. 1161-1168. Drab, M., et al., Loss of caveolae, vascular dysfunction, and pulmonary defects in caveolin-1 gene-disrupted mice. Science, 2001. 293(5539): p. 2449-52. Razani, B., et al., Caveolin-I null mice are viable but show evidence of hyperproliferative and vascular abnormalities. J Biol Chem, 2001. 276(41): p. 38121-38. Schubert, W., et al., Caveolae-deficient endothelial cells show defects in the uptake and transport of albumin in vivo. J Biol Chem, 2001. 276(52): p. 48619-22. Schnitzer, J .E., J. Liu, and P. Oh, Endothelial caveolae have the molecular transport machinery for vesicle budding, docking, and fusion including VAMP, NSF, SNAP, annexins, and GT Pases. J Biol Chem, 1995. 270(24): p. 14399-404. Le, P.U. and LR. Nabi, Distinct caveolae-mediated endocytic pathways target the Golgi apparatus and the endoplasmic reticulum. J Cell Sci, 2003. 116(Pt 6): p. 1059-71. Nabi, LR. and P.U. Le, Caveolae/raft-dependent endocytosis. J Cell Biol, 2003. 161(4): p. 673-7. Murata, M., et al., VIP21/caveolin is a cholesterol-binding protein. Proc Natl Acad Sci U S A, 1995. 92(22): p. 10339-43. Pelkmans, L., J. Kartenbeek, and A. Helenius, Caveolar endocytosis of simian virus 40 reveals a new two-step vesicular-transport pathway to the ER. Nat Cell Biol, 2001. 3(5): p. 473-83. Frank, P.G., et al., Influence of caveolin-1 on cellular cholesterol efi'lux mediated by high-density lipoproteins. Am J Physiol Cell Physiol, 2001. 280(5): p. C1204- 14. Uittenbogaard, A., et al., High density lipoprotein prevents oxidized low density lipoprotein-induced inhibition of endothelial nitric-oxide synthase localization and activation in caveolae. J Biol Chem, 2000. 275(15): p. 11278-83. 53 175. 176. 177. 178. 179. 180. 181. 182. 183. l 84. 185. 186. Lisanti, M.P., eta1., Characterization of caveolin-rich membrane domains isolated from an endothelial-rich source: implications for human disease. J Cell Biol, 1994. 126(1): p. 111-26. Cherukuri, A., M. Dykstra, and SK. Pierce, Floating the raft hypothesis: lipid rafts play a role in immune cell activation. Immunity, 2001. 14(6): p. 657-60. Sedwiek, CE. and A. Altman, Ordered just so: lipid rafts and lymphocyte fitnction. Sci STKE, 2002. 2002(122): p. RE2. Schubert, W., et al., Microvascular hyperpermeability in caveolin-1 (-/-) knock- out mice. Treatment with a specific nitric-oxide synthase inhibitor, L-name, restores normal microvascular permeability in Cav-I null mice. J Biol Chem, 2002. 277(42): p. 40091-8. Gonzalez, E., et al., Small interfering RNA-mediated down-regulation of caveolin- I differentially modulates signaling pathways in endothelial cells. J Biol Chem, 2004. 279(39): p. 40659-69. von Haller, P.D., et al., Mass spectrometric characterization of proteins extracted from Jurkat T cell detergent-resistant membrane domains. Proteomics, 2001. 1(8): p. 101 0-2 1 . Nebl, T., et al., Proteomic analysis of a detergent-resistant membrane skeleton from neutrophil plasma membranes. J Biol Chem, 2002. 277(45): p. 43399-409. Li, N., et al., Lipid raft proteomics: analysis of in-solution digest of sodium dodecyl sulfate-solubilized lipid raft proteins by liquid chromatography-matrix- assisted laser desorption/ionization tandem mass spectrometry. Proteomics, 2004. 4(10): p. 3156-66. Foster, L.J., C.L. De Hoog, and M. Mann, Unbiased quantitative proteomics of lipid rafts reveals high specificity for signaling factors. Proc Natl Acad Sci U S A, 2003. 100(10): p. 5813-8. Blonder, J ., et al., Proteomic analysis of detergent-resistant membrane rafts. Electrophoresis, 2004. 25(9): p. 1307-18. Sprenger, R.R., et al., Comparative proteomics of human endothelial cell caveolae and rafts using two-dimensional gel electrophoresis and mass spectrometry. Electrophoresis, 2004. 25(1): p. 156-72. Fowler, K.H., R.S. Chapkin, and D.N. MeMurray, Effects of purified dietary n-3 ethyl esters on murine T lymphocyte function. J Immunol, 1993. 151(10): p. 5186- 97. 54 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. Stulnig, T.M., et al., Polyunsaturated fatty acids inhibit T cell signal transduction by modification of detergent-insoluble membrane domains. J Cell Biol, 1998. 143(3): p. 637-44. Stulnig, T.M., et al., Polyunsaturated eicosapentaenoic acid displaces proteins fi'om membrane rafts by altering raft lipid composition. J Biol Chem, 2001. 276(40): p. 37335-40. Zeyda, M., et al., LAT displacement from lipid rafts as a molecular mechanism for the inhibition of T cell signaling by polyunsaturated fatty acids. J Biol Chem, 2002. 277(32): p. 28418-23. Fan, Y.Y., et al., Dietary (n-3) polyunsaturated fatty acids remodel mouse T -cell lipid rafts. J Nutr, 2003. 133(6): p. 1913-20. Fan, Y.Y., et al., Dietary docosahexaenoic acid suppresses T cell protein kinase Ctheta lipid raft recruitment and IL-2 production. J Immunol, 2004. 173(10): p. 615 1-60. Ma, D.W., et al., n-3 PUFA alter caveolae lipid composition and resident protein localization in mouse colon. Faseb J, 2004. 18(9): p. 1040-2. Ma, D.W., et al., n-3 PUFA and membrane microdomains: a new frontier in bioactive lipid research. J Nutr Biochem, 2004. 15(11): p. 700-6. Anderson, R.E., Lipids of ocular tissues. IV. A comparison of the phospholipids from the retina of six mammalian species. Exp Eye Res, 1970. 10(2): p. 339-44. O'Brien, J .S. and EL. Sampson, Fatty acid and fatty aldehyde composition of the major brain lipids in normal human gray matter, white matter, and myelin. J Lipid Res, 1965. 6(4): p. 545-51. O'Brien, J .S. and EL. Sampson, Lipid composition of the normal human brain: gray matter, white matter, and myelin. J Lipid Res, 1965. 6(4): p. 537-44. Fliesler, SJ. and RB. Anderson, Chemistry and metabolism of lipids in the vertebrate retina. Prog Lipid Res, 1983. 22(2): p. 79-131. Lecomte, M., et al., Docosahexaenoic acid is a major n-3 polyunsaturated fatty acid in bovine retinal mierovessels. J Neuroehem, 1996. 66(5): p. 2160-7. Delton-Vandenbroucke, 1., P. Grammas, and RE. Anderson, Polyunsaturated fatty acid metabolism in retinal and cerebral microvascular endothelial cells. J Lipid Res, 1997. 38(1): p. 147-59. 55 200. 201. 202. 203. Futterman, S., R. Sturtevant, and C. Kupfer, Effect of alloxan diabetes on the fatty acid composition of the retina. Invest Ophthalmol, 1969. 8(5): p. 542-4. Litman, B.J., et al., The role of docosahexaenoic acid containing phospholipids in modulating G protein-coupled signaling pathways: visual transduction. J Mol Neurosci, 2001. 16(2-3): p. 237-42; discussion 279-84. Mitchell, D.C., S.L. Niu, and B.J. Litman, Optimization of receptor-G protein coupling by bilayer lipid composition 1: kinetics of rhodopsin-transducin binding. J Biol Chem, 2001. 276(46): p. 42801-6. Niu, S.L., D.C. Mitchell, and B.J. Litman, Optimization of receptor-G protein coupling by bilayer lipid composition 11: formation of metarhodopsin H- transducin complex. J Biol Chem, 2001. 276(46): p. 42807-11. 56 II. Dyslipidemia, but not hyperglycemia, induces inflammatory adhesion molecules in human retinal vascular endothelial cells 1. Abstract 2139933 The initial determinants of retinal microvascular damage in diabetic retinopathy are not well understood, but are likely to be contributed by hyperglycemia and/or dyslipidemia. The purpose of this study was to examine the effect of fatty acids and hyperglycemia on human retinal vascular endothelial cells (hRVE) as means of mimicking diabetic metabolic disorders. Methods: The expression of adhesion molecules in hRVE and human umbilical vein endothelial cells (HUVEC) was assayed by Western Blotting and confirmed by leukocyte adhesion assay. The mechanisms underlying the induction of adhesion molecules by fatty acids were further investigated by using cyclooxygenase (COX), lipoxygenase (LOX) and P450 monooxygenase (MOX) inhibitors. Resulg; Treatment of hRVE cells with n6 polyunsaturated fatty acids (PUFA), 18:2,n6 and 20:4,n6, for up to 24 hrs, resulted in a significant induction of ICAM-1 and VCAM-l protein levels. In contrast, treatment with high glucose (22 mM) for 24 hrs did not affect CAM expression. Induction of CAM by n6 PUFA was correlated with enhanced leukocyte binding to hRVE cells. The effect of n6 PUFA on ICAM-1 and VCAM-l was blocked by an inhibitor of LOX, but not COX or MOX inhibitors. In contrast to hRVE cells, n6 PUFA did not induce ICAM-l or VCAM-1 in HUVEC. 57 Conclusion: The data obtained in this study demonstrate that acute exposure to linoleic or arachidonic acid, but not hyperglycemia induces inflammatory adhesion molecules expression in a LOX-dependent manner in microvascular hRVE cells, but not in HUVEC. These results are consistent with the emerging hypothesis recognizing early stage diabetic retinopathy as a low-grade chronic inflammatory disease. 58 2. Introduction Despite the progress made in the last decade in the understanding of the molecular mechanisms of diabetic retinopathy, the disease is still neither preventable nor curable. Diabetic retinopathy is characterized by capillary occlusions, microaneurysms, selective loss of intramural pericytes, acellular capillaries, hypertrophy of the basement membrane, and finally, angiogenesis and neovascularization. These morphological and pathophysio- logical changes occur late in the disease. The initial determinants of retinal micro- vascular darnage are not well understood. Recently very early stage diabetic retinopathy was recognized as a low-grade chronic inflammatory condition. Support for this notion is based on the finding that leukocytes (including monocytes, neutrophils and some lymphocytes) attach and transmigrate the endothelium in both experimental and human diabetic retinopathy[1-7]. Adhesion molecules, especially ICAM-1 and VCAM-l are involved in leukocyte attachment and transmigration[4-8]. ICAM-l is a member of the immunoglobulin superfamily of adhesion molecules whose ligands include leukocyte [32- integrins CD11a/CD18 (LFA-l) and CD11b/CD18 (Mae-1). Vascular endothelial ICAM-l is associated with adhesion and transmigration of leukocytes in the retina[2, 3] and in other vascular systems. Importantly, leukocyte infiltration and expression of retinal vascular ICAM-l coincide with many of the pathological lesions in diabetic retinopathy[3, 6]. While the mechanisms leading to ICAM-l induction in diabetic microvessels are not known, carbohydrate and lipid metabolic disorders are likely to be key causative factors in this process. Hyperglycemia and dyslipidemia are two major metabolic 59 disorders of diabetes mellitus. Despite considerable progress made in understanding of hyperglycemia-induced pathology over the last decade, the link between diabetic metabolic disorders and retinopathy still eludes us. The role of diabetic dyslipidemia in the development of microvascular complications has received much less attention. Insulin controls an array of enzymes and signaling molecules involved in lipogenesis and lipid metabolism. The insulin resistance of Type 2 diabetes is associated with increased plasma FFA levels, triglycerides, LDL cholesterol and a decrease in HDL cholesterol[9- 15]. FFA levels parallel the blood glucose level in diabetes and FFAs are often considered an indicator of the severity of the diabetic state[16]. In Type 1 diabetes, low portal insulin levels cause a reduction in delta-6-, delta—S- and delta-9-desaturases in the liver and a corresponding change in F F A profile resulting, for instance, in an increase in linoleic acid and in n6 PUFA/n3 PUFA ratio[10, 17-19]. Clinical data support the idea that dyslipidemia could be critical factor for the development of diabetic retinopathy. Thus, a recent clinical trial controlling diabetic dyslipidemia with Simvastatin, a HMG-CoA reductase inhibitor resulted in significant retardation of the progression of diabetic retinopathy[20-21]. The data from another population-based study suggested the association of retinopathy with cardiovascular disease and elevated plasma LDL cholesterol[22]. The Early Treatment Diabetic Retinopathy Study demonstrated that higher levels of serum lipids were associated with an increased risk of developing hard exudates in the macula and visual loss[23]. Randomized controlled clinical trials are currently in progress to examine whether lipid lowering agents will reduce the risk of incidence and progression of diabetic retinopathy[22]. 60 Recent progress in lipid research highlighted several pathways through which FFA can exert cellular and functional alterations. These include changes in membrane composition and function as well as in the regulation of gene expression and protein modifieation[24, 25]. Unsaturated fatty acids are substrates for oxygenases such as COX, LOX and MOX, as well as non-enzymatic oxidation. Fatty acids oxidation leads to generation of a variety of bioactive lipids such as eicosanoids, lipid hydroperoxides and isoprostanes[26-3 l ]. All of these pathways are likely to be relevant to endothelial cells. The role of fatty acids as the inflammatory agents leading to diabetic retinopathy has not been studied in detail and could represent a missing link between diabetes, dyslipidemia, and microvascular damage. The current study was designed to address this question by analysis of the effect of fatty acids on human retinal vascular endothelial (hRVE) cells. Our data strongly support the hypothesis that elevated plasma fatty acids induce an increase in inflammatory adhesion molecules leading to retinal vascular inflammation involving leukocyte attachment and transmigration. 61 3. Materials and Methods 3.1 Reagents and supplies DMEM and F12 culture medium, antibiotics, fetal bovine serum and trypsin were obtained from Invitrogen (Carlsbad, CA); and culture dishes and flasks from Corning. Commonly used chemicals and reagents and chemicals were from Sigma/Aldrich Chemical Co. (St. Louis, MO). TNFa and IL-1 B were from R&D Systems (Minneapolis, MN). PMA was from Sigma. 3.2. Cell culture and fatty acid treatments The present study utilized primary cultures of hRVE cells obtained from 3 separate donors. hRVE cells were prepared as previously described and maintained[32- 35] in growth medium consisting of DMEM/F 12 (Invitrogen, Carlsbad, CA), 5.5 mM glucose, 10% fetal bovine serum (Invitrogen), endothelial cell growth supplement (Upstate Biotechnologies, Inc., Lake Placid, NY), insulin/nansferin/selenium mix (Sigma) and antibiotic/antimycotic solution (Invitrogen). The cells were maintained at 37 °C in 5% C02 in a humidified cell culture incubator and passaged at a density of 40,000- 100,000 cells/cm2 in gelatin-coated 75 cm2 flasks. Passaged cells were plated to yield near-confluent cultures at the end of the experiment. The freshly plated cells were allowed to attach in standard growth medium for at least 72 h. For experimental treatments the cells were transferred to serum-free medium for 18-24 h before addition of the stimulatory agents. Treatment of hRVE cells with fatty acids was performed as follows. Fatty acids stocks were prepared by adding fatty acids (NuCheck Prep. Inc., 62 Elysian, MN) to charcoal treated, solvent extracted, fatty acid-free bovine serum albumin (Serologicals Inc., Norcross, GA) in serum-free media to the final concentration of 100 mM fatty acid, 60 M BSA, stabilized with vitamin E (400 uM) and BHT (0.04%) and neutralized with NaOH, as described previously[36]. The fatty acid stock solutions were diluted into serum-free medium to give fatty acid concentrations of 10-100 M with corresponding BSA concentrations of 2-20 uM. The fatty acid-to-albumin molar ratio was 5:1[36]. Cells were incubated for the times indicated in the Results. Equivalent amounts of BSA alone were added to control plates. Inhibitors of COX, LOX (Cayman Chemical, Ann Arbor, MI) and MOX (Sigma) were added to the cells at the time of addition of the fatty acids. For hyperglycemia experiments the cells were incubated in normal (5.5 mM) or high (22 mM) glucose for 24 h. 3.3. Electrophoresis and immunoblotting hRVE cells were grown in 6-cm plates in experimental media for up to 24 hrs. Each plate was rinsed twice with 3 ml of ice-cold phosphate-buffered saline (PBS) containing 130 mM NaCl, 8.2 mM Nazl-IPO4, and 1.8 mM NaHzPO4 (pH 7.4). The cells were harvested in 100-300 pl of the lysis buffer (50 mM HEPES (pH 7.5), 150 mM NaCl, 1.5 mM MgC12, 1 mM EGTA, 1% Triton X-100, 10% glycerol) with the freshly added protease and phosphatase inhibitors (1 mM sodium orthovanadate, 0.15 U/ml aprotinin, and 100 pg/ml PMSF). Homogenates were centrifuged at 13,000rpm for 15 min at 4°C. Proteins were fractionated by electrophoresis on SDS-polyaerylarnide (10%) nrini-gels. The separated proteins were electrophoretically transferred to nitrocellulose (BioRad, Hercules, CA) and blocked for 60 min at room temperature in Tris-Buffered Saline (TBS) 63 (130 mM NaCl, 100 mM Tris/HCl, pH 7.5) containing 5% powdered milk and 0.1% Tween-20. The membranes were then probed overnight at 4°C in a blocking buffer containing antibody against VCAM-l, ICAM-l or E-selectin (rabbit polyclonal antibodies, Santa Cruz Biotechnology, Santa Cruz, CA), followed by anti-rabbit horseradish peroxidase conjugate (BioRad, Hercules, CA). Immunoreactive bands were visualized by enhanced chemiluminescence using an ECL kit (Amersham, Piscataway, NJ). Blots were quantitated by scanning densitometry using ImageJ software (version 1.29). 3.4. Leukocyte adhesion assay U937 cells were labeled with 2 pM 2',7'-bis-(2-carboxyethyl)-5-(and-6)- carboxyfluoreseein acetoxymethyl ester (BCECF-AM, Molecular Probes, Eugene, OR) at 37°C for 30 min, washed twice with fresh RPMI-1640 media with 10% fetal bovine serum and 1 mM L—glutamine. The fluorescent U937 cells were then concentrated to 106 cells/m1 by centrifugation and resuspended in hRVE culture medium without fetal bovine serum. For the adhesion assay, control or treated hRVE cells in 6 well plates were washed with PBS (23°C) followed by addition of 106 fluorescent U937 cells to each well. The cells were then incubated at 23°C on a rotating plate for 2 hrs. The hRVE cells were carefully washed with PBS by decanting and aspirating twice, lightly fixed with 0.5% paraformaldehyde for 15 mins at 23°C and washed with PBS two more times. Adherent florescent U937 cells were directly counted with a fluorescent microscope. For each experiment the total number of fluorescent cells was obtained from 12 to 14 random fields containing a full monolayer of hRVE cells. 64 3.5. Statistical Analysis Data were expressed as meaniSEM. Repeated-measures ANOVA was used for comparison of multiple values obtained from the same plate, factorial ANOVA was used for comparing data obtained from 2 independent samples. The Bonferroni procedure was used to control type I error. Significance was established at P<0.05. 65 4. Results 4.1. Polyunsaturated n6 fatty acids induce inflammatory adhesion molecule expression in hRVE cells We have tested the hypothesis that fatty acids typically found in blood affect the expression of adhesion molecules in hRVE cells. Accordingly, hRVE cells were exposed to 100 1.1M of BSA bound fatty acids (5 mol fatty acid to 1 mol of BSA) for 12 and 24 hrs followed by immunoblot analysis of ICAM-1 and VCAM-1 expression. Treatment with BSA alone was used as a control. Saturated palmitic (16:0) acid and n3 PUFA docosahexaenoic (22:6,n3) acid treatment had no effect on either VCAM-l or ICAM-l expression in hRVE cells (Fig. 1A and B). In contrast, treatment with n6 PUFA linoleic (18:2,n6) or arachidonic (20:4,n6) acid increased VCAM-l (Fig. 1A and C) and ICAM-1 (Fig. 1A) expression. ICAM-1 and VCAM-1 expression increased after 12 hrs and was maximally induced after 24 h of fatty acid treatment (Fig. 1A). The induction of VCAM- 1 was seen with as little as 10 uM linoleic acid and arachidonic acid (Fig. 1C). While ICAM-l was detectable with no fatty acid treatment, VCAM-1 expression was usually below detectable levels (Fig. 1A, lane 1). The maximal fold induction of VCAM-1 (4- fold) was greater than for ICAM-l (3-fold) (Fig. 1A, lanes 2, 6 and 8 and B). Fig. 18 shows a quantitation of the results obtained from 3 separate donor cells at the 4 to 6 passage stage. 66 4.2. Hyperglycemia does not affect CAM expression in hRVE cells As hyperglycemia is a key metabolic disorder of diabetes, we next examined the effect of high glucose on CAM expression in hRVE cells. In contrast to fatty acid treatment, exposure to high glucose (22 mM) for 24 h did not affect ICAM-1 and VCAM-1 expression (Fig. 2A). As a control ILl-B induced VCAM-1 expression under euglycemic conditions (Fig. 2A). 4.3. Induction of inflammatory adhesion molecules by cytokines and PMA in hRVE cells Since little is known about CAM expression in hRVE cells, we compared the effects of fatty acids with those obtained following treatment with inflammatory cytokines (TNFa, IL-IB) and the PKC activator, PMA (Fig. ZB). Both cytokines induced strong expression of the ICAM-1 and VCAM-l molecules in the time range of 6 to 24 h. In contrast, PMA treatment had little effect on CAM expression in the time periods studied. Interestingly, the expression of E-selectin in hRVE cells was not affected by these stimulants (Fig. ZB). 4.4. HUVEC do not respond to fatty acid treatment HUV EC are a primary cell culture derived from umbilical cord veins widely used as a model for human vascular endothelial cells. Like the hRVE cells, treatment of HUVEC cells with TNFa and IL-lB increased ICAM-1 and VCAM-1 expression (Fig. 3A, lanes 3 and 7). However, in sharp contrast to the hRVE cells, treatment of HUVECs 67 with fatty acids under the same conditions as used for hRVE cells resulted in no induction of VCAM-1 or ICAM-l (Fig. 3B). The significant differential sensitivity to fatty acids may represent a fundamental difference between the responses of hRVE cells and other endothelial cells to inflammatory stimuli. 4.5. Inhibition of fatty acid oxidation suppresses fatty acid-induced adhesion molecule expression in hRVE cells Experiments in progress have shown that radioactive tracer labeled fatty acids taken up from the media by hRVE cells entered a number of metabolic pathways, including esterification and elongation, leading to changes in the intracellular neutral and polar lipid pools. A high percentage of fatty acids remained in NEFA pool providing a possible substrate for oxygenases (data not shown). Since both linoleic and arachidonic acids are precursors of inflammatory mediators such as leukotrienes, thromboxanes and prostaglandins, a possible mechanism for induction of adhesion molecules by fatty acids involves lipid oxidation. This hypothesis was addressed by use of specific inhibitors of the COX, LOX and MOX pathways. Nordihydroguaiaretic acid (NDGA, a general LOX inhibitor; ICso = 3-5 uM[37, 38]) at 5 pM attenuated 18:2,n6 and 20:4,n6 induced ICAM- 1 (not shown) and VCAM-1 (Fig. 4A, lane 3 compared to 2) expression by >80%. Higher doses of NDGA completely inhibited ICAM-l (not shown) and VCAM induction (Fig. 4A, lane 4, 5 and 6). In contrast to NDGA, flurbiprofen, a general COX inhibitor (ICso = 0.04 pM, COX-1 and 0.51 uM, COX-2[39]), at 5 to 10 M had no effect on fatty acid-induced ICAM-l (not shown) and VCAM-1 expression (Fig. 4B, lanes 5 and 6 compared to 2). Only at higher, non-specific, concentrations (SO-100 uM) did 68 flurbiprofen inhibit the response (lanes 7 and 8). The specific COX-2 inhibitor NS-398 (IC50=1.77uM)[39] had no effect on the fatty acid-mediated induction of VCAM-1 (Fig. 4B, lanes 3 and 4). We also tested the effect of the MOX P450 inhibitor, l-ABT, on fatty acid induced adhesion molecule expression. l-ABT (500 M) had only a small effect on fatty acid induced ICAM-l (not shown) and VCAM-1 expression (Fig. 4C). Taken together the data strongly implicate the LOX pathway as a requirement for 18:2,n6 and 20:4,n6-mediated induction of adhesion molecules in hRVE cells. 4.6. Leukocyte adhesion correlates with fatty acid induction of inflammatory CAMs A leukocyte adhesion assay was used to confirm that fatty acid induced CAMS were functionally expressed (Fig. 5). Human U937 cells with monocytic properties40 were added to fatty acid, high glucose, or IL-1 [3 treated and untreated monolayers of hRVE cells. Treatment of hRVE cells with 20:4,n6 and 18:2,n6, or with IL-lB resulted in a significant increase of adherent cells as compared to BSA control. In contrast, high glucose or 16:0 treatment did not significantly increase the number of adherent cells. High glucose and 16:0 do not induce CAM (Fig. 1 and 2). The dramatic increase in adhesion correlates well with increased CAM expression observed by immunoblotting. 69 A“ palmitic linoleic arachidonic Treatment control 16:0 18:2,n6 22:6,n3 Time (hrs) 12 24 12 24 24 —J-B— ' VCAM-1 - - 7": *4" lCAM-1 ~ — Actin 1 2 3 4 5 6 7 8 9 B. 3 5 * ":7: m 4 - g ii if a o 3‘ * a: .s ‘T 'o 2 “ E B 5 LL 1 a > 0— Treatment 0 Treatment (1M) 0 10 50 100 18:2,n6 In m—_ VCAM-1 9.5-. 20=4m6 it ~ 2 :- si- VCAM-1 70 Fig. 1. Induction of endothelial cell adhesion molecules by free fatty acids in hRVE cells. (A) Treatment of hRVE cells with BSA control or with 100 pM of the free fatty acids (16:0, 18:2n6, 20:4n6 and 22:6n3 ) was performed for the indicated time before analysis of cell lysates for VCAM-1 and ICAM-1 expression by immunoblot (IB). (B) Quantitative compilation of the data on VCAM-1 induction in hRVE cells after 24 hrs of treatment with 100 M of free fatty acids. The results are obtained from the cells isolated from three independent donors. *P<0.05 compared to BSA control. (C) Induction of VCAM-1 in hRVE cells after treatment with different doses of free fatty acids (18:2,n6 and 20:4,n6). A and C are representative results from one donor. Equal amounts of protein were added to each lane and confirmed by probing for actin (representative loading control shown in A). Results are representative samples from greater than three independent experiments. 71 Fig.2 A, Treatment glucose (mM) ”-1 5.5 22 5.6 22 ’B C i r QlICAM-1 ' ' *L_E£VCAM-1 3. Treatment ctrl _T___NFa PMA lL-1B hours6 24624 624 6 24 IB g E S H VCAM—1 , fig , .3 i lCAM-1 I: _M 1! E-se'ect'" M actin 12345678 Fig. 2. Evaluation of cell adhesion molecule expression after treatment of hRVE cells with hyperglyeenric conditions, cytokines and PMA. The induction of adhesion molecules was assessed by immunoblot analysis (IB) after treatment of cells for the indicated times with each potential stimulant. (A) hRVE cells cells were grown in euglycemic (5.5 mM) or hyperglycemic (22 mM) conditions for 24 hours before analysis. ILl-B (5 ng/ml) was used as a positive control. (B) hRVE cells were stimulated for the indicated times with TNFa (20 ng/ml), PMA (10 ng/ml) or IL-IB (5 ng/ml) followed by immunoblot. For all experiments equal amounts of protein were loaded to each lane and loading was confirmed by probing for actin (shown in B). 72 Fig.3 A, HUVEC Treatment none TNFa _P__MA lL-1E Time (hrs) 24 6 24 6 24 6 24 m:- - g— VCIAM- 1 .59.. H £531. 2 .4 k; lCAM-1 _ — - i g a E-selectln 1 2 3 4 5 6 7 3, Fatty acid treatment (HUVEC) none 16:0 18:2,n6 20:4,n6- IB VCAM-1 2T— 3 4 £1 eic- .=-_ 1 Fig. 3. Fatty acid treatment of HUVEC cells fails to induce endothelial cell adhesion molecule expression. (A) HUV EC cells were treated with TNFa (20 ng/ml), PMA (10 ng/ml) and IL-IB (1 ng/ml) for the indicated times followed by an analysis of VCAM-1, ICAM-1 and E-selectin expression. (B) HUVEC cells were treated with BSA control (none) and with 100 uM fatty acids for 24 h time before analysis of cell lysates for VCAM-1 and ICAM-1 expression. No VCAM-1 was detected and no induction of ICAM-1 was observed in greater than three experiments. Equal amounts of protein were loaded to each lane. 73 Fig.4 A. fatty acid .. + + + + + NDGA (11M) - - 5 1o 50 100 IB 18:2,n6 u—I VCAM-1 20:4,n6 - ~ . «4 . I VCAM-1 ~ ran.“ __ - _.... B. fatty acid - + + + + + + + Flurbiprofen (pM) - - - - 5 10 50 100 NS-398 (uM) - - 5 10 - ‘ - .. - 18:2,n6 ' hits—- a... _ g... ._,.. g VCAM-1 20:4,n6 E - — - - i . VCAM-1 C. Treatment BSA 18:2,n6 20:4,n6 1-ABT (500 W) I - f + - -_ _ _ + _ - - + A .m Egg-Q VCAM-1 1 2 3 4 5 6 Fig. 4. Induction of cell adhesion molecule expression by fatty acids is inhibited by LOX, but not COX and MOX inhibitors. (A) hRVE cells were treated with 18:2,n6 or 20:4,n6 (100 M) for 24 h in the presence of increasing amounts of the inhibitor, NDGA, followed by lysis and immunoblot analysis of VCAM-1. (B) hRVE cells were treated with 18:2,n6 or 20:4,n6 for 24h in the presence of increasing amounts of either Flurbiprofen or NS-398, followed by lysis and immunoblot analysis of VCAM-1. (C) hRVE cells were treated with 18:2,n6 or 20:4,n6 for 24 h in the presence of 500 M 1- ABT followed by lysis and immunoblot analysis of VCAM-1. See methods for details of the treatment. Equal amounts of protein were added to each lane. 74 Fig.5 treatment .. 1.,_8:2,nQ 20:4,n6 '.o VIDA?“ .3 ‘ " V- 1 BSA .‘mt «1v . 7 . . $ch 828883‘ lL-1 c-z {I 52"“ r. s. P-.-., .. , . ’ _ t 1 00X $0 6““ ,‘ .6 l v ‘, . . . as - - . m w w o . 2&8 «=8 RS 3233. u . I. n .. .T,.1 K... mecca 82883:»: B 75 Fig. 5. Increased adhesion of leukocytes after induction of cell adhesion molecule expression by fatty acids. hRVE cells were treated with either BSA; 18:2,n6; 20:4,n6; IL-lB; 16:0; or 22 mM glucose for 24 hours followed by addition of fluorescent-tagged U937 cells (see Methods). (A) The number of adherent U937 cells was determined by fluorescent microscopy. A representative fluorescent and phase contrast field is shown for different treatments (all photographs are at 100X). (B) The number of fluorescent U937 cells from randomly selected microscope fields exhibiting a monolayer of hRVE cells was detemrined. The mean and SEM was determined for 12 to 14 fields for each treatment from four independent experiments with hRVE cells from three different donors. *P<0.05 compared to BSA control. 76 5. Discussion Retinal microvascular damage in early stage diabetic retinopathy has been proposed to be the result of a low-grade chronic inflammatory condition involving endothelial attachment and transmigration of leukocytes[1-3]. Support for this view is provided by the finding that high doses of aspirin are associated with decreased severity of diabetic retinopathy in humans[41] and that a marked increase in leukocyte density and retinal vascular ICAM-l immunoreactivity was found in human eyes with diabetic retinopathy[6]. In addition, it was recently demonstrated in a dog model that aspirin prevented certain classic histopathological features of diabetic retinopathy, including acellular capillary formation, retinal haemorrhage development, and an indicator of cell degeneratrion, capillary sudanophilia[42]. In rodent models anti-inflammatory agents suppressed diabetic retinal ICAM-l expression, leukocyte adhesion, and blood-retinal breakdown[S]. However, the molecular steps linking the diabetic state to retinal ICAM-l expression are not well understood. These causative events could be different in humans compared to animal models. As early inflammatory changes in human eyes do not have any clinical manifestations, human primary cell culture provides an important model to study diabetes-induced low-grade inflammation in human retina. To investigate the mechanism(s) leading to inflammation in human microvessels we performed experiments with primary human retinal vascular endothelial cells. We first considered hyperglycemia and dyslipidemia, which are two major metabolic disorders of diabetes, as likely contributors to the pathogenesis of retinopathy. We have found no evidence to support a direct causative relationship between hyperglycemia and 77 inflammatory effects in hRVE cells. Instead, our studies point to dyslipidemia as an important contributor to inflammatory events. Diabetic dyslipidemia is the result of an imbalance of the complex regulation of fatty acid uptake, metabolism, release by adipocytes and clearance from circulation. Insulin inhibits adipocyte hormone-sensitive lipase and activates lipoprotein 1ipase[43, 44]. In liver, insulin stimulates conversion of fatty acids to triglycerides followed by secretion as VLDL, as well as the induction of delta-5, delta-6 and delta-9 desaturases[9- 12, 14, 17-19, 44-46]. Thus, insulin resistance in Type 2 diabetes and low portal insulin levels in Type 1 diabetes would be predicted to have a profound effect on plasma fatty acid levels and composition. Indeed, Type 2 diabetes is characterized by an elevation of blood levels of cholesterol, esterified and non-esterified fatty acids[12, 14, 15, 45, 47-50], and Type 1 diabetes causes marked changes in FFA profile with an increase in n6 PUFA/n3 PUFA ratio[l7]. In our experiments we modeled dyslipidemia by exposure of hRVE cells to n6 PUFA, linoleic and arachidonic acids. Treatment of hRVE cells with either linoleic or arachidonic acid led to a robust increase in VCAM-1 and ICAM-1 expression. The effect was specific for these n6 PUFA, as other fatty acids tested, such as saturated pahnitic (16:0), and n3 PUFA docosahexaenoic (22:6, n3) failed to yield a response. Human plasma contains substantial amounts of linoleic (30%) and arachidonic (8%) acid in the triglyceride and FFA pools[51]. As total FFA levels in diabetes are 2 600 uM[48, 49], the concentrations of FF As used in this study (100 uM) are comparable to the concentrations expected in diabetic patients. 78 Both linoleic and arachidonic acids are precursors of inflammatory mediators including leukotrienes, thromboxanes and prostaglandins, as well as other bioactive lipid mediators, such as hydroxyl- and epoxy fatty acids. Our inhibitor studies indicate that the lipoxygenase, but not the cyclooxygenase or P450 monooxygenase pathways may be important for the fatty acid-mediated induction of adhesion molecules in hRVE cells. This conclusion is based on the fact that a LOX inhibitor (NGDA) at specific concentrations was effective at blocking the PUFA-mediated induction of CAM expression. LOXs are a diverse family of nonheme ferroproteins that catalyze the hydroperoxidation of polyunsaturated fatty acids. Thus far, six LOXS have been identified in humans: 12-LOX (platelet type), 12(R)-LOX, 15-LOX—l, 15-LOX-2, e- LOX-3, and 5-LOX[52]. LOX products, such as the hydroperoxyeicosatetraenoic acids (HPETE), hydroxyeicosatetraenoic acids (HETE) and their metabolites the leukotrienes, play a role in inflammation, especially in modulating cell-cell interactions in human aortic endothelial cells[53]. 12-LOX activity and expression were highly increased in a diabetic pig model[20]. With regard to the LOX pathway involving 18 :2,n6 and 20:4,n6 mediated induction of CAMS, it is not clear if the exogenous fatty acid, per se, is the substrate for this reaction or whether exogenous fatty acids stimulate other mechanisms to generate a substrate for LOX action. Such mechanisms might involve activation of phospholipase A2 or membrane remodeling resulting in release of substrates for LOX action. How LOX products lead to CAM expression is also unknown. Both detailed fatty acid metabolism and signaling studies will be required to define the metabolic pathway involved in PUFA regulation of CAMS. Among the known factors in the transcriptional regulation of VCAM-1 and ICAM-1 by inflammatory cytokines are NFKB, interferon 79 regulatory factor (IRF)-1, Spl and others[54-57]. How these known transcriptional regulators contribute to fatty acid stimulation is currently under investigation. ICAM-1 and VCAM-1 play an important part in the rolling and attachment of leukocytes to endothelial cells and in normal herneostatic processes. VCAM-1 is not generally constitutively expressed on endothelial cells and is induced significantly following treatment with inflammatory ligands, which include LPS, TNFa and IL-lB. Induction of VCAM-1 is a critical event in the adhesion and diapedesis of leukocytes resulting in localized inflammation. Our results suggest that fatty acids may also be an important component of the inflammatory process. Although ICAM-l is constitutively expressed on hRVE cells, n6-PUFA also promoted a greater level of ICAM-1 expression. Vascular endothelial ICAM-l is associated with adhesion and transmigration of leukocytes in the retina[2, 3] and in other vascular systems. Importantly, leukocyte infiltration and expression of retinal vascular [CAM-1 coincide with many of the pathological lesions in diabetic retinopathy[3, 6]. In our study the physiological relevance of the induction of CAMS was confirmed by performing adhesion assays using hRVE cells and fluorescent—tagged U937 cells. U937 cells were derived from a human histiocytic lymphoma and retain properties of monocytic cells[40]. The adhesion of U937 cells to vascular endothelial cells is primarily dependent on 01.4131 integrin interacting with endothelial VCAM-1[58, 59]. The fact that fatty acid treatment induced a large increase in the number of adherent U937 cells strongly supports the notion that 18:2,n6 and 20:4,n6 play an important role in microvascular inflammation. 80 A significant observation of our study is the finding that while hRVE cells were sensitive to n6-PUFA augmentation of CAM, these same n6-PUFAs did not induce CAM levels in HUV EC cells. However, both hRVE and HUVEC cells were fully capable of producing VCAM-l and ICAM-1 after treatment with cytokines such as TNFa or IL-lB. HUVEC cells are a primary cell culture derived from umbilical cord veins and represent a macrovascular system. It will be important to determine whether other microvascular and aortic macrovascular cells exhibit higher sensitivity to fatty acids compared to umbilical vein endothelial cells. In conclusion, our data suggest that diabetic dyslipidemea serves as inflammatory stimulus to initiate and contribute to microvascular complications. This model includes the increase in total lipid in Type 2 diabetes and the shift in fatty acid profile with an increase in n6-PUFA in Type 1 diabetes. N6-PUFAs are known substrates of LOX (and COX) pathways that lead to production of an array of oxidized lipids and bioactive metabolites. Based on our results we propose that chronic exposure of hRVE cells to elevated n6-PUFA associated with the diabetic condition results in longstanding chronic inflammation that gradually progresses to retinopathy. Note: This part of research has been published at Invest Ophthalmol Vis Sci. 2003 Nov;44(11):5016-22. 81 6. References 10. 11. 12. Sehroder S, Palinski W, Schmid-Schonbein GW. Activated monocytes and granulocytes, capillary nonperfusion, and neovascularization in diabetic retinopathy. Am J Pathol. 1991;139(1):81-100. Joussen AM, Poulaki V, Qin W, et al. Retinal vascular endothelial growth factor induces intercellular adhesion molecule-1 and endothelial nitric oxide synthase expression and initiates early diabetic retinal leukocyte adhesion in vivo. Am J Pathol. 2002;160(2):501-9. Adamis AP. Is diabetic retinopathy an inflammatory disease? Br J Ophthalmol. 2002;86(4):363-5. Joussen AM, Murata T, T sujikawa A, Kirchhof B, Bursell SE , Adamis AP. Leukocyte-mediated endothelial cell injury and death in the diabetic retina. Am J Pathol. 2001;158(l):147-52. J oussen AM, Poulaki V, Mitsiades N, et al. Nonsteroidal anti-inflammatory drugs prevent early diabetic retinopathy via TNF-alpha suppression. Faseb J. 2002;l6(3):438-40. McLeod DS, Lefer DJ, Merges C , Lutty GA. Enhanced expression of intracellular adhesion molecule-1 and P-selectin in the diabetic human retina and choroid Am J Pathol. 1995;147(3):642-53. Miyarnoto K, Khosrof S, Bursell SE, et al. Prevention of leukostasis and vascular leakage in streptozotocin-induced diabetic retinopathy via intercellular adhesion molecule-1 inhibition. Proc Natl Acad Sci U S A. 1999;96(19):10836-41. Lu M, Amano S, Miyarnoto K, et al. Insulin-induced vascular endothelial growth factor expression in retina. Invest Ophthalmol Vis Sci. 1999;40(13):3281-6. Goldberg RB. Lipid disorders in diabetes. Diabetes Care. 1981 ;4(5):561-72. Goldberg RB , Capuzzi D. Lipid disorders in type 1 and type 2 diabetes. Clin Lab Med. 2001 ;21(1): 147-72, vii. Goldberg U. Clinical review 124: Diabetic dyslipidemia: causes and consequences. J Clin Endocrinol Metab. 2001;86(3):965-71. Poisson JP. Essential fatty acid metabolism in diabetes. Nutrition. 1989;5(4):263- 6. 82 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Rossetti L , Goldberg U. A new piece in the diabetes puzzle. Nat Med. 2002;8(2):1 12-4. Vessby B. Dietary fat and insulin action in humans. Br J Nutr. 2000;83 Suppl 1:S9l-6. Zenobi PD, Holzmann P, Glatz Y, Riesen WF , Froesch ER. Improvement of lipid profile in type 2 (non-insulin-dependent) diabetes mellitus by insulin-like growth factor I Diabetologia. 1993;36(5):465-9. McGarry JD. What if Minkowski had been ageusic? An alternative angle on diabetes. Science. 1992;258(5083):766-70. Decsi T, Minda H, Hermann R, et al. Polyunsaturated fatty acids in plasma and erythrocyte membrane lipids of diabetic children Prostaglandins Leukot Essent Fatty Acids. 2002;67(4):203-10. Brown J E, Lindsay RM , Riemersma RA. Linoleic acid metabolism in the spontaneously diabetic rat: delta6-desaturase activity vs. product/precursor ratios. Lipids. 2000;35(12):1319-23. Brenner RR, Bemasconi AM , Garda HA. Effect of experimental diabetes on the fatty acid composition, molecular species of phosphatidyl-choline and physical properties of hepatic microsomal membranes. Prostaglandins Leukot Essent Fatty Acids. 2000;63(3):167-76. Natarajan R, Gerrity RG, Gu JL, Lanting L, Thomas L , Nadler J L. Role of 12- lipoxygenase and oxidant stress in hyperglycaemia-induced acceleration of atherosclerosis in a diabetic pig model. Diabetologia. 2002;45(1):125-33. Sen K, Misra A, Kumar A , Pandey RM. Simvastatin retards progression of retinopathy in diabetic patients with hypercholesterolemia. Diabetes Res Clin Pract. 2002;56(1):1-1 l. Klein R, Marino EK, Kuller LH, et al. The relation of atherosclerotic cardiovascular disease to retinopathy in people with diabetes in the Cardiovascular Health Study. Br J Ophthalmol. 2002;86(1):84-90. Ferris FL, 3rd, Chew EY , Hoogwerf BJ. Serum lipids and diabetic retinopathy. Early Treatment Diabetic Retinopathy Study Research Group. Diabetes Care. l996;19(l 1):1291-3. Jump DB. Dietary polyunsaturated fatty acids and regulation of gene transcription Curr Opin Lipidol. 2002; 1 3(2): 1 55-64. 83 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Jump DB. The biochemistry of n-3 polyunsaturated fatty acids. J Biol Chem. 2002;277(11):8755-8. Selhnayer A, Hrboticky N , Weber PC. Lipids in vascular firnction. Lipids. 1999;34 Suppl:S13-8. Selhnayer A , Koletzko B. Long-chain polyunsaturated fatty acids and eicosanoids in infants--physiologieal and pathophysiological aspects and open questions. Lipids. 1999;34(2): 199-205. Hwang D. Fatty acids and immune responses--a new perspective in searching for clues to mechanism Annu Rev Nutr. 2000;20:431-56. Dubois RN, Abramson SB, Crofford L, et al. Cyclooxygenase in biology and disease. Faseb J. l998;12(12):1063-73. Laneuville O, Breuer DK, Xu N, et al. Fatty acid substrate specificities of human prostaglandin-endoperoxide H synthase-1 and -2. Formation of 12-hydroxy-(9Z, 13E/Z, lSZ)- octadeeatrienoic acids from alpha-linolenic acid J Biol Chem. 1995;270(33):19330-6. Malkowski MG, Thuresson ED, Lakkides KM, et al. Structure of eicosapentaenoic and linoleic acids in the cyclooxygenase site of prostaglandin endoperoxide H synthase-1. J Biol Chem. 2001;276(40):37547-55. Grant MB, Caballero S , Millard WJ. Inhibition of IGF-I and b-FGF stimulated growth of human retinal endothelial cells by the somatostatin analogue, octreotide: a potential treatment for ocular neovascularization Regul Pept. l993;48(1-2):267- 78. Grant MB , Guay C. Plasminogen activator production by human retinal endothelial cells of nondiabetic and diabetic origin. Invest Ophthalmol Vis Sci. l991;32(l):53-64. Grant MB, Tamuzzer RW, Caballero S, et al. Adenosine receptor activation induces vascular endothelial growth factor in human retinal endothelial cells. Circ Res. 1999;85(8):699-706. Busik JV, Olson LK, Grant MB , Henry DN. Glucose-induced activation of glucose uptake in cells from the inner and outer blood-retinal barrier. Invest Ophthalmol Vis Sci. 2002;43(7):2356-63. Jump DB, Clarke SD, MacDougald O , Thelen A. Polyunsaturated fatty acids inhibit S14 gene transcription in rat liver and cultured hepatoeytes. Proc Natl Acad Sci U S A. 1993;90(18):8454-8. 84 37. 38. 39. 40. 41. 42. 43. 45. 46. 47. 48. Hope WC, Welton AF, Fiedler-Nagy C, Batula-Bemardo C , Coffey JW. In vitro inhibition of the biosynthesis of slow reacting substance of anaphylaxis (SRS-A) and lipoxygenase activity by quercetin Biochem Pharmacol. 1983;32(2):367-71. Argentieri DC, Ritchie DM, Ferro MP, et al. Tepoxalin: a dual cyclooxygenase/5- lipoxygenase inhibitor of arachidonic acid metabolism with potent anti- inflammatory activity and a favorable gastrointestinal profile. J Pharmacol Exp Ther. 1994;271(3):1399-408. Barnett J, Chow J, Ives D, et al. Purification, characterization and selective inhibition of human prostaglandin G/H synthase 1 and 2 expressed in the baculovirus system. Biochim Biophys Acta. 1994;1209(1):130-9. Ralph P, Moore MA , Nilsson K. Lysozyme synthesis by established human and murine histiocytic lymphoma cell lines J Exp Med. 1976;143(6):1528-33. Powell EDU. Diabetic retinopathy in rheumatoid arthritis. Lancet. 1964;2z17-18. Kern TS , Engerman RL. Pharmacological inhibition of diabetic retinopathy: aminoguanidine and aspirin Diabetes. 2001;50(7):1636-42. Coppack SW, Evans RD, Fisher RM, et al. Adipose tissue metabolism in obesity: lipase action in vivo before and after a mixed meal Metabolism. 1992;41(3):264- 72. Weinstock PH, Levak-F rank S, Hudgins LC, et al. Lipoprotein lipase controls fatty acid entry into adipose tissue, but fat mass is preserved by endogenous synthesis in mice deficient in adipose tissue lipoprotein lipase. Proc Natl Acad Sci USA. l997;94(19):10261-6. The effect of aggressive versus standard lipid lowering by atorvastatin on diabetic dyslipidemia: the DALI study: a double-blind, randomized, placebo-controlled trial in patients with type 2 diabetes and diabetic dyslipidemia. Diabetes Care. 2001;24(8):1335-41. Zechner R. The tissue-specific expression of lipoprotein lipase: implications for energy and lipoprotein metabolism. Curr Opin Lipidol. 1997;8(2):77-88. Vessby B, Karlstrom B, Ohrvall M, J arvi A, Andersson A , Basu S. Diet, nutrition and diabetes mellitus Ups J Med Sci. 2000;105(2):151-60. Baldeweg SE, Golay A, Natali A, Balkan B, Del Prato S , Coppack SW. Insulin resistance, lipid and fatty acid concentrations in 867 healthy Europeans. European Group for the Study of Insulin Resistance (EGIR). EurJ Clin Invest. 2000;30(1):45-52. 85 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. Lewis GF, Carpentier A, Adeli K , Giacca A. Disordered fat storage and mobilization in the pathogenesis of insulin resistance and type 2 diabetes Endocr Rev. 2002;23(2):201-29. Laws A, Hoen HM, Selby JV, Saad MF, Haffner SM , Howard BV. Differences in insulin suppression of free fatty acid levels by gender and glucose tolerance status. Relation to plasma triglyceride and apolipoprotein B concentrations. Insulin Resistance Atherosclerosis Study (IRAS) Investigators. Arterioscler Thromb Vasc Biol. l997;17(1):64-71. Cetinkale O , Yazici Z. Early postbum fatty acid profile in burn patients. Burns. 1997;23(5):392-9. Brash AR. Lipoxygenases: occurrence, functions, catalysis, and acquisition of substrate. J Biol Chem. 1999;274(34):23679-82. Patricia MK, Kim J A, Harper CM, et al. Lipoxygenase products increase monocyte adhesion to human aortic endothelial cells. Arterioscler Thromb Vasc Biol. l999;19(11):2615-22. Ochi H, Masuda J , Girnbrone MA. Hyperosmotic stimuli inhibit VCAM-1 expression in cultured endothelial cells via effects on interferon regulatory factor- 1 expression and activity. Eur J Immunol. 2002;32(7): 1821-31. Iademarco MF, McQuillan JJ, Rosen GD , Dean DC. Characterization of the promoter for vascular cell adhesion molecule-1 (VCAM-1). J Biol Chem. 1992;267(23): 16323-9. Schindler U , Baichwal VR. Three NF-kappa B binding sites in the human B- selectin gene required for maximal tumor necrosis factor alpha-induced expression Mol Cell Biol. 1994;14(9):5820-31. van de Stolpe A, Caldenhoven E, Stade BG, et al. 12-O-tetradecanoylphorbol-l3- acetate- and tumor necrosis factor alpha-mediated induction of intercellular adhesion molecule-1 is inhibited by dexamethasone. Functional analysis of the human intercellular adhesion molecular-1 promoter. J Biol Chem. 1994;269(8):6185-92. Ramos CL, Huo Y, Jung U, et al. Direct demonstration of P-selectin- and VCAM- 1-dependent mononuclear cell rolling in early atherosclerotic lesions of apolipoprotein E-deficient mice. Circ Res. 1999;84(1 1): 1237-44. Node K, Huo Y, Ruan X, et al. Anti-inflammatory properties of cytochrome P450 epoxygenase-derived eicosanoids Science. 1999;285(5431):1276-9. 86 III. Anti-Inflammatory Effect of Docosahexaenoic Acid (DHA22;5n3) and Peroxisome Proliferator-activated Receptors (PPARS) on Cytokine Induced Adhesion Molecules Expression in Human Retinal Vascular Endothelial Cells 1. Abstract Early stage diabetic retinopathy has been recognized as a low-grade chronic inflammatory condition. As such, it is characterized by an increase in inflammatory cytokines including TNFa, IL-10 and VEGF165. Diabetes induces changes in lipid metabolism that result in a decrease in the principal n3-polyunsaturated fatty acid (PUFA) in the retina, docosahexaenoic acid (DHA22;(,n3) in diabetic eyes. DHA22:6n3 has been shown to have pronounced anti-inflammatory effect in several inflammatory models. A decrease in DHA22:6n3 in the face of increased cytokine production would be expected to further promote inflammatory response in the retinal vascular endothelial cells. The effects of cytokines and DHA22:6n3 on inflammatory response in human retinal vascular endothelial cells (hRVE) were not addressed. We report herein that VEGF165, TNFa and IL-lB caused significant induction of ICAM-1 and VCAM-1 expression in hRVE cells. Pre-treatrnent of the cells with 100 M of BSA-bound DHA22;6,,3 for 24 hours remarkably inhibited cytokine-induced ICAM-1 and VCAM-1 expression compared to BSA (carrier control) or palmitate (lipid control) treated cells. All three cytokines (IL-113, TNFa and VEGF165) induced NFKB binding to the VCAM-1 promoter in hRVE cells by activating specific NFKB isoforms: p65 and p50. DHA22;6,,3 pretreatment inhibited cytokines induced NFKB binding to the VCAM-l promoter by about 25% with IL-1 [3, and by 40% 87 with VEGF165 respectively. Moreover, DHA22:6n3 diminished IL-IB induced Icht phosphorylation thus preventing IKBa degradation compared to palmitate treated control cells. We further addressed if DHA22:6n3 exerted its anti-inflammatory role through activation of the nuclear receptor PPARs. Among three PPAR isoforms ((1, B and y) expressed in hRVE cells, only specific PPARa agonists WY14,643 and fenofibrate downregulated VEGF165 and TNFa induced VCAM-1 expression. DHA22:6n3 as well as other fatty acids all activated PPARa in hRVE cells to a similar degree. These data suggest that both DHA22:6n3 and activation of PPARor can potently suppress proinflammatory cytokine-induced adhesion molecules expression in hRVE cells. Whether activation of PPARa is the principal pathway that DHA22:6n3 acts through to inhibit proinflammatory cytokine signaling in hRVE is discussed. 88 2. Introduction Diabetic retinopathy (DR) is a leading cause of blindness in adults[1, 2]. The early stage of DR has recently been recognized as a result of chronic inflammatory conditions involving attachment and transmigration of leukocytes to the retinal microvasculature[3-5]. Several inflammatory pathways are activated at the early stage of diabetic retinopathy. The pro-inflammatory cytokines TNFa[6-9] and IL-lB[10] are found to be elevated in the extracellular matrix, endothelium, vessel walls and vitreous of eyes with proliferative diabetic retinopathy and in the retinas of rats with 2 months of diabetes. Inhibition of TNFa and IL-lB signaling with a TNFOL receptor/Fe construct[4] or ILRa[11] significantly reduced leukocyte adhesion and endothelial cell (EC) injuries. These data suggests possible roles of TNFa and IL-lB and receptors mediated signaling pathway in inducing EC injury that contributes to the development of diabetic retinopathy. Another mediator, vascular endothelial cell growth factor (VEGF), has also been strongly implicated in the pathogenesis of both background and proliferative diabetic retinopathy[12-15]. Increased intraocular VEGF levels as well as VEGF receptor 1 and 2 were detected in rat and human diabetic retina[12-20]. In addition to its well known mitogenic and angiogenic activity, VEGF was recently recognized as a proinflammatory cytokine[Zl, 22]. As such, the induction of adhesion molecues expression such as ICAM- 1, VCAM-1 and E-selectin in endothelial cells (HUVEC) and rat retina by VEGF was observed[21][22]. Specific inhibition of VEGF activity inhibited ICAM-l expression, leukocyte adhesion, blood-retinal barrier breakdown and neovascularization in STZ 89 diabetic rats[21]. However, the effect of inflammatory cytokines especially VEGF on htunan retinal endothelial cells is not well studied. Inflammatory cytokines function through their receptors to initiate a series of signal transduction events that lead to the phosphorylation and degradation of Inhibitor of nuclear factor Kappa B (IKB) followed by the translocation and activation of nuclear factor Kappa B (NF KB) in the nucleus[23]. NFKB is an important transcription factor controlling the expression of an array of inflammatory response genes including adhesion molecules. Activation of NFKB (p65 and p50) has been well documented in diabetes, especially in the retinal vasculature of diabetic patients and in animal models[21, 24]. In vitro high glucose has been shown to cause the activation of NFKB in bovine retinal endothelial cells or pericytes. However, the role of NFKB in response to inflammatory cytokines in hRVE cells awaits to be clarified. Hyperglycemia and dyslipidemia are two major metabolic disorders of diabetes mellitus. Despite considerable progress in understanding of hyperglycemia-induced pathology over the past decade, the link between diabetic metabolic disorders and retinopathy still eludes us. The role of diabetic dyslipidemia in the development of microvascular complications has received much less attention. Dyslipidemia is a major metabolic syndrome prevalent in both Type 1 and Type 2 diabetes. Type 2 diabetes is characterized by the elevation of blood levels of LDL-C, triglycerides and free fatty acids along with decrease in HDL cholesterol[25-31]. Changes in serum lipids and lipoproteins have also been observed in Type 1 diabetes mainly with a reduced level of long chain PUFA such as DHA22:6n3 in the plasma of diabetic children[32] and in the 90 diabetic human eye[33, 34]. Clinical data suggest that dyslipidenria could be a critical factor in the development of diabetic retinopathy[35][36-42]. n3-PUFAs (abundant in marine fish oils) have long been recognized to modulate the inflammatory immune response and are widely applied clinically as an adjuvant immunosuppressant in the treatment of inflammatory disorders (reviewed in [43, 44]). The decrease in the most abundant long chain n3-PUFA in the eye, DHA22:6n3, could expose diabetic eyes to a more proinflammatory environment. The anti-inflammatory role of n3-PUFAs in the regulation of inflammatory cytokines and the induction of adhesion molecules expression in hRVE has not been studied. PUFAs or their eicosanoid derivatives are natural ligands for the nuclear receptors such as PPARs[45-47]. Recent evidence has indicated that all three PPARS (a, B, y) are actively involved in regulating inflammatory responses. PPARa, B, 7 agonists have been shown to inhibit inflammatory cytokines and adhesion molecules production in a cell type and isoform specific manner[48-53]. It is known that n3-PUFAs are able to activate PPARS effectively[54]. This implies that PPARs and their regulation by n3-PUFAs may play important roles in n3-PUFA mediated anti-inflammatory responses. Herein, we demonstrate that inflammatory cytokines (TNFa, IL-lB and VEGF165) induce VCAM-1 and ICAM-1 expression in hRVE cells through activation of NFKB pathway. This induction was inhibited by treatment with n3-PUFA (DHA22:6n3) and specific PPARa agonists. DHA22:6n3 might not just target PPARa to exert the anti- inflammatory effect. In fact, DHA22:6n3 is acting upstream of IKBa phosphorylation and degradation to suppress cytokine induced NFKB signaling. These results suggest a 91 beneficial role of DHA22:6n3 and PPARa in downregulating inflammatory responses in hRVE. 92 3. Materials and Methods 3.1. Reagents DMEM and F 12 culture medium, antibiotics, fetal bovine serum, and trypsin were obtained from Invitrogen (Carlsbad, CA). Commonly used chemicals and reagents were from Sigma-Aldrich Chemical Co. (St. Louis, MO). TNFa and IL-18 were from R&D Systems (Minneapolis, MN). VEGF165 was purchased from Calbiochem. Trigolitazone (TGZ), Fenofibrate, WY14,643, GW510516 and Alzoyl-PAF were obtained from Caymen chemical (Ann Arbor, MI). 3.2. Cell culture Primary cultures of hRVE cells obtained from at least three donors were prepared and cultured as previously described[55]. Passages 1-6 were used in the experiments. Primary human umbilical vein endothelial cells (HUVEC, multiple donors) were obtained from Cascade Biologicals (Portland, OR) and cultured in DMEM containing 10% FBS, macrovascular endothelial cell growth supplement (MVGS) and 100 ug/ml penicillin/streptomycin, 100 pg/ml antimycotics in a humidified incubator at 37 °C with 5% C02. For experimental treatments, cells were transferred to serum-free medium for 18 to 24 hours before addition of the stimulatory agents. Treatment of cells with fatty acids was performed as follows. Fatty acid stocks were prepared by dissolving fatty acids (NuCheck Prep, Inc., Elysian, MN) in 100% ethanol to a final concentration of 100 mM fatty acid as described previously. The fatty acid stock solutions were diluted in serum-free medium to reach fatty acid concentrations of 100 uM with corresponding 93 bovine serum albumin (BSA) concentration of 20 uM. Charcoal-treated, solvent- extracted, fatty acid—free BSA was obtained from Serologica Inc., Norcross, GA. The fatty acid-to-albumin molar ratio was maintained at 5:1. Cells were incubated for the times indicated in the Results section. Equivalent amounts of BSA alone were added to control plates. 3.3. SDS-PAGE and western blot analysis Cells were lysed in the lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 1.5 mM MgC12, 1 mM EGTA, 1% Triton X-100, 10% glycerol) with freshly added protease inhibitor cocktail (Sigma) and phosphatase inhibitors (1 mM Na3VO4, 100 M glycerophosphate, 10 mM NaF, 1 mM Na4PPi). Proteins were resolved by SDS-PAGE and transferred to nitrocellulose, immunoblotted usingappropriate antibodies followed by secondary horseradish peroxidase conjugated antibody (Bio-Rad). Irnmunoreactive bands were visualized by enhanced chemiluminescence (ECL kit; Amersham Pharmacia Biotech, Piscataway, NJ). Blots were quantitated by scanning densitometry using ImageJ software, ver. 1.29 (available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). 3.4. Electrophoretic mobility gel shift assay The double-stranded oligonucleotide containing the NFKB binding sequence derived from human VCAM-1 promoter were designed and synthesized as follows: 5’ TGCCCTGGGTTTCCCCTTGAAGGGATTTCCCTC3’ and 94 3’GACCCAAAGGGGAAC'ITCCCTAAAGGGAGGCGG5’. The oligonucleotides were annealed and labeled in the presence of P32dCTP with the Random Primer Kit from Invitrogen used according to manufacturer’s protocol. For binding reactions, nuclear extracts (6 pg) were incubated in 25 pL of total reaction volume with 32P-labeled NFxB oligonucleotides for 20 minutes at room temperature. DNA-protein complexes were resolved on a 6% nondenaturing polyacrymide gel and the bands were examined by autoradiography. Incubation of the nuclear extracts with excess cold NF re B oligonucleotides was used to confirm the specificity of binding activity. 3.5. Real time RT-PCR Early passage hRVE (pl-p6) from three donors were cultured to 95% confluent and serum starved overnight. RNA was extracted using Trizol reagent. 1 pg RNA was reverse transcribed using oligo d(Tl8VN) and 1/20 volume of cDNA was used for real time PCR in each reaction. Real-time quantitative RT-PCR primers targeting human PPARa, 8, y and B-actin were designed by using Primer Express software (Applied BioSystems, Foster City, CA) and sequences are listed as follows: (5' to 3'): PPAR 0:: Forward: GGAAAGGCCAGTAACAATCC; Reverse: CTGGCAGCAGTGAAAGATG; PPARS: Forward: GGGCTTCCACTACGGTGTT; Reverse: TTGTTGCGGTTCTTCTTCTG; PPARy: Forward: AGCCCAAGTTTGAGTTTGCT; Reverse: AATGTCTTCAATGGGCTTCA; B-actin: Forward: CTCTTCCAGCCTTCCTTCCT; Reverse: TGTTGGCGTACAGGTCTTTG. The specificity of each primer to the sequence of choice was checked by National Center for Biotechnology Information (NCBI) Blast module. To assure the specificity of each 95 primer set, amplicons generated from PCR reactions were analyzed for specific melting point temperatures by using the first derivative primer melting curve software supplied by Applied BioSystems. The SYBR Green I assay and the ABI Prism 7700 sequence detection system (Applied Biosystems) were used for detecting real-time quantitative PCR products. PCR reactions for each sample were done in triplicates for both target gene and B-actin control. 3.6. Transfection of hRVE using lipofectamine 2000 Cells were plated in 6-well plates at 0.08x106/well. The cells were transfected in Opti-MEM using Lipofectamine 2000 (1.5 pl/pg DNA) (Invitrogen) according to the manufacturer’s instructions. pM-rPPARa-LBD was a fusion of the Gal4-DNA-binding domain fused to the ligand-binding domain of rPPARa. The TKMHlOOx4-Luc reporter contains four binding sites for the Gal4-DNA-binding domain. 24h after transfection, cells were changed to serum free media with 100 pM fatty acid (NuChek Prep, Elysian, MN), and bovine serum albumin (BSA to fatty acid ratio was 1:5) or the PPARa agonist, WY14,643. After 24 h of treatment, the cells were harvested for luciferase assays. Each treatment involved triplicate samples, and each study was repeated at least twice. The results were expressed as relative luciferase activity normalized to protein levels. 96 4. Results 4.1. TNFa, lL-1B and VEGFm induce adhesion molecules expression In hRVE cells In hRVE cells, both TNFa (5 ng/ml) and ILl-B (1 ng/ml) acutely stimulated the expression of ICAM-1 and VCAM-1 (Fig. 1A). Recombinant VEGF165 (20 ng/ml), an important angiogenesis factor in diabetic retinopathy, also induced adhesion molecules expression (both ICAM-1 and VCAM-1) in hRVE cells (Fig.lB). The induction of VCAM-1 and ICAM-1 were time dependent, with VCAM-1 expression peaking at 24 hrs and ICAM-1 expression persisting for up to 48 hrs. Less effect of cytokine stimulation on E-selectin expression was observed in the time points checked. To compare the potency of the principal cytokines such as IL-IB with VEGF165, we treated hRVE cells with VEGF165 and IL-lB at the doses used above (the commonly recommended doses in cell culture settings for each cytokine) and performed side by side western analyses. As in Fig. 1C, VEGF165 has at least one magnitude lower potency than IL-lB, suggesting VEGF165 is a weaker proinflammatory cytokine compared to the principal cytokines in hRVE cells. 4.2. DHAzmm Inhibits TNFa, lL-1B and VEGF155 induced CAM expression n3-PUFAs (EPA20;5,,3 and DHA22:6n3) have been shown to be anti-inflammatory in a number of different cell types. Here we investigated the effect of DHA22;6n3 on cytokine induced inflammation in hRVE cells. Pre-treatrnent of hRVE cells with 97 DHA22:6n3 (100 pM of BSA-bound DHA22:6n3 for 24 hours) significantly inhibited ILl-B and TNFa induced VCAM-1 expression by about 40% and 50% respectively (Fig. 2A and B). In contrast, pre-treatrnent with the saturated palmitic acid (16:0) used as a lipid control did not exhibit a significant effect on cytokine-induced VCAM-1 expression (Fig. 2A, and quantitated in B). Similarly, DHA22:6n3 pretreatment inhibited VEGF165 induced VCAM-1 and ICAM-1 expression while pahnitic acid (16:0) pretreatment had no effect (Fig. 2C). The anti-inflarmnatory effect of DHA22:6n3 was also confirmed in HUVEC cells. DHA22:6n3 inhibited VEGF165 and IL-10 induced VCAM-1 expression in a dose dependent fashion (Fig. 2D), suggesting a common role of DHA22:6n3 functioning as an anti-inflammatory agent in human endothelial cells. 4.3. NFKB is an important transcription factor regulating adhesion molecules expression in hRVE To investigate the role of NFKB in cytokine induced adhesion molecules expression in hRVE, electrophoretic mobility shift assay (EMSA) was performed. A double stranded DNA probe containing the specific NFKB binding site from human VCAM-1 promoter was used to study the activation and binding of NFKB to the promoters of adhesion molecules. As shown in Fig 3A, all three cytokines induced NFKB binding to the VCAM-1 promoter. VEGF165 induced a delayed NFKB activation in the nucleus, with the NFKB induced shifts starting from 1 h and peaking at 2 h (Fig. 3A). Moreover, two specific isoforms of NFKB family members: p65 and p50 accumulated in the nucleus upon stimulation by IL-lB and TNFa (Fig. BB). Phosphorylation of p65 at Ser 536 required for optimal transactivation of NFKB[56] was 98 also observed in the nucleus of IL-lB and TNFa stimulated cells (Fig. 38). Likewise, VEGF165 also induced a minor translocation of p50 and p65 into the nucleus (Fig. 3B) with no obvious p65 phosphorylation observed (data not shown). 4.4. DHAmm pretreatment inhibits cytokine induced NFxB binding to the VCAM-1 promoter To address the molecular mechanism underlining the inhibitory effect of DHA22:6n3, we analyzed whether DHA22:6n3 is acting through inhibiting NFch signaling to downregulate CAM expression in hRVE cells. VEGF165 induced binding to VCAM-1 promoter at 2 h was decreased about 40% by pre-treatrnent with DHA22:6n3, but not with BSA, palmitic acid (16:0) or linoleic acid (18:2n6) (Fig. 4A). Similarly, DHA22:6n3 pretreatment inhibited IL-l B induced NFKB binding to the VCAM-1 promoter by 25% compared with pahnitic (16:0) treated controls as shown in Fig. 4B and quantitated in Fig. 4C. This was concomitant with a decrease in the nuclear level of p65 and p50 in DHA22:6n3 pretreated cells (Fig. 4D), implying that DHA22:6n3 decreases IL-lB induced nuclear translocation of p65 and p50 thus inhibiting their binding to the VCAM-1 promoter. 99 4.5. DHAzzzm pretreatment inhibits lea phosphorylation and degradation, an immediate event upstream of NFxB nuclear translocation The specific step that DHA22:6n3 acts on to inhibit cytokine induced NFKB activation were further dissected by examining the upstream IKBa phosphorylation and its ubiquitin mediated proteosome degradation. DHA22:6n3 pretreatment caused an inhibition of IL-1 B induced IKBa phosphorylation compared with palmitic (16:0) treated controls at the time points checked (Fig. 5). The inhibition was correspondent to the prevention of IKBa degradation by DHA22:6n3 pretreatment. VEGF165, even at the highest dose used, was not as potent an activator of the NFKB pathway as IL-lB and TNFa. Therefore, VEGF165 induced IKBa phosphorylation and degradation was below the sensitivity level to be detected by our analyses used in this study. 4.6. Expression pattern of PPAR isoforms in hRVE by RT-PCR Since PUFAs and the eicosanoids are natural ligands for PPAR isoforms, we first wanted to characterize the expression pattern of PPAR isoforms in hRVE. Reverse transcription followed by real time PCR was utilized to compare the mRNA level of PPAR isoforms in a semi quantitative fashion. The abundance of PPAR mRNA levels was compared to B-actin copies (Fig. 6A). hRVEs express all three isoforms with PPARS appearing as the most abundant. Western blot analysis was used to verify the protein expression levels in hRVEs isolated from three donors (Fig. 6B). 100 4.7. PPARa specific agonists partially inhibit cytokine induced CAMs expression All three isoforms of PPAR have been shown to modulate immune responses[S 7]. To investigate in hRVE cells which PPAR isoform may play a role in regulating inflammatory cytokine induced immune responses, specific agonists for each isoform were used. Fig. 7 showed that specific PPARa agonist WY14,643 dose dependently attennuated VEGF165 induced both VCAM-l and ICAM-1 expression in hRVE cells, while the PPARy agonist, TGZ, had no effect. Similar results were obtained with TNFa induced VCAM-1 expression (Fig. 8) which was inhibited by specific PPARa agonists WY14,643 and Fenofibrate, but not by specific PPARS (GW510516) and PPARy (TGZ and azoyl-PAF) agonists. These results demonstrate that in hRVE cells, activation of PPARa, but not PPARS and 7 may specifically prevented the cytokine induced adhesion molecules expression. 4.8. DHAzzzm activates PPARa in hRVE To study whether DHA22:6n3 could activate PPARa in hRVE, we used the chimeric receptor Gal4-rPPARa-LBD (which shares high homology to hPPARaLBD) to assess the sensitivity of PPARa to fatty acid regulation. Accordingly, cells were transfected with Gal4-MI-ITK-luciferase together with the chimeric receptor Gal4- rPPARa-LBD plasmid. PPARa specific agonist WY14,643 induced about a 5 fold increase in Luc activity, suggesting that WY14,643 specifically targeted to PPARa in hRVE cells (Fig. 9). Interestingly, exogenous free fatty acids such as saturated pahnitic 101 acid (16:0), proinflammatory n6-PUFAs (LA13;2n6 and AA20;4,,6), together with anti- inflammatory n3-PUFA (DHA22;6,,3) all increased Luc activity to a similar level (about 2 fold). Docosapentaenoic acid (DPA22;5,,3) and EPA20;5,,3 were less potent activators for PPARa in hRVE. This implies that different free fatty acids have different affinities for PPARa ligand binding domain. However, anti-inflammatory properties of DHA22:6n3 might not be only acting through activating PPARa in hRVE. 102 A. Treatment none TNFa lL-1B r—*—\ r—-&—\ f—*—\ Time (h) 6 24 6 24 6 24 _LL High-i VCAM-1 F ‘J': 1. _a u lCAM-1 m g '3'. a E'se'e‘ii" M Adi" B. Treatment none VEGF165 Time (h) o _'B - m VCAM-1 ‘ i f- 3 lCAM-1 , -~ ‘_ ‘ ‘ E-selectin w Actin C. Treatment -- VEGF165 lL-1B IB " "' ._ VCAM-1 .1:- c... —. Actin 103 Fig. 1. Induction of cell adhesion molecules after treatment of hRVE cells with cytokine TNFa, IL-lB and VEGF165. hRVE cells were serum starved overnight and stimulated with 5 ng/ml TNFa, 1 ng/ml IL-lB (panel A), 20 ng/ml VEGF165 (panel B) for different time periods as indicated. The potency between VEGF165 (20 ng/ml) and IL-lB (1 ng/ml) were compared in panel C. The induction of adhesion molecules, VCAM-l, ICAM-1 and E-selectin was assessed by immunoblot analysis. Equal amounts of protein were added to each lane and confirmed by actin blot. Representative results from at least tree independent experiments were presented for each panel. 104 A. fatty acid - - 16:0 22:6n3 cytokine - + + + IB IL-1B — - 1 _— ‘ __ . ,VCAM-1 TNFa E! i J B. 1207 fillim ‘ TNFa 100 ~ 1 F c .g 80 ‘ g 4 E 60 ‘ * °\° 40 ‘ 20 ‘ 0 e o 0') g o "a o 59 .5 g .‘2 <5 0 N N m N C D ~59 (0'69 '93? VESFrss fatty acid (pM)- - (93" 635} if r 1 cytokine - + '3' 1‘" f9 ———'§— My?! 18 . .4 E“... VCAM-1 ‘ __ _E‘f‘? f ‘1‘"? WM“ lL-1B E! “m w ._.,... .._.. .. . - - Actin we -Il' ICAM_1 hum-1- *- «.— ~ W‘ cud VCAM-1 - VE F ”“0- ACtll‘l G 165 1 F_'“ Actin 105 Fig. 2. n3-PUFA pretreatment specifically downregulates the induction of VCAM-1 by proinflammatory cytokines. (A) hRVE cells were serum starved overnight and then cells were either untreated or treated with 100 pM 16:0 and 22:6n3 for 24 hrs. Cells were then stimulated with 1 ng/ml IL-lB, 5 ng/ml TNFa and 20 ng/ml VEGF165 for 6 h. Lysates were prepared and same amounts of protein were used for analysis by immunoblot to detect the expression of VCAM-1 and ICAM-1 (panel A and. C). Representative data were presented from at least 3 independent experiments using hRVE cells fiom different donors and quantitated as in panel B for TNFa and IL-lB. *P<0.05. (D) HUVEC cells were treated with BSA alone or BSA bound 16:0 and 22:6n3 in the presence of 1% FBS media as indicated for 24 h. Cells were then stimulated with 0.2 ng/ml IL-lB and 20 ng/ml VEGF165 for 6 h. Lysates were prepared and analyzed as above and representative data from at least three independent experiments were presented 106 A. Treatment TNFa IL-IB VEGF165 . F—~—\ F'hh l—h Time (h) o 0.5 1 0.5 0.5 o 0.5 1 o 0.5 1 2 Coldprobe - - -10x100x ‘ .1 _,_ _, .1 Ia—NFKB no. } NS ~..-m <7 -— Free probe B. TNFa (min) 0 3O 60 '3 -——-—~-- _.... p-p65 Ser535 ----w p65 --—-— p50 lL-1|3 (min) 0 30 60 *——-—--————- p - p65 Ser536 -m —— p50 VEGF165(h) o 0.5 1 2 ' - -- H' p50 ...... p65 107 Fig. 3. Inflammatory cytokines activate NFKB signaling to induce adhesion molecules expression in hRVE cells. (A) hRVE cells were serum starved overnight and treated with TNFa (10 ng/ml), IL-l B (1 ng/ml) and VEGF165 (20 ng/ml) as indicated. Nuclear extracts were prepared. EMSAs were performed using probes containing specific NFch binding motif to the human VCAM-1 promoter. Arrows indicate the NFKB induced shift, NS stands for the nonspecific band. The specific NFKB induced shift band was confirmed by adding cold probes to compete away the labeled probe. (B) Equal amounts of nulear extracts were loaded for western blot analyses against p65, p-p65Se 36 and p50 for their transport into nucleus. Representative results were presented from at least three independent experiments. 108 VEGFm a. lL-1B - + + 4 m BSA 16:0 22:6n3 * ~er 34 NS —’ m 1: BSA 2: BSA 3: 16:0 4: 18:2 5: 22:6 ~12“ IL-1B stimulation D. g... I51 3 ' . . F i001 4: cont. 16.0 22.603 '3 g 00* ——-—- p50 2 401 ° 16:0 22:6n3 Fig. 4. DHA22:6n3 inhibits VEGF165 and IL-1 B induced NFKB signaling. (A) hRVE cells were serum starved overnight and treated with 100 pM BSA bound 16:0, LAlsgnfi and DHAum for 24 h. Cells were then stimulated with 20 ng/ml VEGFm for 2 h and nuclear extracts were prepared. EMSAS were performed as before. (B) hRVE cells were treated with 100 pM BSA bound 16:0 and 22:6n3 for 24 h and then stimulated with lng/ml IL-lB for 30 min followed by EMSA. Arrows indicate the NFKB induced Shift, NS stands for the nonspecific band. Representative results from three independent experiments were shown and statistic analyses were performed as panel C with p<0.005. (D) The same amounts of nuclear extracts from B were analyzed against p65 and p50 by western blot. 109 lL-1B(min) 0 2 5 10 2 5 10 IB Fig. 5. Inhibition of IL-lB induced Icha. phosphorylation and degradation by DHA22:6n3 pretreatment in hRVE. hRVE cells were serum starved ovemight and treated with 100 pM BSA bound 16:0 and DHA22:6n3 for 24 h. Cells were then stimulated with 1 ng/ml IL-l B for the indicated time periods and harvested. Western blots were performed to detect the IKBct phosphorylation and degradation. Representative data were presented fiom at least 3 independent experiments. 110 PPARa PPARS PPARy DNA r—H r—*—\ [s-actin ladder . 200 bp 300 bp £5000 s £4000 1 33000 copies of mRNAHo‘O O PPARa PPARd PPARQ 3' Donor #1 #2 #3 F—k—‘fif ‘ lB ~Uu... “Hf-v6. PPARa .5. me ._ ‘ '1“ “W” PPAR‘Y 111 Fig. 6. Relative mRNA abundance and protein expression of PPAR isoforms in hRVE. (A) hRVE cells were serum starved overnight and total RNAS were extracted for reverse transcription and quantitative real time PCR analyses using Specific primers against human PPARa, 5 and y. The results were calculated relative to B-actin to compare the mRNA abundance of three major PPAR isoforms averaged from hRVE cells isolated from three different donors and the amplified DNA bands for each gene were presented. (B) Same amounts of whole cell lysates of hRVE cells from three donors were applied for SDS-PAGE and blotted using antibodies against each PPAR isoform. Represent blot was shown. 112 vggr... \ r WY14,643(pM) - - 10 50100 - - - Tenn/1);“; - - - 1 5 1o — ‘F’F—vv'vvw ‘0 2333:" VCAM" . Mw-‘fi. Actin Fig. 7. PPARa agonists inhibit VEGF165 induced CAMS expression. hRVE cells were serum starved overnight and pretreated with PPARa agonist WY14,643 and PPARy agonist Trigolitazone (TGZ) as indicated for 45 min. Recombinant VEGF165 (20 ng/ml) was then added. Cells were harvested after 6 h and total lysates were prepared. Same amounts of protein were submitted for western blot analyses against VCAM-1, ICAM-1 and actin. A representative blot from 3 independent experiments was presented. 113 TNFa \ - - Feno WY GW TGZ PAF IB '9") 1...... I-gi—Ii g5 55 VCAM-1 140 1 120 - 100 1 80 - 60 - SSOfTNFahfllflnducflon 20 ~ 0 4 'HSZ PAF Feno Fig. 8. Effect of PPAR ligands on TNFa induced VCAM-1 expression in hRVE cells. hRVE cells were serum starved overnight and pretreated with PPARa agonist 300 pM Fenofibrate (Feno) and 100 pM WY14,643 (WY), PPARS agonist 1 pM GW501516 (GW), PPARy agonists 10 pM Trigolitazone (TGZ) and 1 pM Azoyl-PAF (PAF) for 45 min and stimulated with 5 ng/ml TNFa for another 6 h. Cells were harvested and the same amounts of protein were submitted for western blot analysis against VCAM-1 and actin as a loading control. A representative blot from 3 independent experiments was presented and expressed as the means 4: SD by normalizing to TNFa stimulation alone according to experiments using hRVE derived from two donors (p<0.05). 114 1 Fold change dNUhC’IO’N L 1 1 a1: 6 '5 e x Q 6‘ ‘3 6 a“ (5% <2 0 a1: is: * a1: raw-vs- street a 0 Fig. 9. Activation of PPARa by WY14,643 and exogenous free fatty acids in hRVE cells. hRVE cells were transfected with pMN-MHTK-Luc with pMN-rPPARa-LBD. After an overnight trarrsfection period, the cells were treated with or without WY14,643 (100 pM) or free fatty acids as described in Methods for 24 h. The cells were harvested for protein and luciferase assays. The results were reported as the relative luciferase activity (RLA, firefly luciferase activity/ pg protein) normalized to the untreated control. The results were expressed as the means i SD. of three separate studies with triplicate samples per group. ‘: p<0.05. 115 5. Discussion In chronic inflammatory conditions, endothelial cells actively recruit blood borne leukocytes such as monocytes and T lymphocytes to the underlying tissue in response to the activation by cytokines and growth factors. This process is mediated by the increased expression of adhesion molecules on both immune cells (leukocytes and endothelial cells). The early stage of Diabetic retinopathy has been recognized as a chronic inflammatory disease[3-5]. Upregulation of inflammatory cytokines especially TNFct[7- 9], IL-1B[10, 11] and VEGF[12-14] along with their correspondent receptors have been well documented in diabetic eyes of human subjects and animal models. However, the effect of these principal cytokines on human retinal endothelial cells adhesion molecules expression, especially VCAM-1, the Specific vascular inflammatory marker, has not been tested. In this paper, for the first time we report the effect of cytokines on adhesion molecules expression in human primary retinal endothelial cells isolated fi’om different donors. Also the role of a family of important transcription factors essential to mediate inflammatory response, NFKB, was investigated in regards to their response to different cytokines. Our work further showed the anti-inflammatory properties of the principal n3- PUFA in the retina, DHA22:6n3, on cytokine triggered inflammatory signaling. It suggests that the decrease of DHA22:6n3 in both plasma and retina in Type 1 diabetes may exacerbate the proinflammatory environment due to the elevated levels of the proinflammatory cytokines in diabetic eyes. NFKB has been suggested as a potential therapeutic target in atherosclerosis and thrombosis due to its important role in regulation of inflammatory diseases[23, 58]. 116 NFKB is involved in the development of diabetic microvascular complications. Retinal NFKB is activated in diabetes and mediates retinal capillary cell death[21, 24]. Our data suggests that retinal endothelial cells contain the cognate receptors for TNFa, IL-l B and VEGF165, and activation of these receptors could lead to the increased binding of the important transcription factor NFKB to the VCAM-l promoter in the nucleus. The major NFKB isoforms activated by the inflammatory cytokines in hRVE are p65 and p50, which usually form a classical p65/p50 heterodirner to mediate DNA binding. Previous reports demonstrated increased accumulation of only p50, but not the p65 subunit of NFKB in nuclei of retinal endothelial cells from diabetic animals[59]; and p65 was shown to be increased in retinal pericyte nuclei but not in endothelial cells from diabetic retinopathy patients and/or in cells from the STZ diabetic rat model[60]. The apparent differences could come from the different systems used. Our study using cultured human retinal endothelial cells demonstrates that both p65 and p50 are important DNA binding transcription factors that can be induced by proinflammatory cytokines in hRVE cells as a route for activation of VCAM-1 and ICAM-1 expression. Moreover, our data agrees with other reports Showing that ICAM-l is a critical adhesion molecule increased in the retinas and that VEGF165 is a proinflammatory cytokine in inducing its expression[21][22]. However, our data also suggests that VCAM-1 is also important in mediating leukastasis of the human retina since VEGF165, IL-lB and TNFor were potent inducers of VCAM-1 expression in hRVE cells. n3-PUFAS have long been recognized to modulate immune response and are widely applied clinically as adjuvant immunosuppressant in the treatment of inflammatory disorders[43, 44]. Numerous studies in various cells have shown that 117 treatment with n3-PUFAS could inhibit adhesion molecules and cytokine expression induced by inflammatory agents[43]. Our study using primary human retinal endothelial cells contributes to the evidence supporting a common role for DHA22:6n3 as an anti- inflammatory agent in ameliorating endothelial cells response to cytokines. The specific mechanisms underlying this inhibitory effect have been intensively sought for decades. Several possible mechanisms have been suggested including the displacement of AA20;4,,6, the major substrates for the synthesis of proinflammatory eicasanoids due to n3-PUFAS incorporation into membrane phospholipids[61]. Direct activation of the nuclear receptors such as PPARS is also involved. Indeed, activation of PPARa. by Specific agonists WY14,643 and F enofibrate can suppress cytokine induced VCAM-1 expression in hRVE cells. However, in vitro assay demonstrated that not only anti-inflammatory n3- PUFA (DHA22:6n3), but also saturated palmitate and proinflammatory n6-PUFAS (LA18;2n6 and AA20;4,,6) could activate PPARa to a similar degree. This implies that in hRVE cells, activation of PPARa is not the sole target that DHA22:6n3 impacts to exert its anti- inflammatory function and that additional mechanisms are possibly involved. Further studies to compare the mechanism(s) of PPARa and DHA22:6n3 in suppressing cytokine induced inflammatory signaling in hRVE cells are needed. Moreover, the fact that EPA20;5,.3, and its elongated metabolite DPA22;5,,3, have been shown to downregulate the expression of VEGF R2 in endothelial cells, underscores one of the suggested pathways of suppression of VEGF signaling by n3-PUFAs[62, 63]. However, we did not observe the same effect with DHA22:6n3, suggesting that DHA22:6n3 inhibition of cytokine induced NFKB activation is through a different mechanism in hRVE. 118 Recently, DHA22;6,.3 has been shown to suppress LPS induced activation of NFKB through toll-like receptor 4 in murine macrophages[64]. Biochemical studies demonstrate that the molecular targets of DHA22;6.,3 are probably at the level of receptor localized plasma membrane, upstream of components MyD88 and Akt[64]. Our data suggest that the DHA22:6n3 effect could not only prevent the translocation of specific NFKB isoforms into the nucleus (where it binds to the Specific promoter) but also inhibit the phosphorylation of IKBO. and prevention of its degradation. This implies that DHA22:6n3 acts upstream of IKBG. to inhibit inflammatory Signaling. The more specific steps in DHA22:6n3 action are under further study. An interesting hypothesis will be to test whether DHA22;6,,3 could modify Specific plasma membrane domains, such as lipid rafts or caveolae, as reported for T cells[65, 66]. A number of inflammatory signaling proteins have been found to be localized in caveolae/lipid rafts, such as TNFR1[67, 68] and VEGFR2 (F LK-l) in endothelial cells[69, 70]. Whether DHA22:6n3 could modify the lipid composition of plasma membrane or caveolae/lipid rafts in hRVE and thus interfere with the inflammatory signaling awaits to be fully determined and is the focus of future study. In summary, our data demonstrates that three important inflammatory cytokines upregulated in diabetic eyes, TNFa, IL-lB and VEGF165 induce VCAM-1 and ICAM-1 expression through activating NFKB in hRVE cells. n3-PUFA (DHA22:6n3) and PPARa contribute to antagonizing the cytokine induced inflammatory response possibly through different molecular mechanisms. 119 6. References 10. 11. 12. Frank, R.N., Diabetic retinopathy. N Engl J Med, 2004. 350(1): p. 48-58. Stefansson, E., Diabetic retinopathy. N Engl J Med, 2004. 350(24): p. 2525-6; author reply 2525-6. Schroder, S., W. Palinski, and G.W. Schmid-Schonbein, Activated monocytes and granulocytes, capillary nonperfusion, and neovascularization in diabetic retinopathy. Am J Pathol, 1991. 139(1): p. 81-100. J oussen, A.M., et al., Nonsteroidal anti-inflammatory drugs prevent early diabetic retinopathy via INF-alpha suppression. Faseb J, 2002. 16(3): p. 438-40. Adamis, A.P., Is diabetic retinopathy an inflammatory disease? Br J Ophthalmol, 2002. 86(4): p. 363-5. Limb, G.A., et al., Distribution of INF alpha and its reactive vascular adhesion molecules in fibrovascular membranes of proliferative diabetic retinopathy. Br J Ophthalmol, 1996. 80(2): p. 168-73. Limb, G.A., et al., Evidence for control of tumour necrosis factor-alpha (TNF- alpha) activity by TNF receptors in patients with proliferative diabetic retinopathy. Clin Exp Immunol, 1999. 115(3): p. 409-14. Limb, G.A., et al., Platelet expression of tumour necrosis factor—alpha (TNF- alpha), TNF receptors and intercellular adhesion molecule-1 (ICAM-1) in patients with proliferative diabetic retinopathy. Clin Exp Immunol, 1999. 118(2): p. 213-8. Spranger, J ., et al., [TNF-alpha level in the vitreous body. Increase in neovascular eye diseases and proliferative diabetic retinopathy]. Med Klin (Munich), 1995. 90(3): p. 134-7. Kowluru, RA. and S. Odenbach, Role of interleukin-1 beta in the development of retinopathy in rats: eflect of antioxidants. Invest Ophthalmol Vis Sci, 2004. 45(11): p. 4161-6. Kowluru, RA. and S. Odenbach, Role of interleukin-1 beta in the pathogenesis of diabetic retinopathy. Br J Ophthalmol, 2004. 88(10): p. 1343-7. Adamis, A.P., et al., Increased vascular endothelial growth factor levels in the vitreous of eyes with proliferative diabetic retinopathy. Am J Ophthalmol, 1994. 118(4): p. 445-50. 120 l3. 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. Murata, T., et al., The relation between expression of vascular endothelial growth factor and breakdown of the blood-retinal barrier in diabetic rat retinas. Lab Invest, l 996. 74(4): p. 8 1 9-25. Aiello, L.P., et al., Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc Natl Acad Sci U S A, 1995. 92(23): p. 10457-61. Qaum, T., et al., VEGF-initiated blood-retinal barrier breakdown in early diabetes. Invest Ophthahnol Vis Sci, 2001. 42(10): p. 2408-13. Ellis, E.A., et al., Increased H202, vascular endothelial growth factor and receptors in the retina of the BBZ/Wor diabetic rat. Free Radic Biol Med, 2000. 28(1): p. 91-101. Aiello, L.P., et al., Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med, 1994. 331(22): p. 1480-7. Amin, R.H., et al., Vascular endothelial growth factor is present in glial cells of the retina and optic nerve of human subjects with nonproliferative diabetic retinopathy. Invest Ophthahnol Vis Sci, 1997. 38(1): p. 36-47. Frank, R.N., et al., Basic fibroblast growth factor and vascular endothelial growth factor are present in epiretinal and choroidal neovascular membranes. Am J Ophthalmol, 1996. 122(3): p. 393-403. Obrosova, I.G., et al., Antioxidants attenuate early up regulation of retinal vascular endothelial growth factor in streptozotocin-diabetic rats. Diabetologia, 2001. 44(9): p. 1102-10. Joussen, A.M., et al., Retinal vascular endothelial growth factor induces intercellular adhesion molecule-1 and endothelial nitric oxide synthase expression and initiates early diabetic retinal leukocyte adhesion in vivo. Am J Pathol, 2002. 160(2): p. 501-9. Usui, T., et al., VEGF(I64(165)) as the Pathological Isoform: Diflerential Leukocyte and Endothelial Responses through VEGFRI and VEGFR2. Invest Ophthalmol Vis Sci, 2004. 45(2): p. 368-374. Monaco, C. and E. Paleolog, Nuclear factor kappaB: a potential therapeutic target in atherosclerosis and thrombosis. Cardiovasc Res, 2004. 61(4): p. 671-82. Kowluru, R.A., et al., Diabetes-induced activation of nuclear transcriptional factor in the retina, and its inhibition by antioxidants. Free Radic Res, 2003. 37(11): p. 1169-80. 121 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Laws, A., et al., Differences in insulin suppression of free fatty acid levels by gender and glucose tolerance status. Relation to plasma triglyceride and apolipoprotein B concentrations. Insulin Resistance Atherosclerosis Study (IRAS) Investigators. Arterioscler Thromb Vasc Biol, 1997. 17(1): p. 64-71. Poisson, J .P., Essential fatty acid metabolism in diabetes. Nutrition, 1989. 5(4): p. 263-6. Vessby, B., et al., Diet, nutrition and diabetes mellitus. Ups J Med Sci, 2000. 105(2): p. 151-60. Vessby, B., Dietary fat and insulin action in humans. Br J Nutr, 2000. 83 Suppl 1: p. S91-6. Zenobi, P.D., et al., Improvement of lipid profile in type 2 (non-insulin-dependent) diabetes mellitus by insulin-like growth factor I. Diabetologia, 1993. 36(5): p. 465-9. Lewis, G.F., et al., Disordered fat storage and mobilization in the pathogenesis of insulin resistance and type 2 diabetes. Endocr Rev, 2002. 23(2): p. 201-29. Baldeweg, S.B., et al., Insulin resistance, lipid and fatty acid concentrations in 86 7 healthy Europeans. European Group for the Study of Insulin Resistance (EGIR). Eur J Clin Invest, 2000. 30(1): p. 45-52. Decsi, T., et al., Polyunsaturated fatty acids in plasma and erythrocyte membrane lipids of diabetic children. Prostaglandins Leukot Essent Fatty Acids, 2002. 67(4): p. 203-10. F utterman, S. and C. Kupfer, The fatty acid composition of the retinal vasculature of normal and diabetic human eyes. Invest Ophthalmol, 1968. 7(1): p. 105-8. Goldberg, I.J., Clinical review 124: Diabetic dyslipidemia: causes and consequences. J Clin Endocrinol Metab, 2001. 86(3): p. 965-71. Davis, M.D., et al., Risk factors for high-risk proliferative diabetic retinopathy and severe visual loss: Early Treatment Diabetic Retinopathy Study Report #18. Invest Ophthahnol Vis Sci, 1998. 39(2): p. 233-52. Ferris, F .L., 3rd, E.Y. Chew, and B.J. Hoogwerf, Serum lipids and diabetic retinopathy. Early Treatment Diabetic Retinopathy Study Research Group. Diabetes Care, 1996. 19(11): p. 1291-3. Chew, E.Y., et al., Association of elevated serum lipid levels with retinal hard exudate in diabetic retinopathy. Early Treatment Diabetic Retinopathy Study (ETDRS) Report 22. Arch Ophthalmol, 1996. 114(9): p. 1079-84. 122 38. 39. 40. 41. 42. 43. 45. 46. 47. 48. 49. 50. Natarajan, R., ct al., Role of IZ-lipoxygenase and oxidant stress in hyperglycaemia-induced acceleration of atherosclerosis in a diabetic pig model. Diabetologia, 2002. 45(1): p. 125-33. Sen, K., et al., Simvastatin retards progression of retinopathy in diabetic patients with hypercholesterolemia. Diabetes Res Clin Pract, 2002. 56(1): p. 1-11. Chowdhury, T.A., et al., The role of serum lipids in exudative diabetic maculopathy: is there a place for lipid lowering therapy? Eye, 2002. 16(6): p. 689-93. Gupta, A., et al., Lipid-lowering drug atorvastatin as an adjunct in the management of diabetic macular edema. Am J Ophthalmol, 2004. 137(4): p. 675- 82. Lyons, T.J., et al., Diabetic retinopathy and serum lipoprotein subclasses in the DCC T /EDIC cohort. Invest Ophthalmol Vis Sci, 2004. 45(3): p. 910-8. Calder, P.C., N—3 polyunsaturated fatty acids and inflammation: fiom molecular biology to the clinic. Lipids, 2003. 38(4): p. 343-52. Calder, P.C., Polyunsaturated fatty acids, inflammation, and immunity. Lipids, 2001. 36(9): p. 1007-24. Pawar, A., et al., Fatty acid regulation of liver X receptors (L20?) and peroxisome proliferator-activated receptor alpha (PPARalpha) in HEK293 cells. J Biol Chem, 2002. 277(42): p. 39243-50. Desvergne, B. and W. Wahli, Peroxisome proliferator-activated receptors: nuclear control of metabolism. Endocr Rev, 1999. 20(5): p. 649-88. Devchand, P.R., et al., PPARs: nuclear receptors for fatty acids, eicosanoids, and xenobiotics. Adv Exp Med Biol, 1999. 469: p. 231-6. Neve, B.P., et al., PPARalpha agonists inhibit tissue factor expression in human monocytes and macrophages. Circulation, 2001. 103(2): p. 207-12. Staels, B., et al., Activation of human aortic smooth-muscle cells is inhibited by PPARalpha but not by PPARgamma activators. Nature, 1998. 393(6687): p. 790- 3. Jackson, S.M., et al., Peroxisome proliferator-activated receptor activators target human endothelial cells to inhibit leukocyte-endothelial cell interaction. Arterioscler Thromb Vasc Biol, 1999. 19(9): p. 2094-104. 123 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. Marx, N., et al., PPARalpha activators inhibit cytokine-induced vascular cell adhesion molecule-1 expression in human endothelial cells. Circulation, 1999. 99(24): p. 3125-31. Pasceri, V., et al., Modulation of vascular inflammation in vitro and in vivo by peroxisome proliferator-activated receptor-gamma activators. Circulation, 2000. 101(3): p. 235-8. Rival, Y., et al., PPARalpha and PPARdelta activators inhibit cytokine-induced nuclear translocation of NF -kappaB and expression of VCAM-1 in EAhy926 endothelial cells. Eur J Pharmacol, 2002. 435(2-3): p. 143-51. Diep, Q.N., R.M. Touyz, and EL. Schiffi’in, Docosahexaenoic acid, a peroxisome proliferator-activated receptor-alpha ligand, induces apoptosis in vascular smooth muscle cells by stimulation of p38 mitogen-activated protein kinase. Hypertension, 2000. 36(5): p. 851-5. Chen, W., et al., Dyslipidemia, but not hyperglycemia, induces inflammatory adhesion molecules in human retinal vascular endothelial cells. Invest Ophthalmol Vis Sci, 2003. 44(11): p. 5016-22. Viatour, P., et al., Phosphorylation of NF -kappaB and IkappaB proteins: implications in cancer and inflammation. Trends Biochem Sci, 2005. 30(1): p. 43- 52. Daynes, RA. and DC. Jones, Emerging roles of PPARs in inflammation and immunity. Nat Rev Immunol, 2002. 2(10): p. 748-59. Kuldo, J .M., et al., Molecular pathways of endothelial cell activation for (targeted) pharmacological intervention of chronic inflammatory diseases. Curr Vase Pharmacol, 2005. 3(1): p. 11-39. Zheng, L., C. Szabo, and T.S. Kern, Poly(ADP-ribose) polymerase is involved in the development of diabetic retinopathy via regulation of nuclear factor-kappaB. Diabetes, 2004. 53(11): p. 2960-7. Romeo, G., et al., Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes. Diabetes, 2002. 51(7): p. 2241-8. Lee, T.H., et al., Eflect of dietary enrichment with eicosapentaenoic and docosahexaenoic acids on in vitro neutrophil and monocyte leukotriene generation and neutrophil function. N Engl J Med, 1985. 312(19): p. 1217-24. Yang, SR, 1. Morita, and SJ. Murota, Eicosapentaenoic acid attenuates vascular endothelial growth factor-induced proliferation via inhibiting F lk-I receptor 124 63. 64. 65. 66. 67. 68. 69. 70. expression in bovine carotid artery endothelial cells. J Cell Physiol, 1998. 176(2): p. 342-9. Tsuji, M., S.I. Murota, and I. Morita, Docosapentaenoic acid (22:5, n-3) suppressed tube-forming activity in endothelial cells induced by vascular endothelial growth factor. Prostaglandins Leukot Essent Fatty Acids, 2003. 68(5): p. 337-42. Lee, J .Y., et al., Reciprocal Modulation of T oll-like Receptor-4 Signaling Pathways Involving MyD88 and Phosphatidylinositol 3-Kinase/AK T by Saturated and Polyunsaturated Fatty Acids. J. Biol. Chem., 2003. 278(39): p. 37041 -3705 l. Stulnig, T.M., et al., Polyunsaturated eicosapentaenoic acid displaces proteins fiom membrane rafts by altering raft lipid composition. J Biol Chem, 2001. 276(40): p. 37335-40. Zeyda, M., et al., LA T displacement from lipid rafts as a molecular mechanism for the inhibition of T cell signaling by polyunsaturated fatty acids. J Biol Chem, 2002. 277(32): p. 28418-23. Ko, Y.G., et al., TNF-alpha-mediated apoptosis is initiated in caveolae-like domains. J Immunol, 1999. 162(12): p. 7217-23. Legler, D.F., et al., Recruitment of TNF receptor 1 to lipid rafts is essential for TNFalpha-mediated NF -kappaB activation. Immunity, 2003. 18(5): p. 655-64. Cho, C.H., et al., Localization of VEGFR—2 and Phospholipase D2 in Endothelial Caveolae is Involved in VEGF-induced Phosphorylation of MEK and ERK. Am J Physiol Heart Circ Physiol, 2004. Feng, Y., et al., VEGF -induced permeability increase is mediated by caveolae. Invest Ophthalmol Vis Sci, 1999. 40(1): p. 157-67. 125 IV. Proteomic analyses of detergent resistant caveolae/lipid rafts from cultured human retinal vascular endothelial cells (hRVE) 1. Abstract Endothelial cells (EC) contain caveolae and lipid rafts, the specialized glycosphingolipid and cholesterol enriched micromembrane subdomains that crucially participate in various essential cellular fimctions such as signal transduction, vesicular trafficking and cholesterol homeostasis. The protein constituents of caveolae/lipid rafts isolated by buoyant density methods from cultured human primary retinal endothelial cells (hRVE) were determined using liquid chromatography-electrospray ionization tandem mass spectrometry (LC-MS/MS) analysis. About 70 proteins were subsequently identified in combination with LC-MS/MS after in-gel and in-solution digestions. Proteins present included a large number of known caveolae/lipid raft residents, such as caveolin and flotillin. Overall, together with western blot analysis, known glycosylphosphatidylinositol (GPI)-anchored proteins, transmembrane proteins, cytoskeleton and associated proteins, and cell signaling proteins including G proteins, small GTPases, Src-farnily kinases were found to be present. This was highlighted with the identification of several important EC functional proteins such as CD44, a hyrarunan receptor; CD36, a scavenge receptors for oxidized lipids and free fatty acids; p63; Bene; CD147 and brain acid soluble protein together with retinoic acid induced 3 protein (RAIGl) and Signal proteins induced under serum starvation. These findings suggest crucial involvement of caveolae/lipid rafts in numerous retina vascular functions including permeability, migration, angiogenesis, vesicle trafficking, lipid homeostasis 126 and inflammation. The possible link to the pathogenesis of important retinal vascular diseases such as diabetic retinopathy is discussed. 127 2. Introduction Caveolae were originally described as flask-like membrane invaginations of 50- 100nm on the surface of endothelial and epithelial cells by electron microscopy 50 years ago[l]. Characterized by the presence of the principal structural and regulatory protein caveolin[2], caveolae and the so-called lipid rafts (devoid of caveolin) share a similar distinct lipid composition notable for high concentrations of sphingolipids and cholesterol. These glycosphingolipid enriched membrane domains which pack in a liquid-ordered structure and are resistant to solubilization by nonionic detergents at low temperatureare also called detergent-resistant membrane raft (DRM) structures. Endothelial cells are one of the cell types that contain a large number of caveolae and express a high level of caveolin-H3, 4]. Caveolae/lipid rafts play important roles in endothelial permeability, vesicle trafficking[5-7], cholesterol homeostasis[8], and endothelial cell Signaling[9-11]. Recent in vivo data from caveolin-l (CAV1)-null mice showed a lack of caveolae formation in the microvascular endothelium, which significantly alters microvascular permeability[12-l4]. More importantly, caveolae/lipid rafts are identified as sites for the sequestration of diverse membrane-targeted signaling proteins (reviewed in [11]). Various Signaling molecules, including endothelial nitric oxide synthase (eNOS), G-protein coupled receptors (GPCRS), VEGF receptor 2, PDGFR and EGFR and GPI-linked proteins are partially segregated into these domains with or without activation or ligand binding, suggesting a critical role of caveolae/lipid rafts in endothelial cells signaling transduction[11]. Given their crucial firnctions in transport, cholesterol homeostasis and signal transduction, caveolae and lipid rafts are 128 suspected to play an important role in various diseases such as atherosclerosis[15, 16]. More importantly, increased numbers of caveolae in retinal endothelium and pericytes in hypertensive diabetic rats have been demonstrated to be one of the possible mechanisms in hypertension enhanced diabetic microvascular disease, especially diabetic retinopathy[ l 7]. Retinal microvascular endothelial cells form a tight inner blood-retina barrier to regulate the movement of molecules in and out of the retina. Caveolae have particular significance for the highly specialized, continuous endothelium of the retinal vasculature since these organelles are thought to have a unique role in regulating uptake and transcytosis of proteins across the blood retina barriers. Diabetes causes increased permeability of retina mierocirculation which results in water, albumin and lipid leakage, with accumulation of lipid exudates and intraretinal fluid[l8]. Considering the importance of caveolae/lipid rafts in regulating endothelial cell permeability and signal transduction, further research is needed to determine the function and character of the protein components in caveolae/lipid rafts from human retinal microvascular endothelial cells. Recent advances in proteomics and protein identification technology have permitted the identification of numerous proteins contained in subcellular fi'actions such as lipid rafts. The protein components in lipid rafts from Jurkat T cells[19], neutrophils[20] and monocytes[21] or in caveolae/lipid rafts from Cos cells[22], Vero cells[23] and human endothelial cells (HUVEC)[24] have all been characterized by mass spectroscopy (MS). Here we have isolated caveolae/lipid rafts from human primary 129 retinal vascular endothelial cells using standard methods based on detergent insolubility in Triton X-100 at 4°C. The protein components associated with these cold detergent resistant vesicles were characterized by proteomics using both in-gel digestion and in- solution digestion followed by LC-MS/MS. About 70 proteins were identified including cytoskeletal proteins, ion-channel proteins, signaling proteins such as GPI-linked proteins, G-proteins, proteins involved in angiogenesis, adhesions, and inflammation and some specific retinal endothelial cell proteins. Western blot analysis was also used to identify several less-abundant proteins which had escaped proteomic analysis such as Src-farnily kinases, and proteins involved in cholesterol and lipid transport. This is the first report that systematically characterized the protein components of the important lipid microdomains involved in variable functions of primary retinal endothelial cells, suggesting caveolae/lipid rafts play key roles in mediating the blood-retina barrier. 130 3. Materials and Methods 3.1. Reagents and antibodies Sequence-grade trypsin was purchased from Promega (Madison, WI, USA). Thiourea, triton X-100 Ultra Pure, MES, ammonium bicarbonate and formic acid were purchased from Sigma (St. Louis, MO, USA). Mouse anti-caveolin-l and flotillin-l were obtained from Upstate Biotechnologics (Lake Placid, NY); Rabbit anti-ERK1/2 was from Cell Signaling; mouse anti-PKCa, c-Src; Rabbit antibodies against c-yes, CD36, CD44, Na+/K+ ATPase and mouse antibodies against F yn were purchased fiom Santa Cruz. 3.2. Cell culture Primary cultures of hRVE cells were prepared as previously described[25]. Cells were maintained in growth medium consisting of DMEM/F12 (Invitrogen), 5.5 mM glucose, 10% fetal bovine serum (Invitrogen), endothelial cell growth supplement (Upstate Biotechnologies), insulin/Uansferrin/selenium mix (Sigma-Aldrich) and antibiotic—antimycotic solution (Invitrogen). Passages 1-6 cells were plated to yield near- confluent cultures at the end of the experiment. For experimental treatments, the cells were transferred to serum-free medium for 18 to 24 hours before treatment. 3.3. Electrophoresis and immunoblotting Cells were lysed in the lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 1.5 mM MgC12, 1 mM EGTA, 1% Triton X-100, 10% glycerol) with freshly added protease inhibitor cocktail (Sigma) and phosphatase inhibitors (1 mM Na3VO4, 100 pM 131 glycerophosphate, 10 mM NaF, 1 mM Na4PPi). Proteins were resolved by SDS-PAGE and transferred to nitrocellulose, immunoblotted using appropriate antibodies followed by secondary horseradish peroxidase conjugated antibody (Bio-Rad). Irnmunoreactive bands were visualized by enhanced chemiluminescence (ECL kit; Amersham Pharmacia Biotech, Piscataway, NJ). Blots were quantitated by scanning densitometry using Image] software, ver. 1.29 (available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). 3.4. Isolation of lipid rafts/caveolin-rich membrane domains Caveolae/lipid rafts were prepared using a slightly modified sucrose gradient ultracentrifugation protocol [11]. Briefly, 5x106 hRVE cells were washed with cold PBS twice and then lysed in 0.8 ml MNE buffer (25 mM MES, PH 6.5, 0.15 M NaCl and 5 mM EDTA) containing 1% Triton X-100 and fresh protease and phosphatase inhibitors and kept on ice for 20 min. The homogenization was carried out with 10 strokes of a tight-fitting Dounce homogenizer, and then spun at 4,000xg at 4°C for 10 min. Supernatant (0.8 ml) was then mixed with the same volume of 80% sucrose prepared in MNE buffer, and placed at the bottom of an ultracentrifuge tube. 1.6 ml 30% sucrose and 0.8 ml 5% sucrose were overlaid on top of the sample to form a 5-30% discontinuous sucrose gradient. After 16 h of centrifugation at 200,000xg at 4°C using a swinging bucket rotor, 0.4 ml samples was collected from the top for each fiaction. A band confined to fractions 2 to 4 was designated as caveolae/lipid raft-enriched membrane domain. The combined fractions 2 to 4 were further diluted 3 times in MNE buffer and 132 Spun at 200,000xg at 4°C for another 2 h to precipitate the caveolae/lipid rafts and the pellet was designated as the insoluble fraction (1). Fractions 6-10 were also combined and designated as the soluble fraction (S). 3.5. Protein in-gel digestion The combined caveolae/lipid raft fraction (1) was typically separated by SDS- PAGE on a 4-20% gradient gel (Biorad). Protein bands were visualized by sypro-bluc staining (Molecular Probes) and then imaged using a Bio-Rad FX Pro+ laser scanner. The entire 1D SDS sample lane was cut into 30 2 mm sections and dehydrated twice with acetonitrile for 20 min, and dried under vacuum concentrator for about 3 min. The dried gels were then rehydrated in 30 pl of sequencing grade trypsin solution (20 ng/pl) with 50 mM ammonium bicarbonate on ice for 10 min with occasional vortex mixing and digestion was performed overnight at 37 °C. 3.6. Protein in-solution digestion After pelleting the caveolae/lipid rafts fraction, proteins were dissolved in 15 pl buffer containing ammonium bicarbonate (100 mM, pH 8.0) and 2 M tlriourea, 2 mM DTT and incubated at 60°C for 45 min to reduce and denature the proteins. Sequence grade trypsin (100 ng in 85 pL of 50 mM ammonium bicarbonate) was added to digest the protein mixture at 37°C overnight. Tryptie peptides were concentrated by Speed-Vac evaporation. 133 3.7. LCIMSIMS The peptides were extracted from each gel section and desalted on a 1 x 0.2 mm Magic C18 Captrap cartridge. The bound peptides were then flushed onto a 15 cm x 75 pm New Objectives Picofrit column packed with Microm Magic C18 AQ packing material and eluted over 60 minutes with Buffer A (0.1% formic acid) and a gradient of 5% to 70% B (95% Acetonitrile 0.1% formic acid, starting at 10 min) into a Thermofinnigan Deca XP+ Ion trap mass spectrometer with a flow rate of 250 nl/min. The top three ions in each survey scan were then subjected to automatic low energy collision induced dissociation (CID) and the resulting uninterpreted MS/MS spectra were searched against the IPI_human database using the Mascot searching algorithm. The Mascot results from every gel section in the lane were combined into one database using an in-house database program. Identifications are usually considered positive if 2 peptides per protein are identified with a significant Mascot score (p< 0.05). 134 4. Results 4.1. Isolation of caveolae/lipid rafts from hRVE Although a number of studies have demonstrated the importance of caveolae/lipid rafts in regulating endothelial cell firnction, the protein components of these specialized lipid microdomains in human primary retinal endothelial cells have not been determined. We addressed this issue by characterizing endothelial specific caveolae/lipid rafts in cultured hRVE cells. Caveolae/lipid rafts fractions were isolated by sucrose discontinuous gradient ultra centrifugation (Fig. 1) based on their insolubilization in Triton X-100 at 4 °C. The overall raft isolation was confirmed Since caveolin-1 (a caveolae marker) and flotillin-l (a lipid raft marker) were found highly enriched in density gradient fractions 2to 4). In contrast, the general plasma membrane marker, Nal/Kl ATPase, was excluded from the lipid raft fractions. PKCa and ERK1/2 were mainly in the soluble fractions (fractions 6-10) under basal conditions (Fig. 1). These results demonstrate that hRVE contains both caveolae and lipid rafts represented by the enrichment of both caveolin-1 and flotillin-l in the DRMS isolated using detergent insolubility and density gradient fractionation. 4.2. Characterization of caveolae/lipid rafts components by in gel digestion and LCIMSIMS With the advance in proteomics, several methods have been employed to characterize proteins from specialized microdomains or organelles. To further characterize the protein components in caveolae/lipid rafts in hRVE, we performed 135 conventional gel band excision followed by in-gel trypsin digestion and LC/MS/MS. This method was used to primarily identify abundant proteins due to limitations of gel loading. High resolution SDS-PAGE (4-20%) gradient gels were run to first determine the level of complexity of the protein content of caveolae/lipid rafts preparations via syprob blue staining. 20 pg caveolae/lipid rafts protein from about 5x106 cells were loaded on the gel. Fig. 2 shows a number of proteins are present in the caveolae/lipid rafts preparations from serum starved hRVE cells. A total number of about 70 proteins were identified as detailed in Table I. This include proteins representing a number of functional classes, such as caveolae and lipid rafts structural proteins, well known raft-localized GPI-linked proteins, ion channel proteins, cytoskeleton proteins, and signaling proteins, as well as some putative proteins. Some new proteins were also identified especially proteins involved in signal transduction induced by serum starvation, and in retinoic acid-related signal transduction. 4.3. Identification of caveolae/lipid rafts components by in solution digestion and LCIMSIMS The in-solution digestion coupled with LC/MS/MS was used to additionally identify proteins that might otherwise not be evident fiom in-gel trypsin digestion. Table II summarizes the proteins collectively identified from three independent experiments using the caveolae/lipid rafts samples purified from 3 different donors. About 25 proteins were unequivocally identified. 136 4.4. Identification of less abundant proteins in hRVE by western blot Due to the limitations of proteomic approaches from both in-gel and in-solution digestion, several less abundant proteins were identified by western blotting. These included the important Src family kinases, such as Fyn and c-yes, which were about 90% enriched in the caveolin/raft fraction. CD36, the scavenger receptor for modified lipoproteins or free fatty acids was also present exclusively in caveolae/lipid rafts fractions from hRVE (Fig. 3). 137 Caveolae/lipid rafts (I) Soluble (S) Gradient r—H f-——-A—\ Fractions top 1 2 3 4 5 6 7 8 9 bottom blot an a... Flotillin-1 ~ Caveolin-1 “a ' ' t 3 a a m Nat/KtATPase --- PKCa Fig. 1. Characterization of caveolae/lipid rafts in hRVE cells. Caveolae/lipid raft enriched domains were purified as indicated in methods. Gradient fractions were separated by SDS-PAGE. Western blot against the caveolae marker, caveolin-1, and the lipid raft marker, flotillin-l, as well as other proteins such as PKCa and ERKl/2 were performed. The membrane marker, Na+/K+ ATPase, was also analyzed to confirm the purity of the purification. 138 Soluble Caveolae (S) /lipid rafts (I) ' "'" fl . kDa —210 —125 —101 _29 Fig. 2. SyproB Blue staining of caveolae/lipid rafts proteins (1) and soluble (S) proteins isolated from hRVE cells after SDS-PAGE. Caveolae/lipid rafts were obtained from 5x106 hRVE cells as described in Methods. Pooled caveolae/lipid raft fractions (1) and pooled soluble fractions (S) were separated by SDS-PAGE, stained with SyproB Blue and visualized under UV light. 139 Caveolae Soluble llipid rafts (I) (S) blot ..................... Fyn C-Yes C-Src CD36 Flotillin-1 l Caveolin-l Fig. 3. Localization of Src family kinases in hRVE. Caveolae/lipid rafts from hRVE cells were purified as described in Methods. Pooled caveolae/lipid rafts fractions (1) and pooled soluble fractions (S) were separated on SDS-PAGE and analyzed by western blot. 140 TABLE I. Proteins identified by in-gel digestion and LCIMSIMS. Peptides Accession identified # Related functions Mr(kDa) Identity Marker proteins: Plot: i 1 1 in - 1 47.3 5 NP_005794 Structural and regulatory Plot: 1 1 1 in - 2 42.6 4 NP_004466 proteins of lipid rafts Caveol in - 1 20.458 9 NP_001744 Structural and regulatory Caveol in — 2 18.279 3 NP;001224 proteins of caveolae Integral membrane S tomat: in a 3 1.71 9 NP_004090 proteins, scaffolding protein of lipid rafts CPI-linked proteins: Alpha2 macroglobulin/compleme CD109 161.62 9 NP_598000.1 nt C3, C4, C5 gene family (AMCOM) A1 ka 1 i ne phosphatase, tissue non- Exact physiological specific 57‘“ 2 NP_000469 function unknown isoform precursor Potent inhibitor of the CD5 9 co lement membrane glycoprotein 14.17 1 P13987 “‘9 attack complex (mac) precursor action 5 . _ Formation of anti- nuc 1eot idase inflammatory and 63.327 18 NP_002517.1 irnrnunosuppressive gggursor ' adenosine from extracellular nucleotides Cadherin - 1 3 78.224 5 NP_001 2 48 Calcium dependent cell precursor adhesron molecules Reversion- inducing cysteine - rich Negatively regulates protein with 106.38 1 NP_066934 MMP-9, MMP-2 Kazal motifs secretion and activity precursor (hRECK, STIS) 141 Table 1 (cont’d) Cytoskeleton and related proteins: Myosin-IIA HC (non-muscle type 226.39 61 NP_002464 A) MYOlB protein 131.902 9 NP_036355 Myosin Ic 117.96 6 NP_203693 Myosin regulatory light chain 19 78 4 NP_006462 (MRCL3)‘same ' NP_291024 peptides as MRCLZ calmodulin 17.67 1 ..... msssgsss alkali light chain . l NP_002467 motor, p ays a to e m iso form 4 endocytos1s and . Tub ulin, a1 ha 6 intracellular trafficking . P 52.597 3 NP 116093 cham - "Wm: “19"” 3 52.279 3 NP 006000 cham - “8““ beta" 49.727 5 NP 110400 cham - Tubulin beta-4 chain 50.4 2 NP_006078 Actin, cytoplasmic 1(beta Actin) 41.710 4 NP_001092 Actin, cytoplasmic 2 l amma Actin) 41.766 4 P63261 Paramyosin family, integral membrane P63 (CKAP4) 65.98 14 NP_006816 protein, highly enriched in detergent-insoluble fraction Mediate actin F-actin capping cytoskeleton organization protein alpha-1 32.902 2 NP_OO6126 and biogenesis, cell subunit motility and protein complex assembly Ion-channel proteins: Pore forming proteins also present in plasma Voltage-dependent 30.639 1 NP_003365 membrane especially amon channel 1 caveolae other than mitochondria Transient receptor A calcium-activated potential cation nonselective (CAN) channel, subfamily 136.42 2 NP_060106 cation channel that M, member 4 mediates membrane (TRPM4) depolarization Dihydropyridine- sensitive L-type, calcium channel 123.106 15 NP_000713 Calcium channel protein alpha-Zldelta subunits precursor 142 Table 1 (eont’d) flgnallng proteins: G protein-coupled receptors Retinoic acid fl induced 3 protein 40.225 1 NP_003970 Type? G 9mm” . ( 1211161 or coupling receptor family GPRCSA) G proteins G (i) a; (GNAIZ) 40.27 3 NP_002061 Regulates G protein— Gy- 12 subuni t 7.9 1 NP_061329 coupled receptor activity (?TPaM5 IKHwnflwrofflwlbm superfamily of small R-Ras 23.466 1 NP 006261 .GTP?“ ‘1‘"? has ”“9 — implicated in promoting cell adhesion and neurite mMgnwwh R-Rasz 23.385 1 NP!036382 GTPase Other signaling proteins Serum , deprivat ion Binds to . . 47.14 1 NP 004648 phosphatidylserine, response - ( 8 d r ) substrate of PKC lmhmedbysmmni HSRBC 27.625 5 NP 659477 5min”: a bum“? - protein of the protein kinase C, delta (PRKCD) 21:3: 1 i i te G'I'Pase activator induzed 44.394 1 NP_057479 activity, involved in Rac transcript 1) mediated JNK activation Calmodul in 17.5 1 AAB23129 Calcium binding 32:23::Cyte - RhoGEF, crucial for ezr in- 1 ike microfilament domain 118 5 5 l NP_00100171 orgamza' tion, involved in . ' 5 the adhesion, containing . . prote i n proliferation, and ( CDEP) differentiation 143 Table 1 (oont’d) Other proteins: CD44 39.41 2 AAA82949 A hyalumm (HA) receptor Membrane alanine Member of the zinc- aminopeptidase precursor 109.44 8 NP_001141 binding metalloprotease LANPEP) superfamily. Complement C3 precursor 187.1 1 NP_OOOOSS Integral membrane proteins Complement component C9 63.133 2 NP_00172 8 Integral membrane precursor proteins Potential calcium- Protocadherin 10 precursor 118.57 1 NP_065 866 dependent cell-adhesion protein Heat shock 90kDa protein 1, Regulation of protein alpha 84'621 1 “-0053” foldin and function E Solute carrier family 2, facilitated glucose 56.9 1 NP_006507 Glucose transporter transporter, member 1 Proton-linked M°“°°“n::yl“° 49.43 1 NP_004687 monocarboxylate m transporter. Specific ligand for the . (115/B3 and 015/135 Lactadherin precursor 43.09 1 NP_005919 receptors, mediate cell adhesion . . . A regulator of matrix giflilgngcgfiggn 42. 17 1 NP_001 719 metalloproteinase g ’ (MMP) production Calcium-dependent phospholipid-binding . . . NP_00100285 protein regulating Annexm II (lipocortin 1) 40.328 3 8 cellular growth, signal transduction, and exocytosis. Myeloid-associated Mal family, help target differentiation marker 35°25 2 ”-612382 GPI-linked protein Vacuolar ATP synthase, A 1 ti b . t catalytic subunit A, 68.26 4 NP_001681 mm“ 3‘; “m my“ ubiquitous isoform .t me ates Vacuolar ATP synthase aeidiflcation of . . . ’ 56.45 4 NP_001684 eukaryotic intracellular subumt B’ brain isoform organelles necessary for Limiagflp synthasc’ 40.3 2 NP_057078 such intracellular processes as protein V;°“9i“§ 1MP ”mas" 52.121 3 NP_001174 sorting, zymogen su um precursor activation, receptor- V 1 ATP thase ' 8:13:58 1 syn , 1 3.7 5 1 NP_004879 mediated endocytOSis Ubiquitin and ribosomal a fusion protein protein S27a ”953 1 ”—0029“ (ubiquitin and 327:1 Modulates cell-cell and Beta-galactosidase binding cell-matrix interactions; lectin precursor (galectin-l) ”'70 l NP_002296 may be part of novel anti-inflammatory loop 144 Table 1 (cont’d) Possible nonspecific proteins: Main structural protein of . the polyhedral lattice 5111:??? heavy 191.43 2 NP_004850 surrounding coated pits and coated vesicles, oganelles Moderately similar to Hypothe t i ca 1 Neuroblast. . protein FLJ4 6 8 4 6 180.58 6 NP_076965 difl'erentiation assoc1ated protein AHNAK (nucleoprotein) Component of De smoglein 2 intercellular desrnosome precursor 122 5 NP-OO1934 junctions mediating cell- cell adhesion Sodium/potassium- transporting . ATPase alpha _ 3 1 1 1.66 2 NP_689509 Sodium pump chain Splice isoform Intracellular pumps SBRCAzA of P16615 located in the Sarcoplasmic/ endo 109.62 2 NP_001672 sarcoplasmic or plasmic reticulum endoplasmic reticula of calcium ATPase 2 muscle cells KIAA014 3 protein 103.138 2 NP_055952 gglfigfme . . A component of the Splice isoform 1 . . of oswxae Stonin 101.102 1 NP 149095 “Pd“yuc mach“? that 2 (s tonin 2) - likely regulates veSicle endocytoms (12:33:? 3 Of Nuclear protein that Interleukin 82.8 3 NP_703194 ”3“?“ “3° .“ct‘v‘ty °f enhancer-binding protthein-argigine I factor 3 (NFATQO) me ylm erase . Rough endoplasmic RPNZ protein 73.94 3 NP 002942 reticulum-specific (ribophorin II) - . membrane glycoproteins KIAAO 8 3 0 prote i n 59.347 3 XP_290546 ND Dihydrolipoyllysi ne-residue succinyltransfera . . as component of 48.6 1 NP_001924 2243;312:1132? “°er 2-oxoglutarate dehydrogenase complex 145 Table 1 (oont’d) Leucine-zipper RNA polymerase I and (pBBIP) protein FKSGl3 43'“ 5 NP-036364 transcript release factor a-act in 41.992 3 NP_005150 Smooth muscle actin Hypothetical protein 40.42 1 ND ND DKFZpS47P237 Glyceraldehyde-3- . phosphate 36.03 2 NP_002037 mhymgm“ 1“ deh dr na e glucose metabolism y oge s Chromosome 6 open reading frame 188 35'” 2 ND ND Muscle-specific Membrane-bound DNase I - 1 ike 33.8 3 NP_006721 (potential), belongs to the precursor dnase i family Hypothetical protein DKPZp56413227 31.62 2 NP_079119 111’”? mm?’ (cytochrome b e 0° on transpo reductase 1) Cytochrome c . oxidase 25.55 1 NP_653214 mfggmn polypeptide II p0 Golgi integral membrane protein, involved in the RERI protein 22.94 1 NP_008964 ;“éggnffrfimms fmm the early golgi communnwnt Hypothetical protein FLJ46113 21.114 2 NP_060039 ND Note: ND: not identified. 146 TABLE 2. Proteins identified bi in-solution digestion and LCIMSIMS. Structural proteins: Structural and regulatory (+) Caveolln-1 20.458 5 NP_001744 proteins of caveolae Integral membrane (+) Stomatin 31.71 9 NP_004090 proteins, scaffolding protein of lipid rafts GPl-linked proteins: Potent inhibitor of the (+) 0059 . complement membrane glrecuycoprsrztreln 14.17 1 P13987 attack complex (mac) p action Formation of anti- , . inflammatory and (gawfiggm 63.327 18 NP_002517.1 immunosuppressive p ’ adenosine from extracellular nucleotides (+) Cadherin-13 Calcium dependent cell precursor 78224 5 NP-OO1248 adhesion molecules Cytoskeleton and associated proteins: (+) Myosin-IIA HC (non-muscle type 226.39 15 NP_002464 A) (+) MYO1B protein 131.902 2 NP_036355 . (+) Myosin lc 117.96 2 NP £03693 Ads as a" 3“" ”a?“ (+) calmodulin 17 67 8 AAB 23129 "mm" 9'3” a '°'° '" (+) Actin ' — endocytosis and cytoplasmic 1 ( 41.710 16 NP_001092 '"t'aw'm'a' "ammo beta Actin) (+) Actin, cytoplasmic 2 41.766 16 P63261 (gamma Actin) Ion-channel proteins: (+) Dihydropyridine- sensitive L-type, Calcium channel protein, calcium channel 123.106 15 NP_000713 integral membrane alpha-Zldelta protein subunits precursor Signaling proteins: +) G(i)a2 (GNAI2) 40.27 6 NP_002061 G(i)a3 (GNAI3) 40.506 4 NP_006487 fl) Gy-12 subunit 7.9 2 NP_061329 Regulates G protein- ° I 1 ct' ' G('VG.‘S)’G(T) ‘3 37.307 3 NP_002065 °°”p ed "mp °' 8 “my subunit 1 G(i)lG(s)/G(T) a subunit 4 37.303 3 NP_005264 147 Table 2 (oont’d) Other proteins: A hyaluronan (HA) (+) CD448 39.417 4 AAA82949 receptor (+) Membrane alanine 53328331331231}; aminopeptidase 109.44 6 NP_001141 su e [1% mil T p e ll 5° precursor (ANPEP) p Y' Y”. membrane protein . . . Calcium-dependent WNW" A2 ("Wm 38.449 2 NP_001002857 phospholipid-binding motein Integral membrane (+) Myeloid-associated protein (probable). Mal differentiation marker 3525 2 NP_612382 family, help target GPI- linked protein Vacuolar ATP synthase, subunit E 31.0 NP_001687 A multisubunit enzyme Thy-1 membrane Activation-associated glycoprotein precursor 17.923 4 NP_006279 cell adhesion molecule (CD90) in EC . . Myristoylated protein 8' :{25fi'1gg'gfi'fi 22.549 NP_006308 binds to cholesterol-rich pr raft-like domains MAL proteolipid family, an element of the Bone protein 17.39 NP_005425 machinery for raft- mediated trafficking in endothelial cells Nuclear envelope Nesprin 2 79.585 3 QBWXHO spectrin repeat protein 2 Note: (+): indicates the proteins are also identified by in-gel digestion followed by LCIMSIMS. a: Peptides identified are marked according to the results obtained from in solution digestion. ND: not identified. 148 5. Discussion Here we analyzed the protein composition of caveolae/lipid rafis isolated from cultured human primary retinal endothelial cells using both proteomics and biochemical analyses. As a central organelle of vascular endothelial cells, caveolae/lipid rafts serve as platforms for important functions in mediating endothelial cells signaling, vesicle trafficking, lipid transport and homeostasis. Using Triton X-100 insoluble and low density based sucrose floatation, both caveolae and lipid rafts structural proteins such as caveolin-1, 2 and flotillin-l, 2 as well as the stomatin alpha subunit (which forms oligomers and acts as an important structural protein) were identified by MS and biochemical analyses. Besides its structural role in caveolae, the 22 KDa palmitoylated protein, caveolin-1, also regulates cholesterol efflux due to its high affinity for cholesterol. It also acts as a central scaffolding protein for a variety of signaling complexes in the regulation of EC proliferation, migration and vascular permeability. Proteins dually acylated by saturated fatty acids are generally believed to specifically target to the caveolae/lipid rafis. This was true in our study in that several proteins known to be dually acylated were all identified in caveolae/lipid rafts from hRVE. This includes G proteins, especially G(i) 012, 3 subunits and G137 subunits. The heterotrimeric GTP-binding protein G, binds to caveolin-1 and plays a fundamental role in the mechanism of caveolae-mediated endocytosis involving Src family kinases[26, 27]. Small GTPases such as R-Ras and R-RasZ are also present in this fraction. It is worth noting that a new protein called RAIGl (retinoic acid induced protein 3), which is a type 149 3 G protein-coupled receptor, was first identified in caveolae/lipid rafts, suggesting a presence of caveolae/lipid rafts mediated retinoic acid activated GPCR signaling. Our proteomic method failed to identify several well-characterized caveolae/lipid raft localized signaling proteins such as eNOS and members of the Src family kinases. This could be due to their low expression levels in hRVE, since the presence of Fyn and c-Yes were all confirmed by western blot (Fig. 3). We were not able to detect any integral membrane receptors such as PDGFR, EGFR[28] and VEGFR2[29, 30], which were previously reported localized in caveolae in microvascular or cultured endothelial cells by biochemical analyses. This is possible since our caveolae/lipid rafts were isolated fi’om cells under serum starvation that might exclude receptors from caveolae/lipid rafts due to the resting state of the cells. Another possibility is that the expression levels of the receptors in primary cells are not sufficient for detection by both MS and biochemical analyses. Also worthy of mention is that several signaling proteins induced under serum starvation were first identified as being colocalized with caveolae/lipid rafts in hRVE. This includes serum deprivation response (sdr) and HSRBC. The former is a substrate of PKCa which is highly concentrated in caveolae[31]; the latter binds to protein kinase C6 under serum starvation[32]. This suggests a possible role of caveolae/lipid rafts in PKC mediated serum starvation induced control of cell growth. The exact presence and role of PKC, Sdr and HSRBC in caveolae/lipid rafts need to be further characterized. GPI-linked surface proteins partition into the exoplasmic membrane leaflets of caveolae/lipid rafts and mediate compartmentalized signaling. They have been claimed 150 to be mainly present in lipid rafts with a dynamic assembly, with or without ligand binding. A group of GPI-linked glycoproteins were identified including GPI-protein alkaline phosphatase and 5’-nuc1eotidase precursor (CD73). CD109, a protein belonging to alpha2 macroglobulin/complement C3, C4, C5 gene family (AMCOM), and CD59 glycoprotein precursor, which inhibits the complement membrane complex (mac) action are all identified by in-gel digestion and MS. Interestingly, several complement components such as complement C3, C9 precursors were all found in the caveolae/lipid rafts fraction suggesting that caveolae/lipid rafts may act as platforms for concentrating molecules regulating complement complex recruitment, activation and function. A connection between caveolae and cytoskeleton has been well described. Our finding that low-density detergent insoluble fraction in hRVE cells contain actin and actin binding membrane skeletal proteins (such as myosin-HA) is consistent with previous reports in various cells[19, 21, 24]. The identification of all of the known components of the myosin motor (heavy chain, alkali and regulatory light chains) indicates a possible function of caveolae/lipid rafts in the regulation of cytoskeleton reorganization. The presence of tubulin in caveolae/lipid rafts, also reported in Hela cells[24], could have important implications for membrane restructuring since cytosolic tubulin is thought to be added to the plus ends of the microtubules within lipid rafts to maintain the length of the microtubules associating with the plasma membrane. All of this suggests caveolae/lipid rafts are high dynamic microdomains that are associated with cellular microtubule transport machinery to mediate the endocytosis or transcytosis. This notion was further confirmed by the identification of proteins involved in the molecular transport machinery for vesicle budding, docking and fusion including annexins and 151 GTPase etc. Also worthy of mention is the identification by MS of the type-H transmembrane protein p63 (CKAP4, cytoskeleton-associated protein 4), a member of paramyosin family. The specific role of p63 in microvascular endothelial cells was not characterized with a possible role of binding to tPA to mediate plasmogen activation in vascular smooth muscle cells[33]. Involvement of caveolae in cell interaction with the extracellular matrix is through regulation of matrix degradation, an essential process in mediating endothelial cell migration and angiogenesis. Specific MMPs such as MMP] and 2 were all reported to colocalize with caveolae in tumor cell lines[34, 35]. In hRVE, membrane alanine aminopeptidase precursor (ANPBP), a membrane zinc-binding MMP superfamily, is concentrated in caveolae/lipid rafts fractions along with several other proteins modulating matrix metalloprotease secretion and activity, such as GPI-anchored protein reversion- inducing cysteine-rich protein with Kazal motifs (RECK), a membrane anchored inhibitor of MMPs involved in inhibiting EC migration[36]. Recent evidence also suggests that CD147, a regulator of MMP production, specifically associates with caveolin-1 which diminishes CD147 MMP-inducing activity[37]. Furthermore, calcium dependent cell adhesion molecules such as cadherin-13 precursor, and protocadherin 10 precursor were all identified in this special compartment. Taken together, matrix- degrading enzymes or their regulators and cell adhesion receptors are enriched in a limited microenvironment (caveolae/lipid rafts) at the cell surface to mediate EC transmigration. 152 Caveolae/lipid rafts are also shown to be sites for concentration of ion channels that regulate endothelial cell permeability and receptor mediated endocytosis, etc. Voltage-dependent anion-selective channel proteins and vacuolar (H+)—ATPase subunits are all identified in caveolae/lipid rafts of hRVE. These ion-channel proteins mainly exist in mitochondria, but also in plasma membrane especially caveolae to regulate the intracellular acidity as well as plasminogen activation in EC[38, 39]. The transient receptor potential cation channel, subfamily M, member 4 is also present in caveolae/lipid rafts acting as cation channel regulating ion transport. Two forms of integral membrane calcium channels are identified as indicated in Table I suggesting that caveolae may mediate calcium signaling. The presence of ion-channel proteins in caveolae/lipid rafts implicates specific involvement of caveolae/lipid rafts to regulate cation transport and thus permeability and various other functions. Surprisingly, glucose transporter 1 (Glut-1) and monocarboxylate transporters are also present in caveolae/lipid rafts. Both Glut-1 and monocarboxylate transporters, a large family of proton-driven transporters possessing 12 membrane spanning domains, were previously detected in the lipid rafts fraction from both Hela[22] and the monocytic cell line THP-l cells[21]. Like glucose, the monocarboxylates lactate and pyruvate play central roles in retinal cellular metabolism as important energy sources. As stated earlier, CD147, an accessory protein which determines their trafficking, subcellular localization and functional expression was also identified indicating a role of caveolae/lipid rafts in the trafficking and localization of the transporters in endothelial cells and in regulating of energy uptake[40]. This is of particular interest in hRVE cells since hRVE forms an 153 inner blood retinal barrier to serve as an active interface to provide energy and support retinal functions. An interesting finding is that CD44, a phagocytic glycoprotein, is abundantly enriched in caveolae/lipid rafts in hRVE. CD44 is an alpha-helical integral membrane that is recognized as a major cell surface receptor for hyaluronic acid[4l]. CD44 is also involved in different physiological and pathophysiological processes including inflammation and tumor metastasis[42, 43]. Western blot using anti-CD44 confirmed its presence (data not shown). The specific functions of CD44 involving caveolae/lipid rafts in hRVE wait to be further investigated. Due to the high lipid content and existence of hydrophobic membrane proteins in the caveolae/lipid rafts fractions, only a limited number of peptide peaks were detected from in-solution digestion methods. It is believed that this is mainly due to the signal suppression during the ionization and detection processes and also the complexity of the tryptic peptides which were unable to be fully separated in the one dimensional HPLC we applied in our study. It is not possible to discern how much of this data reflects plasma membrane caveolae/lipid rafts, as the detergent insoluble fraction was isolated fi'om whole cells, without excluding intracellular membranes. Previous biochemical isolations proposed the presence of rafts in other membrane compartments. Clearly, both rafts lipids and proteins are synthesized in the reticulum/Golgi before transport to the plasma membrane. Furthermore, a direct transport route from caveolae to ER exists in endothelial cells[8]. All of this verifies our identification of many synthesis proteins and vesicle trafficking 154 proteins in the caveolae/lipid rafts fractions isolated from hRVE. Contamination by abundant cytosolic proteins is also a common problem encountered during proteomic analysis of subcellular organelles. A recent analysis using agents that deplete cholesterol to disrupt caveolae/lipid rafts identified about 37% proteins unaffected suggesting that they probably represent non-raft proteins[22]. In summary, of the 70 proteins we identified here, about 25 proteins are possible contaminants from either different organelles or cholesterol nondependent proteins. Some atypical endothelial proteins such as a-actin, a molecular marker for pericytes were also detected. This is explainable, since our preparation of primary hRVE cells can not ensure 100% purity, although above 90% purity is always achieved. Taken together, our study for the first time systematically characterized the protein components of caveolar/lipid rafts in human retinal endothelial cells using proteomic approaches. The findings of numerous proteins provide compelling evidence for important roles of endothelial caveolae/lipid rafts in regulating the functions of inner blood retinal barrier. The compartrnentalization of important signal transducers, effectors and receptors in caveolae/lipid rafts suggest critical roles of caveolae/lipid rafts as functional platforms for the organization and coordination of signaling pathways involved in retinal endothelial cell growth, migration and especially inflammation. Acknowledgement: The LCIMSIMS experiments were performed by Doug Whitten in MSU genomics facility. 155 6. References 10. ll. 12. 13. Yamada, B., The fine structure of the renal glomerulus of the mouse. J Histochem Cytochem, 1955. 3(4): p. 309. Liu, P., M. Rudick, and R.G. Anderson, Multiple functions of caveolin-1 . J Biol Chem, 2002. 277(44): p. 41295-8. Razani, B. and MP. Lisanti, Caveolins and caveolae: molecular and functional relationships. Exp Cell Res, 2001. 271(1): p. 36-44. Razani, B., S.E. Woodman, and MP. Lisanti, Caveolae: from cell biology to animal physiology. Pharmacol Rev, 2002. 54(3): p. 431-67. Carver, LA. and J.E. Schnitzer, Caveolae: mining little caves for new cancer targets. Nat Rev Cancer, 2003. 3(8): p. 571-81. Frank, P.G., et al., Caveolin, Caveolae, and Endothelial Cell Function. Arterioscler Thromb Vasc Biol, 2003. 23(7): p. 1161-1168. Matveev, S., et al., The role of caveolae and caveolin in vesicle-dependent and vesicle-independent traflicking. Adv Drug Deliv Rev, 2001. 49(3): p. 237-50. Simionescu, M., A. Gafencu, and F. Antohe, T ranscytosis of plasma macromolecules in endothelial cells: a cell biological survey. Microsc Res Tech, 2002. 57(5): p. 269-88. Anderson, R.G., Caveolae: where incoming and outgoing messengers meet. Proc Natl Acad Sci U S A, 1993. 90(23): p. 10909-13. Sargiacomo, M., et al., Signal transducing molecules and glycosyl- phosphatidylinositoI-linked proteins form a caveolin-rich insoluble complex in MDCK cells. J Cell Biol, 1993. 122(4): p. 789-807. Lisanti, M.P., M. Sargiacomo, and RE. Scherer, Purification of caveolae-derived membrane microdomains containing lipid-anchored signaling molecules, such as CPI-anchored proteins, H-Ras, Src-family tyrosine kinases, eNOS, and G-protein alpha-, beta-, and gamma-subunits. Methods Mol Biol, 1999. 116: p. 51-60. Drab, M., et al., Loss of caveolae, vascular dysfunction, and pulmonary defects in caveolin-1 gene-disrupted mice. Science, 2001. 293(5539): p. 2449-52. Razani, B., et al., Caveolin-I null mice are viable but show evidence of hwerproliferative and vascular abnormalities. J Biol Chem, 2001. 276(41): p. 3 8 12 1 -38. 156 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Schubert, W., et al., Microvascular hyperpermeability in caveolin-I (-/-) [mock- out mice. Treatment with a specific nitric-oxide synthase inhibitor, L-name, restores normal microvascular permeability in Cav-I null mice. J Biol Chem, 2002. 277(42): p. 40091-8. Cohen, A.W., et al., Role of caveolae and caveolins in health and disease. Physiol Rev, 2004. 84(4): p. 1341-79. Shaul, P.W., Endothelial nitric oxide synthase, caveolae and the development of atherosclerosis. J Physiol, 2003. 547(Pt l): p. 21-33. Hilhnan, N., et al., Increased numbers of caveolae in retinal endothelium and pericytes in hypertensive diabetic rats. Eye, 2001. 15(Pt 3): p. 319-25. Cohen, A.W., et al., Role of caveolin and caveolae in insulin signaling and diabetes. Am J Physiol Endocrinol Metab, 2003. 285(6): p. E1151-60. von Haller, P.D., et al., Mass spectrometric characterization of proteins extracted from Jurkat T cell detergent-resistant membrane domains. Proteomics, 2001. 1(8): p. 1 010-2 1 . Nebl, T., et al., Proteomic analysis of a detergent-resistant membrane skeleton from neutrophil plasma membranes. J Biol Chem, 2002. 277(45): p. 43399-409. Li, N., et al., Lipid raft proteomics: analysis of in-solution digest of sodium dodecyl sulfate-solubilized lipid raft proteins by liquid chromatography-matrix- assisted laser desorption/ionization tandem mass spectrometry. Proteomics, 2004. 4(10): p. 3156-66. Foster, L.J., C.L. De Hoog, and M. Mann, Unbiased quantitative proteomics of lipid rafts reveals high specificity for signaling factors. Proc Natl Acad Sci U S A, 2003. 100(10): p. 5813-8. Blonder, J., et al., Proteomic analysis of detergent-resistant membrane rafts. Electrophoresis, 2004. 25(9): p. 1307-18. Sprenger, R.R., et al., Comparative proteomics of human endothelial cell caveolae and rafts using two-dimensional gel electrophoresis and mass spectrometry. Electrophoresis, 2004. 25(1): p. 156-72. Chen, W., et al., Dyslipidemia, but not hyperglycemia, induces inflammatory adhesion molecules in human retinal vascular endothelial cells. Invest Ophthalmol Vis Sci, 2003. 44(11): p. 5016-22. 157 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. Minshall, R.D., et al., Vesicle formation and traflicking in endothelial cells and regulation of endothelial barrier function. Histochem Cell Biol, 2002. 117(2): p. 105-12. Shajahan, A.N., et al., Gbetagamma activation of Src induces caveolae-mediated endocytosis in endothelial cells. J Biol Chem, 2004. 279(46): p. 48055-62. Lisanti, M.P., et al., Characterization of caveolin-rich membrane domains isolated from an endothelial-rich source: implications for human disease. J Cell Biol, 1994. 126(1): p. Ill-26. Labrecque, L., et al., Regulation of Vascular Endothelial Growth Factor Receptor-2 Activity by Caveolin-l and Plasma Membrane Cholesterol. Mol. Biol. Cell, 2003. 14(1): p. 334-347. Cho, C.H., et al., Localization of VEGFR-Z and Phospholipase D2 in Endothelial Caveolae is Involved in VEGF-induced Phosphorylation of MEK and ERK. Am J Physiol Heart Circ Physiol, 2004. Mineo, C., et al., Targeting of protein kinase Calpha to caveolae. J Cell Biol, 1998. 141(3): p. 601-10. Izumi, Y., et al., A protein kinase Cdelta-binding protein SRBC whose expression is induced by serum starvation. J Biol Chem, 1997. 272(11): p. 7381-9. Razzaq, T.M., et al., Functional regulation of tissue plasminogen activator on the surface of vascular smooth muscle cells by the type-II transmembrane protein p63 (CKAP4). J Biol Chem, 2003. 278(43): p. 42679-85. Annabi, B., et al., Localization of membrane-type 1 matrix metalloproteinase in caveolae membrane domains. Biochem J, 2001. 353(Pt 3): p. 547-53. Puyraimond, A., et al., MMP-2 colocalizes with caveolae on the surface of endothelial cells. Exp Cell Res, 2001. 262(1): p. 28-36. Oh, J ., et al., Tissue inhibitors of metalloproteinase 2 inhibits endothelial cell migration through increased expression of RECK. Cancer Res, 2004. 64(24): p. 9062-9. Tang, W. and ME. Hemler, Caveolin-I regulates matrix metalloproteinases-I induction and CD147/EMMPRIN cell surface clustering. J Biol Chem, 2004. 279(12): p. 11112-8. Gonzalez-Gronow, M., et al., The voltage-dependent anion channel is a receptor for plasminogen kringle 5 on human endothelial cells. J Biol Chem, 2003. 278(29): p. 27312-8. 158 39. 40. 41. 42. 43. Nishi, T. and M. Forgac, The vacuolar (H +)-A T Pasesunature's most versatile proton pumps. Nat Rev Mol Cell Biol, 2002. 3(2): p. 94-103. Enerson, BE. and LR. Drewes, Molecular features, regulation, and function of monocarboxylate transporters: implications for drug delivery. J Pharm Sci, 2003. 92(8): p. 1531-44. Oliferenko, S., eta1., Hyaluronic acid (HA) binding to CD44 activates Rae] and induces lamellipodia outgrowth. J Cell Biol, 2000. 148(6): p. 1159-64. Oliferenko, S., et al., Analysis of CD44-containing lipid rafts: Recruitment of annexin II and stabilization by the actin cytoskeleton. J Cell Biol, 1999. 146(4): p. 843-54. Annabi, B., et al., Hyaluronan cell surface binding is induced by type I collagen and regulated by caveolae in glioma cells. J Biol Chem, 2004. 279(21): p. 21888- 96. 159 V. Inhibition of retinal endothelial cell inflammatory response: Modification of caveolae/lipid rafts by Docasahexanoic acid (DHA22;5,,3) treatment 1. Abstract An early stage diabetic retinopathy has been recently recognized as a low-grade chronic inflammatory disease involving increased cytokine production, intercellular adhesion molecule (ICAM-1) and vascular cell adhesion molecule (V CAM-1) expression and subsequent leukocyte adhesion and blood retinal barrier breakdown. A decrease in a major n3 PUFA, docosahexaenoic acid (DHA22:6n3) in the retina of animal models of diabetes was reported. DHAzum inhibits cytokine induced ICAM-1 and VCAM-1 expression in human retinal vascular endothelial cells (hRVE). However, the mechanisms underlying anti-inflammatory effect of DHA22:6n3 in hRVE cells remain unresolved. The possibility that DHA22;(,,.3 acts through modifying the lipid composition resulting in change of the affinity of important signaling molecules for a specialized plasma membrane microdomain, the caveolae/lipid rafts, was addressed. We show that the structure of caveolae/lipid rafts as well as its components Src family kinases (SFK) are involved in mediating cytokine induced VCAM-1 expression. Treatment of hRVE cells with DHA22:6n3 resulted in significant displacement of the SFKs, Fyn and c-Yes, from caveolae/lipid rafts. Notably, enrichment of DHA22:6n3 in endothelial cells resulted in significant incorporation of DHA22:6n3 into major phospholipids (PC, PE and PI) that predominantly reside in both exoplasmic and cytoplasmic leaflets of caveolae/lipid rafts which subsequently increased the unsaturation ratio of this specialized high lipid-ordered 160 membrane microdomain. Moreover, DHA22:6n3 enrichment also caused about 70% depletion of cholesterol from caveolae/lipid rafts. DHA22:6n3 modification of fatty acyl chains of phospholipids in caveolae/lipid rafts followed by cholesterol depletion could effectively perturb the lipid environment of caveolae/lipid rafts resulting in the displacement of important signaling molecules thus an immunomodulatory effect in endothelial cells. 161 2. Introduction Polyunsaturated fatty acids (PUFAS), particularly n3-PUFAs such as eicosapentaenoic acid (EPAzosm) and docosahexaenoic acid (DHA22:6n3) have long been demonstrated to be immunoregulatory by suppressing the activation of immune cells and the production of inflammatory cytokines and adhesion molecules[l, 2]. Numerous clinical studies have shown that dietary supplementation with fish oil abundant in n3- PUFAs has beneficial effects in the treatment of inflammatory disorders such as inflammatory bowel diseases, asthma, arthrosclerosis and cardiovascular diseases and thus are widely applied clinically as adjuvant immunosuppressive agents[3]. Several mechanisms have been linked to the immunosuppressive functions of n3- PUFAs. The replacement of AA20;4n6 due to incorporation of n3-PUFAs into membrane phospholipids reduces the amount of AA20;4,.6 available for oxygenases. This will decrease the production of inflammatory eicosanoids, such as series 1 and 2 thromboxanes and prostaglandins, series 4 leukotrienes, and hydroxy and epoxy fatty acids [4]. Recent discovery of potent anti-inflammatory eicosanoids: E series and D series resolvins derived from EPA20;5n3 and DHA22;6n3 respectively by COX-2, constitutes a novel mechanism(s) for the therapeutic anti-inflammatory benefits of n-3 dietary supplementation[5-7] and reviewed in [8]. Moreover, n3-PUFAs could also directly bind to and activate nuclear receptors such as peroxisome proliferator-activated receptor (PPAR) that are shown to play an anti-inflammatory role in various cells (reviewed in [9])- 162 In vitro studies have strongly demonstrated that the inhibitory effects of n3- PUFA on T cell signal transduction are primarily due to eicosanoids independent mechanisms, particularly the modification of functional membrane lipid microdomains called lipid rafts[10]. Lipid rafts are specialized micromembrane domains highly enriched with cholesterol and sphingolipids such as Sphingomyelin and glycolipids. Polar lipids in lipid rafts predominantly contain saturated fatty acyl residues that aggregate spontaneously to form liquid-ordered membrane regions insoluble in non-ionic detergents[11, 12]. Lipid rafts are essential in lymphocyte signal transduction[l3, 14]. Treatment of cultured T cells with PUFAs selectively displaces acylated proteins such as LCK and LAT from the cytoplasmic leaflet of lipid rafts[15, 16] thus inhibiting the T cell activation and proliferation. This displacement of important signaling molecules is due to the incorporation of n3-PUFAs into the lipids of both exoplasmic and cytoplasmic leaflets of lipid rafts[15]. Dietary n3-PUFA or high purified DHA22:6n3 treatment in vivo has also demonstrated a significant displacement of important signaling molecules concomitant with decreases in cholesterol or Sphingomyelin contents in mouse colon or splenic T cell lipid rafts due to n3-PUFAs incorporation[17-20] Vascular endothelium is the interface between blood elements and tissues and plays a vital role in immune responses especially inflammation. Caveolae, a specialized micromembrane domain which shares similar lipid compositions as lipid rafts are highly abundant in endothelial cells[21]. Caveolae are stabilized by a family of structural and regulatory proteins, caveolin, that gives caveolae its characteristic flask-like shape. Caveolae have been shown to be important players in regulating vascular permeability, vesicle trafficking, cholesterol homeostasis and particularly signal transduction[21-25]. 163 A number of cytokine receptors have been found localized in caveolae/lipid rafts, such as TNFR1[26, 27] and VEGFR2 in endothelial cells[28, 29]. Our previous study has demonstrated that n3-PUFAs, particularly DHA22:6n3 inhibits cytokine induced inflammatory response in primary human retinal endothelial cells (hRVE). However, the molecular pathways underlying the inhibition are not well resolved in endothelial cells. In this study we examined the role of caveolae/lipid rafts in inflammatory signaling in hRVE cells. The effect of DHA22:6n3 treatment on localization of SFKs and modification of lipid composition of caveolae/lipid rafts were further addressed. Our data provide strong evidence that caveolae/lipid rafts are required for endothelial cell inflammatory responses. Changes in fatty acyl chains of phospholipids and cholesterol displacement fiom caveolae/lipid rafts induced by DHA22:6n3 could represent a principal mechanism for the anti-inflammatory effect of DHA22:6n3 in hRVE cells. 164 3. Materials and Methods 3.1. Reagents and antibodies HPLC grade acetonitrile, acetic acid, methanol, chloroform and methyl- cyclodextrin (MCD) were purchased from Sigma. 4-amino-5-(4-chlorophenyl)-7-(t- butyl)pyrazolo[3,4-d]pyrimidine (PP2) was obtained from Calbiochem. The following antibodies were used: Mouse anti-caveolin-l and flotillin-l were fi'om Upstate Biotechnology, Inc. (Lake Placid, NY); mouse anti-c-Src, Fyn, rabbit antibodies against c-Yes, and CD36 were purchased from Santa Cruz. 3.2. Cell culture and fatty acid treatment Primary cultures of hRVE cells obtained from at least three donors were prepared and cultured as previously described[30]. Passages 1-6 were used in the experiments. Primary human umbilical vein endothelial cells (HUVEC, multiple donors) were obtained fi'om Cascade Biologicals (Portland, OR) and cultured in DMEM containing 10% FBS, macrovascular endothelial cell growth supplement (MVGS) and 100 rig/ml antibiotics/antimycotics in a humidified incubator at 37 °C with 5% C02. For experimental treatments, cells were transferred to serum-free medium for 18 to 24 hours before addition of the stimulatory agents. Treatment of cells with fatty acids was performed as follows. Fatty acid stocks were prepared by dissolving fatty acids (NuCheck Prep, Inc., Elysian, MN) in 100% ethanol to a final concentration of 100 mM fatty acid as described previously. The fatty acid stock solutions were diluted in serum- free medium to reach fatty acid concentrations of 100 uM with corresponding bovine 165 serum albumin (BSA) concentration of 20 11M. Charcoal-treated, solvent-extracted, fatty acid—flee BSA was obtained from Serologica Inc., Norcross, GA. The fatty acid-to- albumin molar ratio was maintained at 5:1. The final concentration of ethanol in the media is less than 0.1%. Cells were incubated for the times indicated in the Results section. Equivalent amounts of BSA alone were added to control plates. 3.3. SDS-PAGE and western blot Cells were lysed in the lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 1.5 mM MgC12, 1 mM EGTA, 1% Triton X-100, 10% glycerol) with freshly added protease inhibitor cocktail (Sigma) and phosphatase inhibitors (1 mM Na3VO4, 100 M glycerophosphate, 10mM NaF, 1 mM Na4PPi). Proteins were resolved by SDS-PAGE and transferred to nitrocellulose, immunoblotted using appropriate antibodies followed by secondary horseradish peroxidase conjugated antibody (Bio-Rad). Irnmunoreactive bands were visualized by enhanced chemiluminescence (ECL kit; Amersharn Pharmacia Biotech, Piscataway, NJ). Blots were quantitated by scanning densitometry using Image] software, ver. 1.29 (available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). 3.4. Subcellular fractionation Post-nuclear supematants were prepared by lysing fatty acid treated cells in hypotonic buffer (10 mM Hepes pH 7.4, 1 mM EDTA, 1 mM MgC12 freshly added with protease inhibitors) for 30 min on ice, then dounced 20 times in a Teflon vortexer and 166 spun at 500 X g for 5 min. The supematants were subjected to centrifugation (16,900 X g, 60 min, 4°C) and the pellets were collected as total plasma membrane enriched fi‘actions (also called bulk membranes). Isolation of caveolae/lipid raft enriched detergent resistant membrane domains were prepared using a slightly modified sucrose gradient ultracentrifugation protocol [31]. Briefly, 5x106 hRVE cells were washed with cold PBS twice and then lysed in 0.8 ml MNE buffer (25 mM MES, PH 6.5, 0.15 M NaCl and 5 mM EDTA) containing 1% Triton X-100 and fresh protease and phosphatase inhibitors and kept on ice for 20 min. The homogenization was carried out with 10 strokes of a tight-fitting Dounce homogenizer, and then spun at 4,000xg 4°C for 10 min. Supernatant (0.8 ml) was then mixed with the same volume of 80% sucrose prepared in MNE buffer, and placed at the bottom of an ultracentrifuge tube. 1.6 ml of 30% sucrose and 0.8 ml of 5% sucrose were overlaid on top of the sample to form a 5-30% discontinuous sucrose gradient. After 16 h centrifiigation at 200,000xg at 4°C using a swinging bucket rotor, 0.4 ml samples were collected carefillly from the top for each fraction. A band confined to fiactions 2 through 4 was designated as caveolae/lipid rafts enriched membrane domains. The combined 2-4 fiactions were further diluted 3 times in MNE buffer and spun at 200,000xg at 4°C for another 2 h to precipitate the caveolae/lipid rafts and the pellet was designated as insoluble fiaction (I). The fractions 6-10 were also combined and designated as soluble fraction (S). 167 3.5. Fatty Acid Metabolism hRVE cells were plated at 0.12x106 cells/6 cm plate and cultured as indicated above to about 90% confluence. Cells were then serum starved for 18-24 hours and treated with ["c122:6n3 in 3 m1 of DMEM/F12 containing 100 11M 22:6n3, 0.5 uCi, (1.7 Ci/mol) for 1.5 and 24 h in the presence of 20 11M BSA. l4C-labeled fatty acids were purchased from PerkinEhner Life Sciences. The fatty acid-to-albumin molar ratio was 5:1. At harvest, cells were washed once with phosphate-buffered saline + 20 11M BSA followed by PBS alone once. Cells were then resuspended in 500 pl of 40% methanol and lipids were extracted with chloroform:methanol (2:1); dried under nitrogen and dissolved in chloroform for storage at -80 °C. Total lipids were separated by thin layer chromatography (LK6D Silica G 60A, Whatrnan) in hexanezdiethyl etherzacetic acid (90:30:1). Polar lipids were separated in chloroform:methanol:acetic acid (30:20z4). Location of lipids was compared with authentic standards for diacylglycerol (TAG), diacylglycerol (DAG), cholesterol esters (CE), fatty acids, fatty acid (wax) esters (Sigma), and glycerol- and sphingo-phospholipids (Avanti Polar Lipids). 3.6. Fatty Acid and Cholesterol Analysis Total lipids of caveolae/lipid rafts and total plasma membranes corresponding to equal amounts of protein (measured by Bradford assay (Biorad)) were extracted with chloroform-methanol [2:1], dried and stored in chloroform at —80°C. A fraction of the total lipids was further fractionated on an arnino-propyl (Alltech) column to obtain neutral lipids, neutral phospholipids, acidic phospholipids and nonesterified fatty acids (NEFA) as described[32]. The neutral lipids were separated by Waters YMC-Diol- 168 120NP 5 11M, 250x4.6 mm column followed by evaporative light scatter detection system to detect cholesterol. Pure lipid standards for cholesterol, triacylglycerol, diacylglycerol obtained from Avanti Polar Lipids were used after each experiment to confirm retention times and purity. In order to quantify the data, calibration curves were prepared for each class of lipids at the end of each experiment. A fraction of neutral phospholipids and acidic phospholipids were saponified (0.4 N KOH in 80% methanol, 50°C for 1 h) for the fatty acid composition analysis respectively. After saponification, lipids were acidified and extracted with diethyl ether with 0.1% acetic acid, dried under nitrogen and stored in methanol. Saponified lipids were fractionated and quantitated by reverse phase HPLC (RP-HPLC) using a YMC J- Sphere (ODS-H80) column and a sigrnoidal gradient starting at 86.5% acetonitrile + acetic acid (0.1%) and ending at 100% acetonitrile + acetic acid (0.1%) over 50 min with a flow rate of 1.0 ml/min using a waters 600 controller. Fatty acids were detected using both UV absorbance at 192 nm (Waters model 2487) and evaporative light scatter (Waters model 2420). Fatty acid standards for RP-HPLC were obtained fi'om Nu-Chek Prep (Eysian, MN). The relative concentration for each fatty acid was obtained by normalizing to the standards. 3.7. Mass spectrometry of phospholipids Phospholipids analyses by mass spectrometry (MS) were performed using a Thermo LTQ linear quadrupole ion trap mass spectrometer (model LTQ, Therrno- Finnigan, San Jose, CA), equipped with a nanospray ionization source (nanoESI). Total lipid extracts from each sample were dissolved in 50:50:1 methanol/chlorofomi/28% 169 ammonia hydroxide prior to introduction to the mass spectrometer by direct infusion through non-coated silica tips with internal diameters of 30 pm (New Objective, Inc. Woburn, MA) at a flow rate of 0.2 uL/min. NanoESl conditions were optimized to maximize the sensitivity and stability of the ions of lipids while minimizing “in-source” fragmentation. Typical nanoESI conditions were: heated capillary temperature 150°C, spray voltage 1.3kV, capillary voltage 20V and tube lens voltage 50V. Quantitative analysis of every sample was based on triplicate MS experiments, each being an average of 50 spectra with 3 micro-scans acquired per spectrum, in both positive and negative ionization modes under the same instrument settings. Intensities of all the ions detected in each sample were first normalized to the internal standards (dimyristoylphosphatidylcholine for positive ionization mode and dimyristoylphosphatidylethenolarnine and dimyristoylphosphatidylglycerol for negative ionization mode), and the log of the ratios of normalized intensities between lipids from cells treated with BSA, palmitic acid 16:0 or DHA22:6n3 were then plotted to determine those lipids whose abundance were affected most significantly by the different treatments. MS/MS and MS/MSfMS (MS3) data for each sample were acquired by a data-dependent tandem mass spectrometry experiment, during which an initial MS scan was followed by CID MS/MS (collision energy 30%) of the most abundant precursor ion, followed by MS3 of the three most abundant product ions obtained from the MS/MS scan. Following the MS/MS scan event, the selected precursor ion was placed on an exclusion list so that subsequent data-dependent acquisitions allowed the analysis of progressively lower- abundance precursor ions. A total of the 40 most intense precursor ions in each sample were sequentially selected for analysis. The structures of the lipids previously 170 determined to be significantly different between samples from the triplicate MS experiments were identified by interpretation of the spectra fi'om the data-dependent tandem MS/MS experiments. 171 4. Results 4.1. Caveolae/lipid rafts are involved in VEGF165 and TNFa induced CAM expression In this chapter, we used two different approaches to determine whether caveolae/lipid rafts are involved in cytokine induced CAM expression in endothelial cells. Methyl-cyclodextrin (MCD), a cholesterol-depleting agent was first used to disrupt the structure of caveolae/lipid rafts. As shown in Fig. 1, MCD pretreatment downregulated TNFci induced phosphorylation of Icht, a critical step of triggering the downstream NFKB activation to mediate cytokine induced adhesion molecules expression. In contrast, the We triggered ERK phosphorylation was generally not affected, although cholesterol depletion had an effect on the basal activation of ERK1/2 (Fig. 1). A similar effect was observed in IL-lB meditated Ichct phosphorylation in response to MCD treatment (data not shown). This suggests that the integrity of caveolae/lipid rafts is required for cytokine induced NFKB signaling and downstream cell adhesion molecule (CAM) expression. As shown in Chapter III, Src family kinases (SFK) especially Fyn and c-Yes were exclusively localized in caveoae/lipid rafts. Our second approach was to examine whether inhibition of SFKs localized in caveolae/lipid rafts would lead to a decrease of cytokine induced inflammatory response. PP2, a specific inhibitor of SFKs (c-Src, F yn, and c-Yes) was used to pretreat HUVEC for 30 min before the addition of VEGF165 for another 6 h. Fig. 2A demonstrated that PP2 prevented VEGF165 induced VCAM-l 172 expression, suggesting that Src family kinases are involved in VEGF165 mediated inflammatory signaling leading to induction of the expression of adhesion molecules. Furthermore, pretreating cells with PP2 also diminished the TNFa induced VCAM-1 expression in hRVE cells (Fig. 2B), indicating that SFKs localized in caveolae/lipid rafts are important in the cytokine induced inflammatory response. 4.2. DHAmm displaces Src family kinase from caveolae I lipid rafts in endothelial cells The dual acylated Src family kinases are targeted to the cytoplasmic leaflets of caveolae/lipid rafts due to posttranslational modification with fatty acyl moieties. The next experiment was designed to determine whether DHA22:6n3 treatment could alter the association of SFKs with caveolae/lipid rafts, thus preventing SFKs recruitment to activated cytokine receptors. Pre-treatrnent of hRVE cells with 100 11M DHA22:6n3 caused about 60% displacement of SFK Fyn from the caveolae/lipid rafts (Fig. 3A and quantitated in Fig. 3B). Displacement of c-Yes from this specialized microdomain was also obvious after DHA22:6n3 enrichment compared with BSA alone treated cells (Fig. 3C). DHA22:6n3 treatment did not cause the disruption of caveolae/lipid rafts since the major structural proteins of caveolae and lipid rafts: caveolin-1 and flotillin-l were not affected. Moreover, the localization of CD36, a type II integral membrane scavenger receptor involved in oxidized lipid and fatty acids transport, was also not changed after DHA22:6n3 enrichment, suggesting a specific displacement of Src family kinases by DHA22;5n3 enrichment. The displacement of Fyn and c-Yes was endothelial cell specific as in 173 human retinal pigmented epithelial cells (hRPE), treatment with up to 200 M DHA22:6n3 had no effect on SFKs localization in caveolae/lipid rafts (Fig. 3D). 4.3. DHAmm treatment alters fatty acyl compositions of phospholipids residing in caveolae/lipid rafts The biochemical alterations underlying selective displacement of Src family kinases by DHA22:6n3 in endothelial cells are largely unknown. An attractive possibility is that enrichment of caveolae/lipid rafts with DHA22:6n3 modifies lipid composition of the rafts resulting in a change of the affinity of acylated proteins, such as SFKs, to the caveolae/lipid rafts. The following approaches were used to determine the changes in lipid composition of caveolae/lipid rafts after DHA22:6n3 treatment. a) DHAzzsm is metabolized into phospholipids in hRVE Metabolic labeling of hRVE cells using l4C-DHA22;6n3 clearly demonstrated that DHA22:6n3 was taken up into the cell and rapidly incorporated into different intracellular lipid complexes such as DAG, NEF A, triglycerides and, to a large proportion, polar lipids (Fig. 4A). Further separation of polar lipid fiaction demonstrated the presence of several primary phospholipid subspecies, mainly in the form of phosphotidylcholine (PC), and, to a lesser extent, phosphotidylinositol (PI) (Fig. 48). Since phospholipids are important structural and signaling lipids present in caveolae/lipid rafts, the possibility that DHA22:6n3 could affect the lipid environment of caveolae/lipid rafts to displace acylated proteins is very plausible. b) DHA-mm ls incorporated into phospholipids of caveolae/lipid rafts 174 Next, we isolated caveolae/lipid rafts fractions from cells treated with BSA (carrier control), 100 M palmitatelw (Lipid control) or 100 M DHA22:6n3 and assessed the fatty acyl composition of phospholipids from caveolae/lipid rafts in parallel with plasma membranes by RP-I-IPLC (Table 1). Caveolae/lipid rafts from control BSA and palmitate treated cells were particularly enriched in saturated palmitic (16:0) and stearic (18:0) in neutral phospholipids (mainly PC and PE) and acidic phospholipids (PS, PI and PA). The monounsaturated (18:1n9) and polyunsaturated fatty acids (18:2n6, 20:4n6, and 20:5n3) were less abundant compared with bulk membranes. DHA22:6n3 treatment led to its significant incorporation not only in bulk membranes but also in caveolae/lipid rafts (Fig. 5A). The incorporation of DHA22:6n3 into neutral phospholipids was much more efficient than acidic phospholipids in both bulk membrane (26.44% vs 2.76%) and caveolae/lipid rafts (3.19% vs 0.67%) (Fig. 5A). This could be due to the fact that PC and PE are the most abundant phospholipids present in plasma membranes. The unsaturation index was significantly lower in caveolae/lipid rafts compared with bulk membranes suggesting a high lipid-ordered state in caveolae/lipid rafts. DHA22:6n3 treatment caused a considerable increase (50%) in the unsaturation index in neutral phospholipids (0.36 vs 0.18 double bonds per fatty acyl residue) in caveolae/lipid rafts compared with that of control hRVE cells, although this change is to a lesser content compared with general membranes (1.77 vs 0.55 in DHA treated vs control cells)(Fig. 5B). Palmitate treatment led to an increase of 16:0 levels in both caveolae/lipid rafts and total plasma membranes with a concomitant decrease in both 18:0 and 18:1n9 when compared with control cells. Increased level of 16:1n9, the elongated product of 16:0 was also observed in palmitate treated total plasma membranes (Table 1). The 175 unsaturation index of total membrane as well as caveolae/lipid rafts lipids was moderately decreased in palmitate treated vs. control cells. Taken together, DHA22:6n3 is effectively incorporated into phospholipids of caveolae/lipid rafts and significantly alters the lipid environment of these specialized membrane microdomains. c) DHA22;5..3 alters fatty acyl chains of phospholipids residing in both cytoplasmic and exoplamslc membrane leaflets Considering RP-HPLC analyses could only provide information about the overall fatty acyl compositions in general phospholipids species, we performed nanoESI/MS and MS/MS to identify the specific phospholipid subspecies into which DHA22:6n3 was incorporated. The total lipid extracts of caveolae/lipid rafts and bulk membranes fi'om cells treated with BSA, palmitic (16:0) and DHA22:6n3 were analyzed in parallel by nanoESI/MS according to mass to charge (m/z) values. Typical MS profiles were shown for total plasma membranes (Fig. 6A: positive mode; 6B: negative mode). The major phospholipid species identified were listed as in Table 2. Overall, in bulk membranes, substitutions of phospholipids by DHA22;¢,.,3 in the fatty acyl moieties were observed mainly with PC and possibly with P1 and PS. Accordingly, PC 38:6 (16:0/22:6, m/z 806) was markedly more abundant compared with control (BSA) and palmitate treated bulk membranes suggesting the most probable substitution by DHA22:603, Palmitate treatment also leads to an increase of PC species of 32:0 (16:0/16:0) and 34:1 (16:1/18:0 or 18:1/16z0). DHA22:6n3 treatment did not significantly affect the level of Sphingomyelin (SM) present in bulk membranes and no incorporation of DHA22:6n3 into Sphingomyelin was observed since SM is only present with16:0 at its 2-amino fatty acid group (m/z 703). 176 “h Overall, enrichment of DHA22s6n3 resulted in considerable incorporation of DHA22:6n3 into phospholipids residing in exoplasmic (PC) as well as cytoplasmic (PS and PI) leaflets of membranes. The alterations of phospholipids in the total plasma membranes could unequivocally affect the phospholipids components in caveolae/lipid rafts that is now under intensive study. 4.4. DHAmm enrichment causes cholesterol depletion in caveolae/lipid rafts The introduction of bulky chains such as DHA22:6n3 into fatty acyl chains of phospholipids which normally contain highly packed saturated fatty acids could possibly affect the interaction of fatty acyl moieties with cholesterol in both cytoplasmic and exoplasmic leaflets of caveolae/lipid rafts. As such, incorporation of DHA22:6n3 into phospholipids might affect the cholesterol partition in caveolae/lipid rafts. Indeed, Fig. 7 shows that the cholesterol levels in caveolae/lipid rafts are about 10 times higher than total plasma membranes confirming the idea that caveolae/lipid rafts are highly enriched in cholesterol. However, DHA22:6n3 enrichment decreased the cholesterol level in caveolae/lipid rafts by about 70% compared with BSA treated control, while palmitic acid 16:0 had no significant impact. Moreover, the depletion of cholesterol induced by DHA22:6n3 only occured in caveolae/lipid rafts with no significant effect on total cholesterol levels in bulk membranes. 177 Ctrl MCD A A . f N f \ IB TNFa (min) 0 5 10 0 5 10 - a _ H _p—v-b P- IKB a =====3== P-Erk1/2 ~’~—— - - Actin Fig. l. Methyl-B-cycloxydextrin (MCD) pretreatment disrupts TNFa induced NFKB signaling in hRVE cells. hRVE cells were serum starved overnight then pretreated with 8 mM MCD for 30 min before addition of We (20 ng/ml) for the indicated time. Cells were then harvested and same amounts of protein were loaded for SDS-PAGE and western blot analysis. Representative results were presented fiom two independent experiments using hRVE cells from two independent donors. 178 - - _ 428 Q6 _ hh-Fgglg VCAM-1 M, Adi” B. Tlchr r \ PP2, - - - + + 18 --E23 VCAM-1 tease-.1 is intuit. ! “ P '7 "’ ICAM‘]. MW Actin Fig. 2. Src family kinases are involved in VEGF165, TNFa induced VCAM-1 expression. (A) HUVEC cells were pretreated with pp2 (10 M) for 30 min and stimulated with VEGF165 (20 ng/ml) for 6 h. (B) hRVE cells were pretreated with pp2 (10 11M) for 30 min and stimulated with TNFa (5 ng/ml) for 6 h. Cells were then harvested. The same amounts of proteins were loaded for western blot analysis against VCAM-l, ICAM-1 and actin. Representative results from 2 independent experiments were presented. 179 ' BSA 16:0 DHA(22:6) ,_A—‘ fi 4 w r A m IB Fyn 0036 C-Src Caveolin-l Flotillin-l 2-1 a. 0.51 | 0 BSA 16:0 DHA C. BSA DHA (22:6) 1 s I S m 22.—:2 Caveolin-l 1 , _ Flotillin-l D. BSA DHA (22:6n3) r—L—m F—A—n I s I S E3. - ’ Fyn ‘ - d Flotillin-l - - Caveolin-l 180 Fig. 3. Specific displacement of Src family kinase Fyn and c-yes from caveolae/lipid rafts by DHA treatment in hRVE. (A) hRVE cells were serum starved overnight and treated with BSA (control), 100 M BSA bound 16:0 or 22:6n3 for 24 h. Caveolae/lipid rafts were isolated and analyzed by western blot. The amounts of Fyn localized in caveolae/lipid rafts fiom 4 independent experiments were quantitated and normalized to the levels in BSA treated samples as in B. I"P<0.005. (C) hRVE cells were treated with BSA (control) and 100 uM BSA bound 22:6n3 as above. The displacement of c-Yes was analyzed by western blot and a representative result was presented from 2 independent experiments. (D) hRPE cells were treated as BSA (control) and 200 11M BSA bound 22:6n3 as above. The displacement of F yn was analyzed by western blot and a representative result was presented fi'om 3 independent experiments. 181 ll u:- ;-. .- B 1 ‘ . a f . I" f I. . ' I t . I “' ck ~_‘ :.;t- . .4 '.'\!’fi I‘v‘ .‘I‘J "i. "t ‘5 . $3”7-2‘1‘-3t1 “1‘. ' e'gl). .J . fl ' s\ " ‘2 f ‘ q ‘0‘}: '39.".9 .’ so. I': . ' ;“ ‘~o:' ’5': |{T" ’1‘... :f’t ' was» ‘- - .lt- . 4'. z. ‘- ‘ D . }"1. lo-_, \ .' i. .' ..- '-V'\‘ ' If "h‘ n"’{‘;' _. ' . l ._ ' a : l .. h» f , 4 ~-,- - -. -. :1.» . :-.-c '3 f " Vi ' a 'r“ 0:: ‘ ‘ . ' g ‘ ’ l ‘ ." ( ".1'" " p 5‘. c A ,-~-. 1' . k . t \ ,'. a .‘ L. ' 1.. ‘1 ‘- , a , I .l .‘:‘- :4 1‘ ‘P‘ o u f“ . ‘ . '- _~. . .Iv . ' ' “..' W. . .1 0' I' \" ’ I W. 6‘ “. '0‘, ‘\_ ‘ .‘ r.)‘ p a- ,. - ~ -- . ., ’ .o 6‘ o ‘. '1 \ ‘."_ - ~;’ I . t. o. f‘_ .1 ' v .". 0 i - 0" r , . . l - s; . v a ' . .l e ‘," n .‘t. r“"" 1”, :5‘fi ‘ ‘4 I'..- 01" I'. ‘ f "It!!!" 1‘ y o .I“ ' 'I'.-' h' a P" i . '1 ° .1 wad 11 _ d ‘6" V a 0 " . '. .‘ r‘ ‘ m n . .. . . 3 t" '-'o‘ r ~.-"“‘s.s~.: , I" ‘\ ". I . ‘ 0". .'~'I ‘ ’ g...- n'|' $‘1"N . ' " . l . 'V ,-.‘u "I 0 '~ .. ’. - ‘3 " .' -' ,_-- 9,. 1‘ 0- I 51 ,I a , ' P . ‘ . ’ K. ' . - t ‘i‘ I) d. I I ' ‘ir’;k'~" I .a \“ ‘ ‘\ ‘ ',\ ') ‘5 r.) ‘ I "" ‘ J“"" J .' 5.. 4.. ‘1 If 5A.: , f .w- as, is, ,> .--.--i . ‘IO '1'... I " ‘v‘. ' . " w I . . ' . ' a 5‘ ell" O L . \ , a '5 , .. 1’ w, ' If ”"th. r .I" ‘ ’4‘ ‘-.'v\.e’. .f l. ,. “ T , "O- ‘ ’, ‘ .' ' u . .I ,‘ .'.‘ t. {I {“421 ' . “V ‘ ‘ I. ”0“ ‘ r. r E: ..'| A fr if. 3 .‘.;~ “panic,“ s." DAG ""5"" '. rig: (~1 i. c: . 0.1. _ . I, ‘-\9 "‘ >§ .9 a '3 . : PL 3"“ Q -' V :- iv J )- (h) 1.5 ” 4 Fig. 4. Incorporation of l4C-22:6n3 into different lipid complexes in hRVE cells. (A). Serum starved hRVE cells were treated with 14C-22:6n3 and labeled for the time periods as described in methods. Total lipids were extracted and submitted for TLC analyses. (B) The polar lipids were extracted from TLC plates and subjected for another TLC to separate different phospholipids subspecies. 182 TABLE 1. Fatty acid composition of phospholipids from caveolae/lipid rafts and general plasma membranes of hRVE cells treated with control (BSA), lipid control (16:0) and n3-PUFA (22:6n3) Fatty BSA 16:0 22:6n3 acids Rafts Membranes Rafts Membranes Rafts Membranes Neutral phospholipids (% of total lipids) 20:5n3 0.20 0.29 0.07 0.14 0.49 0.15 18:3n3 0.04 0.08 0.03 0.02 0.15 0.01 18:3n6 0.03 0.06 0.01 0.02 0.00 0.01 22:6n3 0.33 1.09 0.25 0.81 3.19 26.37 _ 20:4n6 0.73 4.80 0.50 3.37 0.67 1.41 . " 22:5n3 0.10 0.39 0.07 0.33 0.17 0.15 18:2n6 0.31 1.05 0.08 0.85 0.46 0.61 20:3n6 0.33 1.08 0.18 0.72 0.28 0.40 20:3n9 0.14 0.23 0.06 0.14 0.00 0.09 16: 1n9 0.00 0.85 0.00 1.39 0.00 0.26 16:0 76.54 46.82 85.06 66.18 67.10 46.34 18:1n9 10.07 19.56 6.95 10.49 8.80 8.79 18:0 11.44 23.69 6.85 15.58 19.33 15.40 Acidic phospholipids (% of total lipids) 20:5n3 0 0.59 0 0.14 0 0.80 18:3n3 0.27 0.22 0.07 0.09 0.1 1 0.99 18:3n6 0.00 0.07 0.02 0.02 0.02 0.10 22:6n3 0.26 0.67 0.29 0.40 0.67 2.72 20:4n6 0.60 1.56 0.62 1.39 0.06 0.30 22:5n3 0.11 0.19 0.15 0.15 0.025 0.14 18:2n6 0.43 0.42 0.34 0.49 0.10 0.41 20:3n6 0.46 0.68 0.37 0.77 0.09 0.46 20:3n9 0.06 0.18 0.12 0.00 0.01 0.00 16:1n9 0 0 0 0 0 0 16:0 39.71 36.25 46.23 57.83 37.77 33.54 18:1n9 3.51 6.83 2.93 5.82 0.61 7.20 18:0 53.62 52.33 48.08 32.89 59.80 53.33 hRVE cells from 3 different donors were treated with vehicle control (20 11M BSA), lipid control (100 11M 16:0) or DHA (100 uM 22:6n3). Total plasma membranes and caveolae/lipid rafts were prepared. Samples corresponding to equivalent amounts of protein were extracted for total lipids followed by separation of neutral phospholipids and acidic phospholipids on amino-propyl column in parallel. The fatty acids composition and amounts after saponification were analyzed by RP-HPLC as described in Methods. The representative fatty acid composition from one donor was presented and expressed in % of total lipids in each phospholipid fraction. 183 caveolaellipid Bulk membranes I DHA 1:1 Palmitate [:1 BSA Bulk membranes caveolaellipid rafts DHA% of total fatty acids I DHA I Palmitate [:1 BSA 1 .77 Unsaturation index 184 Fig. 5. DHA22:6n3 alters fatty acyl compositions of phospholipids in caveolae/lipid rafts and bulk membranes. hRVE cells were serum starved overnight and treated with BSA (control), 100 11M BSA bound 16:0 or 22:6n3 for 24 h. Cells were lysed and same amounts of protein were loaded to isolate caveolae/lipid rafts. 2-4 fiactions collected from the gradient were combined and submitted to total lipids extraction and amino- propyl column fractionation. Neutral phospholipids and acidic phospholipids were saponified followed by RP-HPLC analyses. The relative concentration for each fatty acid was obtained by normalizing to the standards and the percentage was calculated according to the total amounts in each fraction. The unsaturation ratio was calculated by the average number of double bonds per fatty acyl residue. A representative set of data from one donor was presented fi'om three independent experiments using cells derived from three different donors. Panel A: the DHA% of total fatty acids in each fraction; panel B: the unsaturation index of phospholipids. 185 “L 3 '8 8 1. s. s l .— . 4 126901 ‘ <1: . 1* ~ 210901 ‘1- t l, jr 26666 , 1"- l" 91266 1 1» 02.9% 21996 1 l» . 1 8.0.1 1' — 02006 82188 1 § § § 88728 ' 82928 f 03949 ‘ 83.999 “'5 178298 1 82.798 ' ”9999 _‘ 82:88 a 28928 -‘ . 82288 . vszB . 82018 p, 82908 —.1 . -‘ -l N 3899/- “j. § 28882 § . 4, § ‘ 3898‘ -i 28282 _' as?“ . E A - 28092 1 28'892’29 1 egg-992 - 28892 ‘- 28'v82 - 28282 .; 28:2; . 82082 28282 8 802 , 82902 ' 82802 2 . 3290; 93969 ‘7 § 28969 1 § 82969 1 § gang 3 821729 _j 35999 _fl‘ 0.) 73.999 Qz'm 93171.9 '0 O E 9V9” 02989 91989 E 172129 82189 g a, 36609 96609 ' 96609 c . g g 8 § 18 c . .52 g ; a) (‘0 .2 “5 <1: c: «5 ‘1 a $9 .-.- a o 2 9: 1 91 l. O. a. 16 .o o llllllTlllTl § llllllllllllllll § 1111111111§ 383393 83883838 88838 P 1- v- F P v- 1- m M." 0]. M98“ WW % 186 fine as .3532 689 6:238 ”one as .35 ion apes 3983898928629oesoszegoSanaapasaseaoasasnspoaofigeoségsagpaeaéece 3.3588338 one 2 .683 coca 388a 69... 858. gunmen... 6e 3.8. on. 3688 963 3830 on. one £8 .23 889 m3... ace seasons .888 co scene so: .93 use. b... saute... 9... 38... Es: 98. 59.2.82... 8.5.8... .33.. .58 co 85.... 92.—mean... e .3... oboe- - coop l coo- 1 con m8 Hoe Hoe Hoe c” oHoor moNNA Hosea/o We: o3 more. . . . o.oopr moo. 80 W. HON . low W me ”on floor Hoar Her Hour 1 e a coin Museum w H8“ HOQN loo“ cheer...mooe.l . . ooe soon m 1 m H8 w m H3 .1111: ”8' Hour 11.1.0: ”8? 23: mas—ma 6 ”mo—95m Beach 92 E 38205 332 Enema a flue—9:8 coach 33: £29m b oeoE serenade gunmen $5-3: mecca cornice 2668 ¥E+$= so. mEgonamona me $20 .6: u .9: .m o5 < o .oE 5 team an asa