THE ROLE OF EXOSOMES IN COMPLEMENT ACTIVATION AND DEVELOPMENT OF DIABETIC RETINOPATHY By Chao Huang 2018 A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Cellular and Molecular Biology – Doctor of Philosophy THE ROLE OF EXOSOMES IN COMPLEMENT ACTIVATION AND DEVELOPMENT OF DIABETIC ABSTRACT RETINOPATHY By Chao Huang Diabetic Retinopathy (DR) is a vision-threatening microvascular complication of diabetes, and more than 50% of diabetic patients develop some degree of retinopathy ten years after the onset of the disease (1). With prevalence of diabetes, DR is a major cause of vision impairment among working age adults and is affecting approximately 4.2 million people in the US and 93 million worldwide, despite remarkable advancement in the diagnostic tools and treatments (2). While progress has been made, detailed molecular mechanisms underlying pathogenesis of DR have not yet been fully deciphered. A number of studies imply that there is a link between progression of DR and complement dysregulation (3). The aim for this dissertation is to provide a novel molecular link connecting complement activation with retinal vascular impairment associated with diabetes. Our studies revealed that extracellular vesicles such as exosomes associate with immunoglobulins in plasma and activate complement pathway, which contributes to the increase in retinal vascular permeability in diabetes. First, we investigated the hypothesis that plasma exosomes activate complement pathway and contribute to retinal vascular damage. We demonstrated that IgG-laden exosomes bind and activate classical complement protein C1 in plasma. Exosome quantification showed an increased number of IgG-laden plasma exosomes in diabetes that may results in greater complement activation. An immunoglobulin deficient mouse model (IgM-KO) showed reduced level of exosome-induced complement activation compared to C57bl/6J mice. Importantly, diabetic IgM-KO mouse model demonstrated a greatly reduced level of retinal vascular permeability as compared to diabetic C57bl/6J. Second, we investigate the hypothesis that exosome-induced complement activation contributes to the increase in membrane attack complex (MAC) deposition and cytolytic damage of retinal endothelium in DR. We demonstrated that diabetic, but not control rat plasma induces MAC deposition and cytolytic damage in human retinal endothelial cells (HRECs). Removal of the exosomes from diabetic plasma abrogated MAC deposition and cytolytic damage in HRECs. Taken together, these findings suggest a novel mechanism in which plasma exosomes activate complement cascade and contribute to the pathogenesis of DR. This may provide a novel therapeutic strategy for treating DR. To my loved ones iv ACKNOWLEDGEMENTS I would like to express my deepest appreciation to those who helped and supported me during this exciting doctorate journey. I am especially grateful to my mentor, Julia V. Buisk, for granting me great autonomy to pursue this project that was novel to most people. This line of work would have been impossible if without her unfailing patience, enduring encouragement, and guidance. I will be forever thankful for all of the valuable lessons that Dr. Busik has taught me in both scientific research and life in general. Outside the laboratory, Julia makes the best roasting duck ever during holidays! It has been a great privilege and a great honor to be part of Dr. Busik’s research team. I would like to thank my committee members Dr. Laura McCabe, Dr. Nara Parameswaran, and Dr. Brian Gulbransen for their great support, invaluable suggestions, and constructive criticisms throughout the committee meetings. I would also like to thank everyone in the Busik Laboratory. I was lucky enough to be surrounded by such a group of supportive and talented people. I would like to thank Svetlana Navitskaya for selflessly helping me out in so many ways and for giving me sound advice throughout the years. I am also especially grateful to Sandra Hammer, who has been a great friend and has been always willing and able to offer help in my experiments, and who also often goes through my writing drafts. Another Busik Lab member whom I would like to thank is Yan Levitsky – also a great friend of mine; he has a strong passion in science and is fun to talk with on almost every subject. I also would like to acknowledge my undergraduate assistant, Kiera Fisher, who is also a good friend, for contributing tremendously in my studies. It was nice to have someone to go through v sufferings together. Best of luck in med school, Kiera! I am also deeply grateful to Dr. Sandra O’Reilly, who provided endless support in in vivo studies and took time out of her busy schedule to teach me in vivo techniques. I am also grateful to some previous members of the Busik lab, Dr. Nermin Kady, Dr. Qi Wang, and Dr. Harshini Chakravarthy for their help and support. More importantly, I would like to thank my parents, whose love and support always sustain me to keep going for whatever I pursue. I would also like to thank my beloved girlfriend Chen Lou, for her steady support, understanding, and love. Without any of your support, I would not have made it this far. Thank you all! vi TABLE OF CONTENTS LIST OF FIGURES…………………………………………………………………………………………………….…………….ix KEY TO ABBREVIATIONS……………………………………………………………………………………………………….xi Chapter1. INTRODUCTION………………………………………………………………………………………………….…1 1.1 Background and Significance……………………………………………………………………………………………1 1.2 Etiology of Diabetic Retinopathy………………………………………….……………………….….……………2 1.3 Diabetic Retinopathy Treatments…………………………………………………………….………………………4 1.4 Inflammation in Diabetic Retinopathy …………………………………………………………….………………6 1.4.1 Pro-Inflammatory Cytokines………………….…………………………………………………………………6 1.4.2 Vascular Endothelial Growth Factor ……………………………………………………….………………7 1.4.3 Adhesion Molecules ……………………………………………………………….………………………………7 1.4.4 Protein Glycation in DR ………………….…………….…………….…………….…………….………………8 1.5 Complement System ..…….…………….…………….……………….…………….…………….…………….………8 1.5.1 Classical Pathway……….…………….……………….…………….…………….….…………………….……11 1.5.2 Mannose-Binding Lectin Pathway ………………….…………….……………………….………………12 1.5.3 Alternative Pathway ……………….…………….……………………….…………….………………………12 1.5.4 Membrane Attack Complex (MAC) Formation………………….…………….………………………13 1.6 Complement Inhibitors……………….…………….……………………….…………….……………………………13 1.6.1 Fluid-Phase Regulators……………….…………….……………………….…………….……………………13 1.6.2 Membrane Bound Complement Regulators………………….…………….………………………….15 1.7 Complement in the Eye………………….…………….……………………….…………….…………………………16 1.7.1 Complement in Diabetic Retinopathy……………….…………….………………………….…..…….16 1.8 Exosomes…………………….…………….……………………….…………….……………………….…………….……17 1.8.1 Exosome Nomenclature…………….…………….……………………….…………….…………….……….18 1.8.2 Exosome Biogenesis………………….…………….……………………….…………….………………….…19 1.8.2.1 ESCRT-Dependent Exosome Biogenesis ……………….…………….…………………………………19 1.8.2.2 ESCRT-Independent Exosome Biogenesis………………….…………….………………………….…20 1.8.3 Exosome Docking on Recipient Cells ……………….…………….………………………………………21 1.8.4 Exosome Cargoes ……………….…………….……………………….…………….……………………………22 1.8.5 Exosome Isolation ……………….…………….……………………….…………….…………………….……23 1.8.5.1 Ultracentrifugation Based Isolation……………….…………….………………………………………23 1.8.5.2 Size-Based Exosome Isolation…………….…………….……………………….…………………………25 1.8.5.3 Immuno-Affinity Capture-Based Exosome Isolation………………….…………….………….…26 1.8.5.4 Precipitation-Based Isolation……………….…………….………………………………………….……26 1.8.6 Exosome Quantification/Characterization ……………….…………….……………………….……27 1.8.6.1 Flow Cytometry ……………….…………….……………………….…………….……………………………27 1.8.6.2 Dynamic and Static Light Scattering……………….…………….……………………………….……28 1.8.6.3 NanoSight Tracking Analysis (NTA) ………………….…………….……………………………………29 1.8.6.4 Western Blot Analysis………………….…………….……………………….…………….…………………29 vii 1.8.6.5 Transmission Electron Microscopy (TEM) ……………….…………….……………………….……30 1.8.7 Exosomes and Diabetes.…………….…………….……………………….…………….……………………30 1.8.8 Exosomes and Diabetic Retinopathy.…………….…………….………………………………….……30 1.9 Objective of the Dissertation ……………..…………….………………….…….…………….…………….……31 1.9.1 Overview of Chapters……………….…………….……………………….…………….……………………...33 Chapter 2. Plasma exosomes contribute to microvascular damage in diabetic retinopathy (DR) by activating classical complement pathway………………….…………….………………………………………34 2.1 ABSTRACT……………….…………….……………………….…………….……………………….…………….…….…34 2.2 INTRODUCTION….……….…………….……………………….…………….……………………….…………….……34 2.3 RESULTS……………….…………………….…………….……………………….…………….……….….………………36 2.4 DISCUSSION……………….…………….……………………….…………….……………………….……………………52 2.5 RESEARCH DESIGN AND METHODS…………………………….…………….…………….…….………………57 Chapter 3. Exosome-induced classical complement activation leads to the retinal endothelial cells damage via MAC deposition……………….…………….……………………….…………….…………….……62 3.1 ABSTRACT……………….…………….……………………….…………….……………………….…………….…….…62 3.2 INTRODUCTION …………….…………….……………………….…………….…………………………………………62 3.3 RESULTS……………….…………….……………………….…………….……………………….…………….……….…64 3.4 DISCUSSION..………….…………….……………………….…………….……………………….…………….…..……71 3.5 METHODS……………….…………….……………………….…………….……………………….…………….…….…75 Chapter 4. Summary and Future Perspectives….…………….…………….………………………………….…80 REFERENCES……………….…………….……………………….…………….……………………….…………….……….…84 viii LIST OF FIGURES Figure 1. Three complement activation pathways…….………………………………………………………….10 Figure 2. Comparing exosomes in conventional TEM with Cryo-EM………………………………….….18 Figure 3. Sequential ultracentrifugation exosome isolation………………………………………………….25 Figure 4. Rationale of the dissertation.………………………………………………………………………….…….32 Figure 5. Comparison of plasma exosomes, isolated with ExoQuick, and 100-nm extruded microvesicles..…………………………………………………………………………………………………………….……….39 Figure 6. Western blot of plasma exosomes isolated through the use of ExoQuick from control and STZ-induced diabetic (3 months) mice…………………..………………………………………………….….40 Figure 7. Characterization and specificity of Ig binding to exosomes.……………………….………….42 Figure 8. IgG bound to exosomes in plasma and an elevated number of exosomes in diabetes leads to greater IgG level in mice with diabetes than in control mice……………………….………….43 Figure 9: Western blot analysis of mouse plasma exosomes isolated via ExoQuick………….….44 Figure 10. Western blot of complement activation exosomes isolated from mouse and human plasma through ultracentrifugation or ExoQuick and purified with OptiPrep Density Gradient....…………………………………………………………………………………………………………….…………….47 Figure 11. Western Blot analysis of C1q binding assay without presence of exosomes isolated by ultracentrifugation or/and OptiPrep density gradient.…………………………………….…………….48 Figure 12. Western Blot analysis of C1 auto-activation after ultracentrifugation or/and OptiPrep density gradient without presence of exosomes………………………………………………………...……….48 Figure 13. Western blot analysis of plasma exosomes isolated from C57bl/6J and IgM KO mice by ExoQuick and OptiPrep density gradient...…………………………………………….……………………….50 Figure 14. Analysis of plasma exosomes, isolated with ExoQuick from control and STZ-induced diabetic (7 weeks) C57BL/6J and IgM-KO mice…………………………………………………………………….51 Figure 15. Characterization and specificity of immunoglobulin binding to exosomes………....65 ix Figure 16. Western Blot analysis of complement activation in rat plasma exosomes isolated via ultracentrifugation and purified by OptiPrep density gradient. …………….……………………….67 Figure 17. Comparison of rat artery plasma exosomes isolated via ExoQuick……………………….68 Figure 18. Diabetic rat plasma induced cytotoxicity and cell death in Human Retinal Endothelial Cells (HRECs)…………………………………………………………………………………………………………….………….70 Figure 19. MAC in the retina of diabetic rats ……………………………….……………………….…………….71 x KEY TO ABBREVIATIONS AGEs……………………………………………………………………………….………Advance glycation end products Apo-A……………………………………………………………………………………………………………Apolipoprotein-A Apo-E…………………………………………………………………….…………………………….……. Apolipoprotein-E BRB…………………………………………………………………………………………….……………Blood retinal barrier C1………………………………………………………………………………………….……………. Complement protein 1 C1INH………………………………………………………………………………………………………………….…C1 inhibitor C2…………………………………………………………………………………………………….…. Complement protein 2 C3………………………………………………………………………………………………….……. Complement protein 3 C4…………………………………………………………………………………………………...……Complement protein 4 C4bp…………………………………………………………………………………………………...……C4b-binding protein C5………………………………………………………………………………………………….….…Complement protein 5 C6……………………………………………………………………………………………….…….…Complement protein 6 C7…………………………………………………………………………………………….…….……Complement protein 7 C8………………………………………………………………………………………….………….…Complement protein 8 C9……………………………………………………………………………………….…………….…Complement protein 9 CCD………………………………………………………………………………………….…………. Charge-coupled device CCL2…………………………………………………………………………………….……………………Chemokine ligand 2 CCL5……………………………………………………………………………….…………………………Chemokine ligand 5 CDC…………………………………………………………………………Centers of Disease Control and Prevention CMOS………………………………………………………….…….Complementary metal-oxide-semiconductor xi CR1………………………………………………………….………………………………………. Complement Receptor 1 DAF…………………………………………………….……………………………………………Decay-accelerating factor DCCT…………………………………………………………………………Diabetes Control and Complications Trial DLS……………………………………………………………………………………………………. Dynamic light scattering DM…………………………………………………….…………………………………………………………Diabetes mellitus DME……………………………………………………………………………………………………Diabetic macular edema DR……………………………………………………………………….…………………………………. Diabetic retinopathy ECM………………………………………………………………………………….………………………Extracellular matrix ELISA……………………………………………………………………….……. Enzyme-linked immunosorbent assay EM………………………………………………………………………………………………….………. Electron microscope ESCRT…………………………………………………………Endosomal sorting complex required for transport EV……………………………………………………….…………………………………………………. Extracellular vesicles FB…………………………………………………………………………………………………….……. Complement factor B FD……………………………………………………………………………………………………….…Complement factor D FH………………………………………………………………………………………………….………Complement factor H GPI………………………………………………………………………………………….…. Glycosylphosphatidylinositol HDL…………………………………………………………………………………….……………. High-density lipoprotein HRECs……………………………………………………………………………….……. Human retinal endothelial cells ICAM-1……………………………………………………………………….……. Intercellular adhesion molecule-1 IgM-KO…….……………………………………………………………………………………………………………Ighmtm1cgn/J IL-1R1….………………………………………………………………………………………………Interleukin-1b receptor IL1-b…………………………………………………………………………………………………………….……Interleukin-1b xii LDH Assay……………………………………………………………………………….…. Lactate dehydrogenase assay LDL……………………………………………………………………………………………………. Low-density lipoprotein MAC/C5b-9……………………………………………………………………………………Membrane attack complex MASP1/2-……………………………………………………...Mannose-binding lectin-associated proteases MBL………………………………………………………….………………………………………. Mannose-binding lectin MCP………………………………………………………………………………………………Membrane cofactor protein MVBs…………………………………………………………………………………….………….Multivesicular bodies MVEs………………………………………………………………………………….…………. Multivesicular endosomes NPDR………………………………………………………………. Non-proliferative stage Diabetic Retinopathy NTA…………………………………………………………….………………………………. NanoSight tracking analysis PDR……………………………………………………….……………………………. Proliferative diabetic retinopathy PEGs………………………………………………….……………………………………………………. Polyethylene glycols PRP…………………………………………………………………………………….…….Pan-retinal photocoagulation RAGEs……………………………………………………………….Receptors for Advance glycation end products RPE…………………………………………………………………………………………Retinal pigment epithelium cells SLS………………………………………………….………………………………………………………Static light scattering STZ……………………………………………….…………………………………………………………………..Streptozotocin TEM……………………………………………………………..……………………..Transmission electron microscopy TNF-α……………………………………………………….…………………………………….Tumor necrosis factor-α UKPDS…………………………………………………………………United Kingdom Prospective Diabetes Study VCAM-1……………………………………………………….…………………..Vascular cell adhesion molecule-1 VEGF…………………………………………………………………….……………Vascular endothelial growth factor xiii WHO…………………………………………………………………………….……………...World Health Organization xiv Chapter 1. INTRODUCTION 1.1 Background and Significance Diabetes mellitus (DM) is a chronic metabolic disease characterized by elevated blood glucose levels. This is in part due to the impairment of insulin secreting pancreatic beta cells, or the body’s failure to respond to the secreted insulin. Worldwide, the World Health Organization (WHO) estimates that 422 million people in 2014 were diagnosed with diabetes and that number is projected to increase to 592 million in 2030 (4). In the US alone in 2015, an estimated 23.1 million people were diagnosed with diabetes as reported by the Centers of Disease Control and Prevention (CDC), constituting 7.2% of the United States population. Diabetes and its complications are the seventh leading cause of death in the U.S (5), and with the increasing prevalence of the disease the health care cost imposes a significant burden on society. Diabetes mellitus is classified into two major categories: type 1 or type 2. Type 1 DM is defined as insulin-dependent DM, in which insulin producing pancreatic beta cells in the islet of Langerhans are destroyed by an autoimmune process. Generally, type 1 DM occurs more frequently in genetically predisposed children and juvenile, but it can appear at any age(5). Type 1 DM accounts for approximately 10% of diabetic patients, while type 2 DM accounts for 90-95% of all cases of diabetes and mostly presents in adults (5).Type 2 DM can be defined as insulin independent or insulin resistant DM, and is due to peripheral tissue having an ineffective response to insulin for glucose uptake. Different from type 1 DM, certain risk factors increase the development of type 2 DM such as obesity, sedentary lifestyle, gender and ethnicity (6). In the recent years, type 2 DM becomes more prevalent in children and younger individuals, which is largely attributed to obesity epidemic. Over time, patients with either type of diabetes have 1 increased risk of developing complications such as retinopathy, nephropathy, cardiovascular diseases, stroke, and neuropathy. Diabetic retinopathy (DR) is a long-term micro-vascular complication of diabetes. Each year 12% of diabetic patients in the U.S. become blind due to DR (NEI website). Moreover, without proper control of blood glucose level, 20-25% of people with diabetes have vision impairment after 10 years onset of disease (7). DR affects men and women about equally and often goes unnoticed until vision impairment occurs (NEI website); it is believed that nearly half of Americans with diabetes will have developed some form of retinopathy at the time of diagnosis(1). Therefore, DR is the number one cause of vision impairment in working-age adults in the U.S. (8). The duration of diabetes is suggested to be the strongest predictor for development and progression of DR. Type 1 and type 2 diabetic patients are equally affected by the DR with more than 60% prevalence after 20 years(1). 1.2 Etiology of Diabetic Retinopathy The retina is a multiple layer tissue where visible light is converted into electrochemical impulses that are transmitted via the optic nerve into the cerebral visual cortex for image interpretation. There are two circulations that supply the retina: the outer retina is nourished by choriocapillaries, while the inner retina is supported by the inner retinal capillaries (9). The choroid capillaries have high blood flow volume and low arteriovenous pO2 difference; this suggests that the choroid capillaries have low metabolic activity (10). On the other hand, the inner retinal capillaries that derive from the central retinal artery have small caliber and are projected throughout multiple retinal layers, providing nutritive needs for the complex neuronal network. Despite of limited blood supply, a large arteriovenous pO2 difference indicates that the 2 inner retinal capillaries have a high metabolic activity (10). These factors suggest that inner retinal capillaries have a low capacity to tolerate the metabolic stress induced by diabetes (11). Diabetes affects all the layers of the retina and leads to microvascular perturbations. Pathology of DR presents itself in distinctive stages which are used for clinical diagnosis. The early stage of the disease, known as non-proliferative stage DR (NPDR), is characterized by remodeling of the retinal microvasculature. In this stage, thickening of the capillary basement membrane results from loss of retinal pericytes, leading to weakening of the retinal blood vessels and formation of acellular capillaries (12,13). Moreover, tiny bulges protruding from the retinal capillary, called microaneurysms, can rupture and result in an increase in retinal vascular permeability. Concomitantly, lipid and/or blood may leak into the retina resulting in intra-retinal hemorrhages and formation of hard exudates. Leakage of fluid/or blood into the macula can cause diabetic macular edema (DME) resulting in blurry vision and loss of visual acuity (14). The alterations in the vascular structure can be observed upon ophthalmoscopic examination including: non-perfused vessels, microaneurysms, dot hemorrhages, cotton-wool spots, beading venous, vascular loop and subtle increases in vascular permeability (15). As NPDR progresses approximately 50% of patients will develop the advanced stage disease named proliferative diabetic retinopathy(PDR) in 1 year (7); this stage is characterized by growth of new blood vessels from the retinal capillaries into the vitreous, known as neovascularization. These newly grown blood vessels tend to be fragile, leaky and exceptionally permeable, and often lead to formation of hemorrhage accompanied by scar tissue formation. Once scar tissue is formed in the vitreous of the eye, this can cause detachment of the retina, resulting in permanent vision loss(16). 3 1.3 Diabetic Retinopathy Treatments Several clinical trials have demonstrated the beneficial effect of tight glycemic control (hemoglobin A1C < 7%) in slowing the progression of DR in diabetic patients (17–19). The Diabetes Control and Complications Trial (DCCT) study demonstrated that intensive blood glucose control reduces the risk of DR progression by 76% in type 1 diabetic patients (17). The United Kingdom Prospective Diabetes Study (UKPDS) also has shown a 25% reduction in rate for DR development in type 2 diabetic patients who underwent tight glycemic control (18). However, intensive tight glycemic control is hard to achieve in most diabetic patients and even diabetic patients with good glycemic control can develop complications (20). Moreover, patients who undergo intensive glycemic control may have life threatening complications such as hypoglycemic episodes and diabetic ketoacidosis (17). Over the last couple of decades, surgical interventions have been developed to prevent the deterioration of vision lost due to DR. Pan-retinal photocoagulation (PRP) has been the main option for sight-threatened DR patients since it was first developed in the 1970s. It utilizes high intensity laser to seal off leaking peripheral retinal blood vessels, as well as retinal tissue to prevent worsening of visual acuity by limiting the metabolic load on the remaining central retina. Although this technique has demonstrated a 50% reduction in vision loss in DR patients with PDR, PRP does not result in vision improvement and may in fact result in complications like loss of peripheral vision, development or worsening of macular edema, and loss of color vision (21). Vascular endothelial growth factor (VEGF), a pro-angiogenic factor, has been shown to be significantly elevated in the vitreous of DR patients. A number of studies have implicated this growth factor as a major causative factor in diabetic macular edema, retinal neovascularization, 4 and other complications that alter the Blood-Retinal Barrier (BRB)(22). Therefore, VEGF is one of the most heavily studied factors in DR research. Recently, a group of new drugs was developed that, by targeting the VEGF activity, have shown great promise in stabilization of the BRB and reduction in retinal vascular leakage. This group of drugs includes VEGF aptamer, pegaptanib (Macugen), the monoclonal antibody fragment ranibizumab (Lucentis), the full-length antibody bevacizumab (Avastin), VEGF-Trap (Regeneron), bevasiranib (Eylia). Anti-VEGF therapies have shown great promise in halting neovascularization in advance stages of DR and diabetic macular edema; however, there are limitations. Anti-VEGF drugs have relatively short effective-duration, and causing tractional retinal detachment in some patients with severe advanced DR stage(23). Other treatment options such as vitrectomy and intravitreal injection of Corticosteroids have also been shown to be beneficial in slowing DR progression and stabilizing vision. However, despite a range of therapeutic approaches (laser photocoagulation, intravitreal anti-VEGF injection, corticosteroid injection) being available, all of these treatments merely focus on slowing the progression of vision loss instead of vision restoration. Additionally, current gold standards of care are aimed at the later more advanced stages of DR where vision complications already exist. Moreover, these approaches involve procedures or the methods of delivery that are not only burdensome to patients, but also can pose a risk of complications such as infection, inflammation, and endophthalmitis (24). Therefore, development of new therapies remains an important goal in the field. Moreover, there is a great need to develop therapies that target the early stage of the disease to prevent progression into more advanced stages. 5 1.4 Inflammation in Diabetic Retinopathy Inflammation is the body’s defense response to noxious stimuli such as pathogens. This complex mechanism involves multiple mediators such as pro-inflammatory cytokines (Tumor necrosis factor alpha (TNF-α) and interleukins) and chemokines (CCL2 and CCL5) which activate and guide immune cells (leukocytes) to the site of injury. Upregulation of adhesion molecules (ICAM-1 and VCAM-1) in the endothelium then allows leukocytes to attach and transmigrate from the endothelium into the infected or injured tissue. Generally, inflammation is beneficial and occurs on an acute basis but can have undesirable effects if it becomes persistent and uncontrollable. Increasing evidence points to inflammation as a key player in pathogenesis of DR. There are interlinked molecular pathways in DR that contribute to the inflammatory response. However, the exact underlying molecular mechanisms causing pathological changes in the early stage and progression to the advance stage of DR are not fully understood. 1.4.1 Pro-Inflammatory Cytokines Although there are no pathogens in DR, a significant increase of pro-inflammatory cytokines, chemokines and adhesion molecules have been reported in vitreous, serum and retina of diabetic patients and experimental animal models (22,25). TNF-α is an inflammatory cytokine that is elevated in vitreous, serum and retina from patients with DR, as well as in retinas of diabetic animal models(15). Moreover, inhibition of TNF-α via a soluble receptor of TNF-α has been shown to reduce leukocyte adherence in the retinal blood vessels as well as blood-retinal barrier breakdown (22). These studies suggest the role of TNF-α in diabetes-induced increase in leukostasis and vascular permeability. 6 Interleukin-1b (IL1-b), a pro-inflammatory cytokine, is cleaved by caspase-1 into its biologically active form and binds to its receptor (IL-1R1) expressed on the cell surface. The level of IL1-b and the activity of the caspase-1 are increased in retinas of diabetic patients, rodent models and high glucose treated Muller cells (26). Mice lacking IL1-b receptor were shown to be protected from diabetes induced retinal capillary degeneration (27), indicating that IL1-b participates in pathological development of DR. 1.4.2 Vascular Endothelial Growth Factor Previously mentioned VEGF is a pro-inflammatory molecule whose levels are closely tied to the progression and severity of DR. VEGF is found in high level in diabetic retina and is mainly produced by Mueller cells. Elevated level of VEGF has been shown to directly contribute to increased retinal leukostasis (22,28) and to alter BRB integrity (29). Intraocular delivery of anti- VEGF therapies is widely used to treat advanced stage of DR as discussed previously. 1.4.3 Adhesion Molecules Increased level of neutrophils in both choroidal and retinal blood vessels have been reported in diabetic patients, in conjunction with upregulation of intercellular adhesion molecules-1 (ICAM- 1)and P-selectin(30). Moreover, in DR patients, loss of endothelial cells and capillary dropout has been correlated with an increased number of leukocytes in the choriocapillaries (31). In animal studies, retinal leukostasis was found in early stage of diabetic animals and was associated with increased retinal vascular permeability and upregulation of adhesion molecules (ICAM-1 and CD18)(32). In addition, Miyamoto et al. have shown that recruitment of leukocytes to the retinal capillaries leads to retinal non-perfusion and increased vascular permeability(33). 7 1.4.4 Protein Glycation in DR In diabetes, hyperglycemia drives protein glycation and the glycated proteins may have modified activity. During this non-enzymatic reaction, a free amino group of a protein interacts with a carbonyl group from glucose to form a Schiff base and this process occurs over a period of hours. Once it is formed, Schiff base rearranges into a more stable form named Amadori product, and formation of Amadori products occurs over days. Glycated protein can undergo further rearrangement and formation of advanced glycation end products (AGEs)(34). Given their slow formation process, glycation occurs at the long-lived extracellular proteins, as well as on short- lived molecules. Receptors for AGEs (RAGEs) are found in many cells and interacts with AGEs. In DR, AGEs have been attached on the lumen of the retinal blood vessel walls and it is suggested to contribute to vascular perturbation and increase retinal vascular permeability (35). Moreover, inhibition of RAGE has been shown to have anti-inflammatory effects in retina(36). 1.5 Complement System The complement system is part of innate immunity, which is a major defense system against infection. It also plays a key role in maintaining host homeostasis, for example, by removing immuno-complex or cellular debris(37,38). Complement system includes more than 30 proteins found in plasma and on cell surfaces that are synthesized by liver or locally in peripheral tissues(39,40). Complement proteins are most abundant in plasma as inactive precursors, and there are more than 3 g of complement proteins per liter of plasma(37). On the cell surface, complement proteins can function as receptors for activated complement proteins or regulatory complement proteins. During inflammation, activation of complement system opsonizes the target and attracts leucocytes to the tissue site. There are three pathways involved in 8 complement activation, namely: classical pathway, lectin pathway, and alternative pathway (Figure 1). Once complement is activated, all three pathways converge and lead to the activation of a central mediator complement protein 3 (C3). With further activation of C3 and complement protein 5(C5), the activation enters into the termination pathway resulting in the formation of a one megadalton membrane attack complex (MAC/C5b-9). MAC is the main effector of complement-mediated tissue damage, which includes cellular lysis and cell death(38). Complement activation is tightly regulated and delicately balanced by regulatory proteins, and in physiological conditions will only be directed against foreign cells, pathogens or cellular debris, but not self-tissue. 9 Figure 1. Three complement activation pathways: Classical, Mannose-Binding Lectin, and Alternative Pathways. The three pathways converge at the point of cleavage of C3. The classical pathway is activated via binding of the C1 complex (consists of C1q, C1r and C1s) to antibody complex. Activated C1s cleaves C4 into C4a and C4b, the later then binds with membrane surface and cleaves C2, leading to the formation of a protein complex C4b2a, also known as C3 convertase of classical pathway. Activation of the mannose-binding lectin pathway is initiated by binding of mannose-binding lectin(MBL) complex to sugar groups on the surface of a bacterial cells, then activated serine proteases mannose-binding lectin-associated proteases (MASPs) of the complex cleaves C4 into C4a and C4b. C3 convertase formed by mannose-binding lectin pathway consists of similar complement components as classical pathway. The alternative pathway is constitutively active at low level leading to spontaneous hydrolysis of C3. Hydrolyzed C3 (C3(H2O)) binds with Factor B (FB), gets cleaved by serine protease Factor D (FD) leading to formation of fluid phase C3 convertase (C3(H2O) Bb). C3b fragment generated by C3 convertase, in turn binds with FB leading to formation of amplification of C3 convertase (C3bBb). C3 is cleaved by C3 convertase releasing anaphylatoxin C3a and leading to opsonization of fragment C3b. Active C3b binds with C3 convertase, forming C5 Convertase (C4b2a-c3b or C3bBb-c3b). C5 convertase cleaves C5 into anaphylatoxin C5a and C5b. C5b initiates formation of membrane attack complex (MAC) via recruitment of complement proteins (C6, C7, C8, and C9) onto the target membrane. 10 1.5.1 Classical Pathway The classical pathway was the first discovered complement pathway; it is commonly initiated by binding of Complement protein 1(C1) to antigen bound IgM, IgG1 or IgG3. C1 is a large complement protein complex consisting of C1q, C1r, and C1s subunits. C1q is the recognition molecule of the classical pathway, which is produced mainly by dendritic cells and monocytes(41). In C1 complex, C1q consists of six globular target recognition domains, allowing C1q to recognize different target molecules including IgG- and IgM-containing immune complexes. IgM is a hexamer, and when it binds to an antigen it leads to exposure of the C1q- binding site (41). Conversely, IgG is a monomer, and it has a low binding affinity for C1q. However, when IgG hexamers form on the antigen surface, the Fc fragment presents an efficient binding site for the C1q(42). Additionally, C1q is also known to recognize molecules present on the surface of dying cells such as pathogen-associated molecular patterns and bacterial porins (43), directly causing complement activation. Once C1q binds to its target surface, this induces auto activation of serine protease C1r, which cleaves and activates another serine protease C1s (44). Then, the activated C1s recognizes and cleaves zymogen C4 resulting in production of C4a and C4b. C4b covalently binds to the bacterial surface in close proximity to the C1 complex-binding site(45,46). C2 complex is subsequently cleaved by C1s, forming C2b; association of C4b with C2b on the membrane surface results in formation of Classical pathway C3 convertase: C4b2b complex. This complex can cleave C3 into membrane binding complement protein C3b and anaphylatoxin C3a. 11 1.5.2 Mannose-Binding Lectin Pathway The lectin pathway is similar to the classical pathway, except lectin pathway relies on members of the collectin family (Mannose-binding Lectin(MBL) and ficolins) to identify sugar patterns that are frequently expressed on microorganisms but rarely present on the surface of host cells(37). MBL consists of six trimeric subunits that have similar structure to C1q in C1 complex(47). Binding of MBL to its target surface leads to activation of mannose-binding lectin-associated proteases (MASP1 and MASP2), and then the complex cleaves C4 and C2. From here C3 convertase (C4b2b) formation is analogous to the classical pathway, which cleaves C3 into C3a and C3b. 1.5.3 Alternative Pathway Different from the classical and lectin pathway which are activated upon recognition of exogenous material, the alternative pathway is constitutively active at low levels to monitor for pathogen invasion in a process known as tick-over (41). Activation of the alternative pathway is initiated by spontaneous hydrolysis of a thioester bond within C3 (47). Hydrolyzed C3 (C3-(H2O)) binds to complement factor B (FB), which is then cleaved by a serine protease, complement factor D (FD), which then forms a fluid phase initiation C3 convertase complex C3(H2O)Bb (41). This fluid phase C3 convertase, C3(H2O) Bb, cleaves C3 into C3a and C3b. Formation of C3b fragments, from initiation C3 convertase, bind to FB and are cleaved by FD, which activates the amplification loop of C3 convertase formation (C3bBb) in fluid phase or on target surface(41). It should be noted that C3b fragment generated from any of the pathways can cause the amplification of C3 convertase in order to augment the conversion of C3. 12 1.5.4 Membrane Attack Complex (MAC) Formation All three pathways converge at the formation of C3 convertase, which cleaves C3 into C3a, an anaphylatoxin, and C3b, which serves as an opsonin. The incorporation of C3b into C3 convertase results in the formation of C5 convertase (Classical and Leptin pathway: C4bC2bC3b, Alternative pathway: C3bBbC3b). Complement protein 5 (C5) is cleaved by C5 convertase into C5a and C5b. Upon cleavage by C5 convertase, C5b binds with C6 and causes a conformational change in C6 which allows the interaction with C7(48). The C5b-7 complex has high affinity to lipids, which allows the complex to be inserted into the membrane and bind with the C8. Assembly of the C5b- 8 complex initiates the C9 incorporation and polymerization, resulting in formation of the C5b-9 membrane attack complex. The C5b-9 complex forms an annular ring structure with an external diameter of 20nm on the target membrane, causing cell lysis and death(48). 1.6 Complement Inhibitors Complement system is important for defending host tissue and maintaining cellular integrity; however, it has a potential to cause tissue destruction if not tightly regulated. Complement regulatory proteins keep the complement system under control by regulating every step of the cascade, and disturbance of this delicate balance can result in damage to self-tissue or even autoimmune disease(49). Complement regulator proteins are present at all levels, and they are categorized into two main classes: fluid-phase regulators and membrane regulators. 1.6.1 Fluid-Phase Regulators Activation of the complement system occurs rapidly and amplifies during the fluid phase; therefore, a relatively short half-life and continuous decay-dissociation of fluid-phase regulators is needed to restrict the activation(50). Fluid-phase complement regulators are present in plasma 13 or in biological fluids such as synovial or vitreous; they include: C1 inhibitor (C1 INH), C4b-binding protein (C4bp), complement factor H (FH), clusterin and S-protein(vitronectin). C1 INH is a regulator protein involved in classical pathway. It is a plasma glycoprotein, synthesized primarily in the liver, as well as by monocytes or skin fibroblasts at lower levels (49). C1 INH is a serine protease inhibitor. It reversibly binds to serine protease C1s and C1r in circulation to prevent the auto-activation of the C1 complex. Once C1 complex is activated, C1r and C1s cleave C1 INH to initiate their protease activity (50). C4b-binding protein (C4bp) is mainly synthesized in the liver and is one of the largest fluid-phase regulators present in the plasma. In complement system, C4bp functions to irreversibly displace C2b from the C4b2b (C3 convertase) and inactivate C4b, together with complement factor I, to prevent further activation of C3 convertase formed by classical or lectin pathway(49). FH is a fluid-phase regulator which inhibits C3 convertase (C3bBb) formation from the Alternative pathway. Factor H inhibits C3 convertase formation via three mechanisms: first, it competes with FB for binding with C3b. FH also functions as a decay- accelerating factor to displace Bb from the C3 convertase. Thirdly, FH acts as a cofactor for complement factor I to cleave C3b into inactive form iC3b(49). Clusterin is mainly synthesized in the liver, and it functions to aggregate cells. S-protein is also synthesized mainly in liver, as well as being produced by platelets and macrophages and is important in cell-matrix interaction. In complement system, clusterin and vitronectin have similar functions; both bind to terminal pathway complex: C5b-7, C5b-8 and C5b-9, preventing their insertion into the cell membranes and subsequent cell death(50). 14 1.6.2 Membrane Bound Complement Regulators Mainly present on the host cell membrane, membrane complement regulators function to protect the cell from autologous complement attack. There are four well characterized membrane complement regulators: complement receptor type 1 (CR1), membrane cofactor protein (MCP), decay-accelerating factor (DAF, CD55), and protectin (CD59). CR1 is a long membrane bound glycoprotein, it inactivates C3/C5 convertases and functions as a cofactor for factor I-mediated cleavage of C4b and/or C3b into less active fragments (50). CR1 is present on erythrocytes, T and B lymphocytes, neutrophils, monocytes and eosinophils, but is notably absent in platelets (50). CR1 also binds with immune complex and is responsible for immune complex processing and clearance (51). MCP is an integral membrane glycoprotein. It is present on all circulating cells, including epithelial and endothelial cells, but not on erythrocytes. MCP binds and cleaves C3b and/or C4b, which prevents C3 convertase formation(51). Decay-accelerating factor (DAF, CD55) is a glycoprotein that is present on all peripheral blood cells, vascular endothelial cells, and many types of epithelial cells. DAF anchors on the phospholipids of the cell membrane via a glycosylphosphatidylinositol (GPI), but is also found in soluble form in biological fluids. DAF with GPI tail can be detached from the membrane, and exerts its function by reincorporated into a new cell membrane(51). DAF inhibits C3 convertase formed through classical and alternative pathways, and accelerates the decay of C3 convertase(50). DAF expression on the cellular membrane is relatively low but can be upregulated via stimuli. CD59 is expressed ubiquitously on cells, including circulating blood cells, endothelial cells, and epithelial cells(50). Like DAF, CD59 also anchors on the cell membrane via GPI and can be released from 15 the membrane and become soluble. CD59 inhibits MAC assembly by binding to the C5b-8 complex, which limits the C9 incorporation and polymerization(50). 1.7 Complement in the Eye In the eye, three ocular compartments reside behind blood-tissue barriers and display immune privilege: the anterior chamber, vitreous cavity, and sub-retinal space(52). As a consequence, T cells, NK cells, macrophages and granulocytes are prevented from entering these compartments, leaving the eye highly vulnerable to various pathogens. Therefore, complement system serves as the first line of defense to protect the eye from immunological insults while simultaneously maintaining the immune-privileged state of the eye. Alternative pathway induces spontaneous complement activation and continuously produces low levels of C3 activation products and MAC deposition that provides immediate protection against various pathogens (53). Spontaneous complement activation is regulated by membrane bound complement regulatory proteins (DAF, MCP and CD59) on different layers of the retina(54). The complement system functions as a double-edged sword. Normally complement activation helps to maintain healthy environment within the eye, but it is finely balanced and suppressed to prevent overactivation and ocular damage. On the other hand, when perturbation occurs, such as diabetes, deleterious effects can lead to overactivation and tissue damage. 1.7.1 Complement in Diabetic Retinopathy Several studies have shown the involvement of complement system in DR. Human vitreous proteomic studies demonstrated an increase in complement proteins (C3, C4b, C9 and FB) in patients with proliferative stage DR(55). Gerl et al. showed that complement activation occurred 16 to completion in the choriocapillaries of the eyes of DR donors by deposition of C3d and MAC (56). In a separate study, Zhang et al. found that MAC deposition on the endothelium of the donors’ eyes with DR; moreover, reduced levels of membrane-bound complement regulatory proteins (CD55 and CD59) were suggested to contribute to MAC deposition(57). Furthermore, glycation induced inactivation of GPI anchored complement protein CD59 contributes to increased MAC deposition in diabetes(58). There are genetic association studies linking complement gene, with DR. Xu et al. demonstrated an intronic SNP rs2269067 in the C5 gene is a risk factor for development of proliferative DR of type 2 diabetes in Chinese Han population (59). Whether observed complement modulation is a causative factor for DR pathogenesis, or rather is a consequence of DR-associated tissue damage still remains unclear(3). 1.8 Exosomes The term “Exosome” was first described in 1981 by Trams et al. as a group of membrane vesicles, of various sizes, exfoliated from cells that contain 5’-nucleotidase activity (60). In 1983, Pan et al discovered vesicles externalized from sheep reticulocytes during maturation that contain transferrin receptor (61). Later they reported that these vesicles are around 50nm in diameter and associate with plasma membrane activity, and named them “Exosomes”(62,63). At that time, the exosome secretion mechanism was suggested to discard unnecessary membrane proteins during reticulocyte maturation. This mechanism was accepted until 1996, when Raposo et al demonstrated that MHC class II containing exosomes from B lymphoblastoid cells play a role in antigen presentation(64). Interest in extracellular vesicles then really sparked after studies showed that exosomes bear proteins, lipids and RNAs, and that they were involved in intercellular communications. Since then, studies surrounding exosome biogenesis, secretion, 17 and composition have increased exponentially. Moreover, exosomes are becoming potentially valuable as biomarkers for various diseases. Exosomes have a diameter ranging from 30 to 200 nm, and under an electron microscope (EM) exosomes normally show a characteristic flattened cup-shaped morphology. It has been suggested that this cup-shaped morphology is most likely caused by the dehydration process during EM preparation that leads to collapse of the exosomes (65). In contrast, exosomes which remain fully hydrated in cryo-EM are found to have a round-shaped morphology(65) (Figure 2). Figure 2. Comparing exosomes in conventional TEM with Cryo-EM (G.van Niel et al. 2017. Nature reviews Molecular Cell Biology). Isolated extracellular vesicles processed for conventional transmission electron microscopy (TEM) shows a cup-shaped morphology (top panel). In cryo- electron microscopy (cryo-EM), extracellular vesicles appear in a round structures and enclosed by double-leaflet membranes (bottom panel). 1.8.1 Exosome Nomenclature Cells release diverse types of vesicles into the extracellular environment and the nomenclature of the vesicles is still being developed. However, extracellular vesicles (EV) are often categorized into two classes: ectosomes and exosomes. Ectosomes are often given other names including shedding vesicles, nanoparticles, exosome-like vesicles and micro-vesicles. Generally, ectosomes 18 refer to EVs of various sizes (150 to 1000nm) that are shed from the plasma membrane of all cells (66). In contrast, exosomes are intraluminal vesicles before they are released extracellularly. Exosomes are generated via inward budding of the endosomal membrane during the formation of multivesicular bodies (MVBs) in late endosomes stage, and are released extracellularly via exocytosis(67). Exosomes have a smaller diameter range (40 to 100 nm) than ectosomes, have a density of 1.13-1.19g/ml, and are released by most cells(67). Despite differences in membrane of origin, ectosomes and exosomes function similarly once they are released into the extracellular space. They both participate in binding and fusion to target cells and play critical roles in cell physiology and pathology. The overlapping size range, similar morphology, common intracellular mechanisms and cargo sorting machineries makes it a challenge to differentiate ectosomes and exosomes in many studies (67). 1.8.2 Exosome Biogenesis 1.8.2.1 ESCRT-Dependent Exosome Biogenesis Exosomes are generated on the lumen of endosomes during the endosome maturation stage, there are several cargo sorting machineries involved in this process. The endosomal sorting complex required for transport (ESCRT) machinery plays an important role in the formation of the MVB(68). ESCRT-dependent exosome biogenesis relies on four ESCRT machineries that act in a stepwise manner; each ESCRT complex comprises several subunits and plays distinct roles. ESCRT-0 binds and sequesters ubiquitylated transmembrane cargoes on the micro-domains of the multivesicular endosomes (MVEs); then, ESCRT-I and ESCRT-II are recruited to the site to initiate local budding of the endosomal membrane; and then ESCRT-III complex performs budding and fission of the micro-domain (68,69). 19 1.8.2.2 ESCRT-Independent Exosome Biogenesis Exosomes can also be formed via ESCRT-independent mechanisms. Ceramide that has been hydrolyzed from sphingomyelin by neutral type II sphingomyelinase has been shown to participate in ESCRT-independent exosome biogenesis (70). Ceramide participates in micro- domain generation on the membranes (71). It is possible that ceramide enriched endosomal membrane micro-domains impose spontaneous negative curvature on the membranes, facilitating the budding process in MVBs. Tetraspanin proteins expressed on various types of endocytic membranes and are involved in a multitude of biological processes including: cell adhesion, motility, invasion or membrane fusion(72). Extracellular vesicles are highly enriched with tetraspanin proteins. This group of proteins is involved in organization of membrane micro-domains by clustering and interacting with transmembrane and cytosolic proteins(72). The tetraspanin family has been shown to regulate ESCRT-independent exosome biogenesis. Tetraspanin protein CD63, which is enriched on the exosomal surface, together with apolipoprotein E(ApoE) has been shown to participate in sorting in melanoma cells (73). Other tetraspanin proteins such as CD81, CD82 and CD9 are also directly involved in cargo sorting process during exosome biogenesis(74,75). In sum, exosome biogenesis is a complex process which involves several distinct mechanisms for recruitment of cargoes and leads to generation of heterogeneous populations of exosomes. Most of these sorting machinery proteins (ESCRT and tetraspanin) are inherited on exosomes and can then therefore be used to identify and characterize secreted exosomes. These main exosome markers include ESCRT-dependent pathway proteins such as ALIX and TSG101, as well as ESCRT- independent/Tetraspanin-dependent pathway proteins such as CD63, CD9, CD81, and CD82. It is 20 possible that the sorting machinery functions either independently or in combination, and as a result exosomes may carry multiple markers from different machineries(66). It is also likely that exosome biogenesis can be influenced by other pathological stimuli that acts on the cells as well, potentially creating an even greater diversity of exosome populations. To overcome this challenge there are several EV databases available, namely EVpedia(76), Vesiclepedia(77), and ExoCarta(78), which allow researchers to compare across previously published exosomal markers and cargos. 1.8.3 Exosome Docking on Recipient Cells Once exosomes are released into the extracellular space, they can travel and deliver their contents to the recipient cells, promoting physiological or pathological changes in the cells. This exosome-mediated inter-cellular communication requires a series of steps including: docking on the plasma membrane, exosome internalization, and signal delivering. One of the key characteristics of the exosomes is their target cell specificity. This is likely determined by specific interactions between enriched proteins on the exosome surface and receptors on the plasma membrane of the recipient cells. As an example, integrins on exosomes interact with adherent molecules ICAMs on the membrane of the recipient cell to facilitate the docking process(79). Similar interactions have been demonstrated with mediators including tetraspanins (80), lipids(81), and extracellular matrix (ECM) (82). Once exosomes dock on the recipient cells, they may remain at the plasma membrane (83) or be internalized via endocytosis such as micropinocytosis(84) or phagocytosis (85). In addition, plasma membrane lipid composition of recipient cells also contributes to exosome internalization(86). Once docked on the plasma membrane, exosomal transmembrane proteins bind and activate receptors on recipient cells to 21 induce functional response; moreover, cargos delivered by exosomes can also activate various responses in recipient cells(67). 1.8.4 Exosome Cargoes Exosomes are typically comprised of luminal cargoes including: proteins, lipids, and nucleic acids, surrounded by a lipid bilayer membrane which protects the luminal cargo from the extracellular environment. Characterization of exosomal cargo is one of the core interests in exosomal studies because these cargoes can provide invaluable information associated with the physiological or pathological states of the parental cells. The nature and abundance of the exosomal cargoes are cell-type-specific, making it extremely challenging to identify the origin of exosomes found in biological fluids such as plasma, urine, or breast milk. With the expansion of the field, more exosome luminal cargoes are being discovered continuously. In general, the composition of the lipid bilayer in exosomes differs from the plasma membrane composition of the cell of origin(87,88). Exosomes are particularly enriched with cholesterol, sphingomyelin, phosphatidylserine, ganglioside, and some saturated fatty acids(66,88). Additionally, exosomal lipid composition has been investigated in different origin cell types and found to vary based on the cell type they originated from(89). However, lipid composition of exosomes in biological fluids is less characterized(89) In 2007, Valadi et al. were the first to show that RNA is presented in extracellular vesicles; moreover, they also showed that mRNA inside extracellular vesicles could be translated into protein in vitro(90). Since then, numerous studies have identified the presence of genetic material in exosomes and other extracellular vesicles, including mRNAs, miRNAs, and non-coding RNAs with low levels of ribosomal 18S and 28S RNAs (66). It is suggested that RNAs and cytosolic 22 proteins are packaged into exosomes during the inward budding process of MVB formation(87). Furthermore, RNAs inside the exosomes are resistant to RNAse digestion; importantly, miRNAs delivered by exosomes can affect gene expression in distant recipient cells(66,90). 1.8.5 Exosome Isolation Extracellular vesicles are present in an environment that contains a wide spectrum of cellular debris and other biological components. Therefore, it is crucial that exosomes are specifically isolated and kept free of other biological contaminants. With fast advances in science, there are many exosome isolation techniques being developed with different outcomes in extracted exosome quantity and purity. Each isolation technique utilizes a particular trait of exosomes such as their size, density, or surface proteins to aid in their isolation. Therefore, each isolation method brings a unique set of advantages and disadvantages to exosome isolation. There are new isolation methods emerging continuously, therefore, this dissertation will only focus on the most commonly used techniques. These include, ultracentrifugation-based isolation, size-based isolation, immunoaffinity capture-based isolation, and precipitation-based isolation. 1.8.5.1 Ultracentrifugation Based Isolation Centrifugation sedimentation in general allows separation of a heterogeneous mixture based on its density, size, and shape. It plays an important role in fractionating small bio-particles such as viruses, bacteria, subcellular organelles as well as extracellular vesicles. Ultracentrifugation based exosome isolation is considered as the gold standard and is the technique of choice used in almost 56% of all exosome research(65,91). Sequential ultracentrifugation exosome isolation starts with cleaning steps that usually involve a series of centrifugation cycles of different centrifugal force and duration. Then, the exosomes are precipitated at 100,000 to 120,000g 23 (Figure 3). This exosome isolation approach is relatively easy to perform, affordable, and has no requirement for sample pretreatments. However, due to the heterogeneity of exosomes and overlap in size of extracellular vesicle populations, sequential ultracentrifugation isolation often suffers from contamination and exosome loss(65). Variants of ultracentrifugation such as density gradient ultracentrifugation have been developed to lessen the contamination and loss of exosomes. With this approach, a pre-constructed density gradient medium is established in a centrifuge tube with progressively increasing density when going from top to bottom. Exosome mixture is layered on to the top of the density gradient and then subjected to ultracentrifugation. The centrifugal force moves the exosome mixture through the density gradient medium towards to the bottom, and exosomes separate in discrete density zones. Density gradient separation is often coupled with sequential ultracentrifugation or other approaches to further purify the exosomes(65); however, density gradient separation is limited by its capacity and is time consuming (65). 24 Figure 3. Sequential ultracentrifugation exosome isolation. Culture supernatant or biological fluid is spun at lower speeds (1000g and 10,000g) to remove live/death cells, alone with cell debris. Then the supernatant is subjected to high speed centrifugation (100,000g) to collect exosomes. Additional wash is needed after exosomes are collected in order to remove other contaminants. 1.8.5.2 Size-Based Exosome Isolation Ultrafiltration is one of the size-based exosome isolation techniques which separates the exosomes based on their size or molecular weight using a relatively low centrifugation speed. It has been suggested that ultrafiltration has the highest exosomal RNA yield from urine when compared to the ultracentrifugation approach(92). In addition, ultrafiltration approach is faster than ultracentrifugation exosome isolation, and does not require special equipment. However, fixed pore sizes in the Nano-membrane may force the deformation and breaking-up of the large vesicles which may skew the downstream analysis(93). 25 1.8.5.3 Immuno-Affinity Capture-Based Exosome Isolation The presence of membrane proteins on the surface of exosomes provides an excellent site to develop highly specific immuno-affinity-based exosome isolation techniques. For example, the enzyme-linked immunosorbent assay (ELISA) based microplate for exosome isolation was developed for capturing and quantifying exosome from biological fluids such as plasma, serum and urine. The ELISA based approach was demonstrated to produce comparable results to those obtained from ultracentrifugation but with much less sample volume (65). ELISA based approach is ideal for isolating pre-enriched exosomes samples, but cannot be used directly to extract out exosomes from original sample due to the abundance of non-exosomal contaminants (65). 1.8.5.4 Precipitation-Based Isolation Polyethylene glycols (PEGs) are polymers that are routinely used for precipitation of viruses and other particles (94). This is the principle used for development of precipitation-based exosome isolations. ExoQuick, a proprietary reagent developed by System Biosciences, and Total Exosome Isolation, launched by Life Technologies, are two well-known precipitation-based exosome isolation reagents. These two reagents share the same core components which function to tie up water molecules and force less-soluble components, such as exosomes, out of solution by utilizing low-speed centrifugation(94). Precipitation-based exosome isolation is easy to use and there is no requirement for specialized equipment, and it can be scalable in clinic usage for large sample size (65). The disadvantage of precipitation-based isolation is the co-sedimentation of other non-exosome contaminants such as protein aggregates or lipoproteins(91). Together, existing exosome isolation approaches each have their own advantages and disadvantages, lacking a one-size-fits-all isolation approach. Since 2012, combined different 26 isolation approaches has been emerging (95). However, combination of these techniques usually results in higher cost, longer duration, and complicated steps, which can all potentially result in greater technical error rates and lower recoveries (65). Hence the users should keep in mind the inherent advantages and disadvantages of each technique. 1.8.6 Exosome Quantification/Characterization As mentioned previously, there are many approaches being developed for exosome quantification including ELISA based quantification, flow cytometry, NanoSight tracking analysis (NTA), and Dynamic/Static Light Scattering (DLS/SLS). With the advancement of the field, more methods are in development with the goal of enabling the user to measure exosome size and quantity simultaneously. 1.8.6.1 Flow Cytometry Flow cytometry is one of the most commonly used techniques for extracellular vesicles detection. In this technique, fluorescently labeled extracellular vesicles such as exosomes are excited by a laser beam at specific wavelength, and scattered fluorescence intensity is collected. Then, recorded parameters are used to analyze the relative vesicle size and concentration. Flow cytometry enables a high throughput and the ability to detect multiple markers simultaneously, which is ideal for quantifying vesicle populations. However, conventional flow cytometry can only detect particle size larger than 500nm, and has difficulties measuring vesicles in the lower size range such as exosome (40-150nm) (96). In this case, smaller vesicles are identified as background noise and reduce the accuracy (97). New techniques have been developed to combat this issue, usually involving the aggregation of exosomes with antibody-coated latex beads, therefore increasing the detection by conventional flow cytometry(98). 27 1.8.6.2 Dynamic and Static Light Scattering The principle of Dynamic Light Scattering (DLS) is based on the analysis of fluctuations in scattered intensities of the nanoparticles that are illuminated by a laser beam. In solution, these fluctuations are caused by the Brownian motion of the particles. Based on the Brownian motion equation, the velocity of suspended particles is related to hydrodynamic radius and other experimental parameters such as temperature and viscosity. Then, the particle size distributions can be extrapolated by fitting the light fluctuation reading to a mathematical model and applying the Stokes-Einstein equation assuming that all objects are spheres (99). On the other hand, Static Light Scatting (SLS), also known as multi-angle light scattering, is a measurement of the light intensity scattered by nanoparticles in a solution. It is usually used to determine the concentration and average size of the nanoparticles(99). When used in conjunction, DLS and SLS are highly sensitive, require very little sample volume, and are easy to perform; therefore, both DLS and SLS are widely used in different scientific fields especially in extracellular vesicle measurement. The greatest advantage of the DLS approach is its ability to measure vesicle range from 1nm to 6um(96), and DLS and SLS are very effective in analyzing homogenous particles in solution (100). However, several drawbacks have been highlighted in the case of polydisperse samples, the results become ambiguous. Additionally, the presence of few high-intensity particles may mask the data of small particles (100). To overcome the limitations of DLS, several recent studies have used DLS in combination with other quantification approaches to improve the accuracy. One of studies is done by Kesimer and Gupta to characterize extracellular vesicles from cultured airway epithelial cells using a combination of DLS and SLS coupled with size exclusion separation and electron microscopy(101). 28 1.8.6.3 NanoSight Tracking Analysis (NTA) NTA is an alternative method to DLS that is widely used in characterization of extracellular vesicle. Similar to DLS, NTA is also based on Brownian motion to measure particle size and concentration. NTA is normally equipped with a complementary metal-oxide-semiconductor (CMOS) camera or a sensitive charge-coupled device (CCD) microscopy which are used to record and visualize illuminated particles over a certain time period (96). The NTA software is then able to identify and track individual nanoparticles moving in solution and calculate particle size distribution based on the Stokes-Einstein equation. In addition, nanoparticle concentration can be determined after estimating the sample volume. Using NTA for extracellular vesicle measurement has several advantages, the greatest advantage over DLS being its ability to measure polydisperse samples as a result of its method of measuring nanoparticles on an individual basis without using ensemble-averaged signal (96). While NTA is more sensitive for detecting nanoparticles with diameter greater than 50nm (96), some smaller exosomes may not be detected using this approach. Moreover, NTA is also limited in measuring complex biological suspensions due to being unable to distinguish from other particles or protein aggregates(99). 1.8.6.4 Western Blot Analysis There are minimal experimental requirements for identify extracellular vesicles(102). Western blot is one of the most commonly used techniques for exosome characterization. Since secreted extracellular vesicles such as microvesicles and exosomes shared similar markers, but differ in relative proportion, therefore, at least 3 exosomal markers are needed to identify isolated exosomes (102). Exosome signals should be enriched when compared with signals from biological 29 fluids or conditioned medium(102). In addition, negative controls such as non-exosomal markers or cellular organelle markers should be reported(102). 1.8.6.5 Transmission Electron Microscopy (TEM) As a general rule in the field, beside semi-quantitative techniques such as western blot, additional techniques are required to characterize isolated exosomes such as TEM (102). 1.8.7 Exosomes and Diabetes Type I diabetes Type I diabetes results from autoimmune destruction of insulin secreting beta-cells in the pancreas. Sheng et al. demonstrated that pancreatic islet releases beta-cell autoantibodies via exosomes and may promote beta cell autoimmunity in type I diabetes (103). Type II diabetes Exosomes have been implicated in the development of insulin resistance commonly found in type II diabetes. Exosomes released from adipose tissue of obese mouse induce differentiation of the monocytes into activated macrophages, which contribute to insulin-resistant phenotype in a mouse model (104). 1.8.8 Exosomes and Diabetic Retinopathy In recent years, the role of exosomes in DR is beginning to unfold. It’s been reported that cytokines and angiogenic factors in circulating exosomes are increased in diabetic patients, and formation of specific cytokines in circulating extracellular vesicles is strongly influenced by duration of the disease (105). In a separate study, increased level of platelet-derived extracellular vesicles was associated with the progression of DR (106). In the vitreous, level of extracellular vesicles increased in patients with DR, which could contribute to the progression of the disease 30 (107). αβ-crystallin, an anti-apoptotic and anti-inflammatory protein, secreted via exosomal pathway on the apical surface of retinal pigment epithelium (RPE) cells, may protect the eye from oxidative stress such as in DR(108). In addition, exosomes from retinal astroglial cells contain anti-angiogenic components, that might protect the eye from neovascularization(109). Extracellular vesicles package cellular information from the secreting cells and their cargo alters based on stimuli such as cytokines, glycemic levels, and drugs. They serve as a reservoir of novel biomarkers studied in various pathological models or when looking at drug effects. The increased release of extracellular vesicles from cells or accumulation in biological fluids is often directly related to cell activation and pathogenesis of a variety of diseases(110). In DR, elevation of monocyte-(111) and platelet-(106) derived extracellular vesicles have been suggested to have a strong correlation to the progression of the disease. 1.9 Objective of the Dissertation It is well accepted that a chronic inflammation contributes to retinal vascular damage in DR, which can lead to vision impairment. Dysregulation of Complement cascade has been implicated in progression of DR. Extracellular vesicles such as exosomes are found in all human biological fluids and they carry cargos that can be exchanged between different cells. Complement proteins are associated with exosomes in circulation. We therefore initially proposed the central hypothesis of the study that circulating extracellular vesicles such as exosomes contribute to complement activation and development of DR. • We found that immunoglobulin laden plasma exosomes activate complement protein C1 and contribute to increased retinal vascular permeability in diabetes. 31 • We further explored the mechanism of exosome-induced DR pathogenesis showing that diabetic exosome-associated complement activation participates in MAC deposition and cytolytic damage in retinal endothelial cells. This dissertation will provide a novel mechanism in which exosomes-mediated complement activation contributes to development of DR. A schematic representation of the rational of the dissertation is shown in Figure 4. Figure 4. Rationale of the dissertation. In diabetic retinopathy, circulating exosomes activate classical complement cascade and lead to cytolytic MAC formation on the retinal vasculature. 32 1.9.1 Overview of Chapters Chapter I is a review of literature encompassing diabetic retinopathy and the potential contribution of complement system and role of exosomes in diabetic retinal vascular damage. Chapter II investigates complement system activation by plasma exosomes and its contribution to retinal vascular damage. In this chapter, a novel exosome quantification method was developed. The experiments demonstrated in this chapter suggest that increased number of IgG laden exosomes in diabetic plasma may contribute to increased complement activation and elevated retinal vascular permeability. Chapter III explores the hypothesis that diabetic plasma exosomes contribute to retinal vascular damage via MAC deposition. The experiments in this chapter elucidated that diabetic IgG-laden exosomes contribute to MAC deposition and participates in cytolytic damage in human retinal endothelial cells (HRECs). Chapter IV summarizes the obtained data, and conclusions that can be drawn from the studies, addresses question raised by the data, and suggests future direction of the research related to presented findings. 33 Chapter 2. Plasma exosomes contribute to microvascular damage in diabetic retinopathy (DR) by activating classical complement pathway This chapter is a modified version of the manuscript accepted by Diabetes Authors: Chao Huang1, Kiera P. Fisher1, Sandra S. Hammer1, Svetlana Navitskaya1, Gary J. Blanchard2, Julia V. Busik1 2.1 ABSTRACT Diabetic Retinopathy (DR) is a micro-vascular complication of diabetes and is the leading cause of vision loss in working-age adults. Recent studies have implicated the complement system as an emerging player in development of vascular damage and progression of DR. However, the role and activation of the complement system in DR is not well understood. Exosomes, small vesicles that are secreted into the extracellular environment, have a cargo of complement proteins in plasma suggesting that they can participate in causing vascular damage associated with DR. We demonstrate that IgG-laden exosomes in plasma activate the classical complement pathway, and that the quantity of these exosomes is increased in diabetes. Moreover, we show that lack of IgG in exosomes results in a reduction of retinal vascular damage in diabetic mice. Together, the results of this study demonstrate that complement activation by IgG-laden plasma exosomes could contribute to the development of DR. 2.2 INTRODUCTION Vascular inflammation and activation of complement system is of significance in many tissues and especially critical for the brain and retina where normal blood brain and blood retinal barrier can be disrupted by complement activation (44,112). As pro-inflammatory changes contributing to blood-retinal barrier breakdown represent an important initiating factor in the pathogenesis of DR(113), and activation of complement system and impairment of complement regulatory 34 proteins was observed in the eyes of DR patients and in animal models (3,112,114), this study addressed the role of complement activation in vascular damage in animal and ex vivo models of DR. The complement system consists of a large group of small, inactive precursors found in the circulation and plays a central role in host defense against infectious pathogens through activation of an inflammatory response. The complement system can be activated via three pathways: classical, alternative, and lectin pathway. All three complement activation pathways lead to generation of C3/C5 convertases, and ultimately lead to formation of the membrane attack complex (MAC). In the classical complement pathway, C1 is the first complement protein that is activated. C1 is a large protein complex comprising C1q, C1r, and C1s. Downstream proteolytic activation is achieved when C1q binds to a classical activator, such as an immunoglobulin complex (37). In DR, elevated complement protein deposition in the retinal vascular lumen has been suggested to be, in part, due to alternative pathway activation and reduced level of complement inhibitor proteins CD55 and CD59 (56,57). Classical pathway complement proteins have not been found in the retinal vascular lumen (57); however, they are significantly elevated in vitreous fluid from patients with DR compared to age-matched non- diabetic patients (115). In addition, C1 inhibitor, a circulating regulatory protein belonging to the classic complement pathway, has been shown to be impaired in diabetes by a hyperglycemic byproduct methylglyoxal (116). Taken together, these studies suggest the involvement of the classical complement pathway in the pathogenesis of DR. The mechanism(s) of complement activation and the contribution of complement activation to DR pathogenesis remain unknown. 35 Intriguingly, proteomic studies have shown that classical complement proteins like C1q, C3, and C4 are found in association with extracellular vesicles such as exosomes in plasma (117–119). Exosomes are small extracellular vesicles that are 40-200 nm in diameter. Exosomes are found in most human biological fluids, including blood (120,121), urine (122), cerebrospinal fluid (123) and ascites (124). Initially, secretion of exosomes was proposed as a mechanism for removal of unwanted proteins from cells (61). In recent years the role of exosomes has been largely expanded due to their involvement in intercellular communication. Exosomes released from different cells were shown to carry cargos that have different effects on target cells, acting in a paracrine manner. For example, in ocular cells, exosomes released from retinal astroglial cells containing anti-angiogenic components might serve as protection from neovascularization, while exosomes from retinal pigment epithelium cells have not shown this anti-angiogenic role (125). Beside shuttling cytokines and growth factors, exosomes also carry genetic material such as mRNAs, miRNAs, and lipids. Recently, it was suggested that immunoglobulins have the capability to associate with exosomes in circulation (126). Immunoglobulin is a potent activator of the classical complement cascade, and was reported to be increased in patients with diabetes when compared to non-diabetic patients (127). In this study, we investigated the role of exosomes in the activation of the classical complement pathway and downstream retinal vascular damage during DR. 2.3 RESULTS Quantification of exosomes in control and diabetic plasma. Quantification of circulating exosomes is challenging due to the lack of stably expressed exosomal markers. As it was important for the goals of this study to have reliable exosome quantification methods, combined 36 dynamic and static light scattering (DLS and SLS) assays were developed using Zetasizer Nano NZ (Malvern Instruments Ltd, Malvern, UK) to quantify and compare the number of exosomes in plasma. Extruded micro-vesicles of known size, composition and concentration were extruded using an Avanti Mini-Extruder (Avanti Polar Lipids, Inc.) with 100-nm pore filters and used to create calibration curves (particle/mL). DLS readings of extruded vesicles were compared with isolated mouse plasma exosomes (Figure. 5A, B). Vesicles extruded through 100nm pore filter had similar diameter to mouse plasma exosomes (Figure. 5C), which was further confirmed by electron microscopy (Figure. 5 D, E). The absolute concentration of extruded micro-vesicles was calculated based on lipid composition and vesicle surface area. Serial dilution of 100nm vesicles at initial concentration of 0.64x10^13 vesicles/mL was used to generate the calibration curve by plotting absolute sample concentration (particle/mL) verses SLS count rate (kcps) (Figure. 5F). The number of circulating exosomes in control and diabetic blood plasma was determined by comparing SLS readings of the plasma exosomes to the standard curve created using extruded vesicles of known size and concentration. To prevent sample loss during ultracentrifugation isolation procedure, exosomes were isolated from equal volume of mouse plasma (100uL) by the ExoQuick precipitation method. We found that there was a significant increase in the concentration of exosomes in STZ-induced diabetic (3 months) mouse plasma when compared to control (Figure. 5G). To confirm the SLS observation, we performed Western Blot analysis of several exosomal markers. Under equal volume-loading conditions, STZ-induced diabetic (3 months) mouse plasma had higher expression of exosomal markers CD63, CD9, and TSG101 compared to age-matched controls (Figure. 6A). We next measured the diameter of mouse diabetic exosomes compared to control exosomes with DLS. We found no difference in average 37 diameter between control and diabetic exosomes (Figure. 6B). Moreover, CD63 and CD9 expression showed no difference between control and diabetic exosomes under equal number of vesicles loading conditions (Figure. 6C). These findings suggested that there are more circulating exosomes in diabetic plasma as compared to control; furthermore, the size and composition of exosomes are less affected by diabetes. 38 Figure 5. Comparison of plasma exosomes, isolated with ExoQuick, and 100-nm extruded microvesicles. (A) and (B): DLS measurement of 100-nm extruded micro-vesicles (A) and plasma exosomes (B). d.nm, diameter, in nanometers. (C) DLS measurement showed no difference in diameter between 100-nm extruded microvesicles (spotted bar, n=9) and control mouse exosomes (grey bar, n=10). (D) and (E): Electron microscopy images of 100-nm extruded vesicles (D) and plasma exosomes (E). (F) Vesicle concentration (particles per milliliter) plotted against SLS measurement (kilocounts per second [kcps]). (G) Number of plasma exosomes was increased significantly in diabetic mouse plasma. The number of exosomes isolated from plasma of control (n=9, grey bar, grey circles) and STZ induced diabetic (3 months) mouse plasma (n=15, white bar and black squares) was measured using SLS (kcps) reading (bottom table). All data points are shown as univariate scatter plots with mean and standard deviation, * P < 0.05. Figure 1 A ) % ( y t i s n e t n I 20 15 10 5 0 0.1 1 10 100 Size (d.nm) 1000 10000 D E 7.E+10 6.E+10 5.E+10 4.E+10 3.E+10 2.E+10 1.E+10 0.E+00 G / l i L m s e c s e v f o r e b m u N 2.5×1011 2.0×1011 1.5×1011 1.0×1011 5.0×1010 0.0 * Control Diabetic Sample Control Diabetes SLS Intensity (kcps) Number of Vesicles/mL 835 1240 7.35x10^10 1.14x10^11 y = 1E+08x - 1E+10 R² = 0.98472 0 500 Intensity (kcps) 1000 39 1 10 100 Size (d.nm) 1000 10000 F L m s e / l i c s e V f o r e b m u N 100nm Vesicles Mouse Exosomes 200 150 100 50 0 10 6 4 2 0 0.1 B ) % ( y t i s n e t n I C ) m n ( r e t e m a i D Figure 6. Western blot of plasma exosomes isolated through the use of ExoQuick from control and STZ-induced diabetic (3 months) mice. Under condition of equal volume, increased amounts of exosomal markers (CD63, CD9 and TSG101) were measured in diabetic mouse plasma (bottom). The markers were quantified on the basis of average band intensity (n = 9, 8, and 7 for the three control measurements (grey bars and grey circles), and n = 15,9, and 9 for diabetes measurements (white bars and black squares). *P < 0.05. (B) DLS (nanometer) readings of mouse plasma exosomes showed no change in exosome diameter (DLS) between control mice (n = 10; grey bar) and diabetes (n = 15, white bar). (C) Western blot analysis of plasma exosomes isolated with ExoQuick from control and STZ-induced diabetic (3 months) mice. Under loading condition with an equal number of vesicles, no change in plasma exosomes (CD 63 and CD9) was observed between control (grey bars and grey circles) and diabetic mice (white bars and black squares). Exosomes were quantified on the basis of average band intensity (n = 10 [middle] and 4 [bottom] for control and n = 12 [middle] and 6 [bottom] for diabetes). All data points are shown as univariate scatter plots with mean and standard deviation. Figure 2 A Equal Volume-Loading Control Diabetes B CD63 CD9 TSG101 y r t e m o t i s n e D e v i t a e R l s t i n U 25000 20000 15000 10000 5000 0 54kDa 24kDa 52kDa * * 25000 20000 15000 10000 5000 y r t e m o t i s n e D e v i t a l e R s t i n U 25000 20000 15000 10000 5000 ) m n ( r e t e m a D i 200 150 100 50 0 Control Diabetic Sample Control Diabetes DLS Diameter (nm) 116.1± 20.4 113.2± 19.3 * s t i n U y r t e m o t i s n e D e v i t a e R l 0 Control Diabetic CD63 0 Control Diabetic CD9 Control Diabetic TSG101 Equal Number of Vesicles Control Diabetes C CD63 CD9 54kDa 24kDa l e v i t a e R 3 6 D C s t i n U y r t e m o t i s n e D l e v i t a e R 9 D C s t i n U y r t e m o t i s n e D 20000 15000 10000 5000 0 15000 10000 5000 0 Control Diabetic Control Diabetic Immunoglobulins are associated with exosomes in plasma. As shown in Figure. 7A, we detected an enrichment of exosomal markers and large amount of IgG in ExoQuick isolated mouse plasma exosomes (Figure. 7A, left lane). Concurrently, we also detected non-exosomal markers such as LDL (ApoE), HDL (ApoA), and ER markers (Calnexin) in ExoQuick isolated exosomes. Several reports have shown that ExoQuick and other polymeric-based exosome isolation methods co- 40 precipitates other proteins or vesicles such as lipoproteins (95,128), and further separation by OptiPrep density gradient centrifugation is required to obtain pure exosomes (31,32). To confirm that immunoglobulins are associated with plasma exosomes, mouse plasma exosomes isolated by ExoQuick were further purified via OptiPrep density gradient centrifugation, and then evaluated via Western blot using exosome markers CD63, CD9, and TSG101. Consistent with previously reported plasma exosomal density range (120,130), OptiPrep gradient separation yielded mouse plasma exosomes in fractions 8 to 10 with a density range of 1.16 to 1.39 g/mL (Figure. 7B). We detected prominent bands of IgG in the same fractions where exosomes are present (Figure. 7B, bottom). After depleting exosomes from mouse plasma by repeating ultracentrifugation, we found a concomitant reduction of IgG level in exosome-depleted mouse plasma (Figure. 7C, right lane). These results suggest that most of immunoglobulins in circulation are associated with exosomes. 41 Figure 7. Characterization and specificity of Ig binding to exosomes. (A) Western blot analysis of exosomes from mouse plasma isolated with ExoQuick, with a 20-ug protein loading volume. Exosomes isolated from mouse plasma showed enrichment of exosomal markers (CD63, ALIX TSG101, CD9) than whole plasma. Specificity of exosome isolation was confirmed using lipoprotein markers (ApoE and ApoA) and the endoplasmic reticulum marker Calnexin. (B) Western blot of total circulating exosomes. Exosomes were separated with OptiPrep Density Gradient after ExoQuick isolation. The exosomal markers CD63, TSG101, and CD9, as well as Ig, are present in fractions with a density of 1.16-1.39 g/ml (fractions 8 to 10). (C) Ig is removed in exosome-depleted plasma (right lane). Figure 3 OptiPrep Gradients 7 1.12 9 1.24 8 1.16 A Exosome Plasma IgG CD63 ALIX TSG101 CD9 ApoE ApoA Calnexin B Fractions # Density (g/mL) CD63 5 1.08 6 1.10 TSG101 CD9 IgG 50kDa 25kDa 54kDa 55kDa 52kDa 24kDa 36kDa 31kDa 75kDa 10 1.39 11 1.42 C Exosome Plasma Exosome depleted plasma CD63 TSG101 CD9 IgG 54kDa 55kDa 24kDa 50kDa 25kDa 54kDa 52kDa 24kDa 50kDa 25kDa Ultracentrifugation remains the gold standard for exosome isolation and less contaminants are reported with this method than with ExoQuick(65). Therefore, to confirm that IgG associates with exosomes in circulation, ultracentrifugation-isolated rat exosomes were stained with immuno- gold labeled IgG and investigated under TEM. Exosomes appeared as a disk shape with diameter of approximately 150nm (Figure. 8A). Immuno-gold labeled TEM showed a strong IgG association with the rat plasma exosomes (Figure. 8B). Interestingly, an increase in the number of circulating exosomes in diabetic animals was associated with the rise in IgG levels in diabetic mouse samples. This increase in IgG was measured via western blot analysis under equal volume-loading conditions (Figure. 8C) but was not observed under equal number of vesicles loading condition 42 (Figure. 9). These data confirm that it is an elevated number of circulating vesicles, but not a change in exosome morphology, that results in higher IgG levels in diabetic exosome samples when compared to non-diabetic controls. Figure 8. IgG bound to exosomes in plasma and an elevated number of exosomes in diabetes leads to greater IgG level in mice with diabetes than in control mice. TEM images of rat plasma exosomes isolated by ultracentrifugation. (A) 138.89-nm-diameter exosome was visualized using TEM with uranyl-oxalate staining. (B) Immunogold-labeled rat IgG colocalized with exosomes (arrowheads). (C) Western blot of mouse plasma exosomes isolated via ExoQuick (left). Under loading conditions of equal volume, more exosomes containing IgG were observed in diabetic mouse plasma (n = 12, white bar, black squares) than in control plasma (n = 10; grey bar, grey circles) (right). Exosomes were quantified on the basis of band intensity. All data points are shown as univariate scatter plots with mean and standard deviation. *P < 0.05. Figure 4 e v i t a l e R G g s t i n U y r t e m o t i s n e D I e s u o M 25000 20000 15000 10000 5000 0 * Control Diabetic A B C Equal Volume-Loading Control Diabetes IgG 50kDa 24kDa 43 Figure 9: Western blot analysis of mouse plasma exosomes isolated via ExoQuick. (left) Under Supplemental Figure 1 equal number of vesicles, no change was observed between control and diabetic mouse plasma exosomes. (right) Quantification based on band intensity. (n=10 for control and n=15 for diabetes, *P < 0.05). Equal Number of Vesicles Control Diabetes IgG 50kDa 24kDa e v i t a l e R G g s t i n U y r t e m o t i s n e D I e s u o M 40000 30000 20000 10000 0 Control Diabetic S-Figure 1. Western blot analysis of mouse plasma exosomes isolated via ExoQuick (left). Under equal number of vesicles, no change was observed between control and diabetic mouse plasma exosomes (right). Quantification based on band intensity. (n=10 for control and n=15 for diabetes, *p<0.05 ). 44 Circulating immunoglobulin bound to exosomes activates C1 complex. To examine if the complement cascade can be activated by immunoglobulin-associated exosomes, C1q binding assay (131) was performed. Mouse plasma exosomes were used in this assay and the results were further confirmed using non-diabetic human plasma. Purified human C1q protein (CompTech, A099) was incubated with isolated exosomes and purified via OptiPrep gradient centrifugation. As shown in Figure. 10A, exogenous human C1q binds to ExoQuick isolated mouse plasma exosomes. To verify the specificity of C1q binding, we fractionated the exosome-bound C1q via OptiPrep density gradient (Figure. 10B). C1q, TSG101, and mouse IgG were detected at the exosomal fraction 9 (1.24g/mL). The purity of exosomes in fraction 9 was confirmed by the absence of lipoproteins (ApoE and ApoA) and ER marker (Calnexin). We also performed the same experiment by using non-diabetic human plasma exosomes and found that endogenous C1q was associated with circulating exosomes and was recognized by anti-human C1q antibody (Figure. 10C, left lane). After OptiPrep gradient centrifugation of human exosomes, we observed that C1q and CD63 was also found in fraction 9 (Figure. 10D). C1q alone was not present in any of these fractions (Figure. 11). To investigate whether the binding of C1q to immunoglobulin-containing exosomes activates C1 complex in the classical complement cascade, we performed the C1 activation assay (132). Non- diabetic human exosomes were incubated with C1 complex protein (CompTech, A098) and showed activation of C1s in fraction 9 (Figure. 10E, left panel). In the presence of C1 inhibitor (INHC1), however, the activation of C1s was reduced (Figure. 10E, right panel). Spontaneous activation of the C1 complex was observed in the reaction via Western blot analysis (Figure. 12); however, spontaneous C1 activation was not present in fractions without exosomes. Our data 45 demonstrate that immunoglobulin-laden exosomes bind to C1 complex via C1q in plasma and causes proteolytic activation of C1. Moreover, under conditions in which equal number of vesicles are loaded, there was no difference in C1 activation between exosomes from control and diabetic mouse plasma (Figure. 10F). Thus, these experiments suggest that the elevated number of exosomes observed in diabetes may lead to the activation of the classical complement pathway. 46 Figure 10. Western blot of complement activation exosomes isolated from mouse and human plasma through ultracentrifugation or ExoQuick and purified with OptiPrep Density Gradient. (A) Binding of human complement protein C1q to ExoQuick-isolated mouse exosomes. (B) C1q specifically bound to mouse plasma exosomes with a density of 1.24 g/mL (fraction 9). This fraction was negative for ApoE, ApoA and Calnexin. (C) Endogenous human C1q was detected in non-diabetic human plasma exosomes isolated via ultracentrifugation. (D) After centrifugation in the OptiPrep gradient, human non-diabetic plasma fractions 7 - 9 were ran. C1q and CD63 were expressed in fraction 9. (E) C1 was activated by ultracentrifugation-isolated human plasma exosomes. A C1 activation assay was performed by incubating C1 complex with human plasma exosome, fractioning them with OptiPrep Density Gradient. The C1 complex bound to exosomes with a 1.24g/mL density (fraction 9) and activated serine protease subcomponent C1s (left). The activity of the C1s was inhibited by a C1 inhibitor (INHC1) (right). (F) Under conditions that used an equal number of vesicles, no difference was observed in C1 activation in plasma exosomes isolated with ExoQuick from control and STZ-induced diabetic (3 months) mice (n=3 for both groups). Figure 5 Exosomes Exosomes C1q C Exosomes Exosomes C1q C1q CD63 IgG A B Exosomes + C1q Fractions Density(g/mL) 7 1.12 8 1.16 9 1.24 C1q TSG101 ApoE Calnexin ApoA IgG E Fractions Density(g/mL) C1s Activated Activated CD63 F TSG101 C1s Activated Activated Exosomes C1 8 7 1.12 1.16 Exosomes C1+INHC1 9 1.24 7 8 1.12 1.16 9 1.24 Equal Number of Vesicles Diabetes Control 86kDa 58kDa 28kDa 54kDa 52kDa 86kDa 58kDa 28kDa C1q CD63 IgG 30kDa 54kDa 54kDa 24kDa D Exosomes + C1q Fractions Density (g/mL) 7 1.12 8 1.16 9 1.24 C1q CD63 30kDa 54kDa 30kDa 54kDa 50kDa 24kDa 30kDa 54kDa 36kDa 75kDa 31kDa 50kDa 24kDa 47 Supplemental Figure 2 Figure 11. Western Blot analysis of C1q binding assay without presence of exosomes isolated by ultracentrifugation or/and OptiPrep density gradient. No C1q present in fraction 9. C1q Fractions Density(g/mL) 7 1.12 8 1.16 9 1.24 C1q C1q CD63 26 kDa 53 kDa Figure 12. Western Blot analysis of C1 auto-activation after ultracentrifugation or/and OptiPrep density gradient without presence of exosomes. (left) C1 shows auto-activation with activated C1s and its prevented by INHC. (right) Without exosomes, no auto-activated C1s is presented in Supplemental Figure 3 fractions 8 to 11. S-Figure 2. Western Blot analysis of C1q binding assay without by ultracentrifugation or/and OptiPrep density gradient. No C1q present in fraction 9. exosomes presence of isolated C1 C1 + INHC1 C1 C1 INHC1 8 9 1.16 1.24 10 1.39 11 1.42 8 1.16 9 1.24 10 1.39 11 1.42 Fractions Density(g/mL) 86kDa 58kDa 28kDa C1s Activated Activated S-Figure 3. Western Blot analysis of C1 auto-activation after ultracentrifugation and/or OptiPrep density gradient without presence of exosomes (left). C1 shows auto-activation with activated C1s and is prevented by INHC1 (right). Without exosomes, no auto-activated C1s is presented in fractions 8 to 11. 48 Lack of classical complement activation in diabetic Ighmtm1cgn/J (IgM-KO) mice results in a reduction of retinal vascular damage. To assess the role of classical complement pathway involvement in DR, a transgenic mouse that lacks complement activator (IgG) was used. Ighmtm1cgn/J (IgM-KO) is a B-cell deficient mouse (Ig- deficient) model available through JAX laboratory. These mice lack mature B-cells due to disruption in the heavy chain of the IgM, causing B-cell maturation to be arrested at the pre-B cell stage which results in no production of immunoglobulin (133). First, we confirmed that IgM- KO mice showed no expression of IgG in ExoQuick isolated plasma exosomes when comparing to exosomes from control mice (C57BL/6J) (Figure. 13A). Furthermore, we found that exosomes from IgM-KO mice (Figure. 13B, right panel) induced lower levels of C1 activation than exosomes from C57BL/6J mice (Figure. 13B, left panel). To investigate the role of classical complement pathway by IgG-laden exosomes in the development of diabetic retinal damage, male IgM-KO and C57BL/6J mice (10-12 weeks old, n=6) were subjected to STZ-induced diabetes for seven weeks. Hyperglycemia was observed in both strains of diabetic mice but not in controls. DLS measurement demonstrated that diabetes did not modify exosome size in both strains of mice when compared to non-diabetic controls (Figure. 14A). Exosome quantification confirmed that as early as seven weeks after onset of diabetes, an elevated number of exosomes is found in both strains of mice when compared to non-diabetic controls (Figure.14B). Higher number of exosomes in diabetic C57BL/6J mice results in greater IgG levels than in non-diabetic C57BL/6J mice (Figure. 14C, left); however, no IgG expression was observed in either control or diabetic IgM-KO plasma exosomes (Figure.14C, right). Importantly, there was a reduction in retinal vascular leakage in diabetic IgM-KO retinas (Figure. 14D, bottom panels, n=5) when compared to 49 diabetic retinas from wild type mice (Figure.14D, top panels, n=6). These results suggest that lack of IgG association with exosomes prevents classical complement pathway activation by exosomes, resulting in reduction of diabetes-induced retinal vascular damage. Figure 13. Western blot analysis of plasma exosomes isolated from C57bl/6J and IgM KO mice by ExoQuick and OptiPrep density gradient. (A) IgM-KO exosomes (right) showed no expression of IgG, but C57bl/6J exosomes did(left). (B) C1 was activated with exosomes from C57bl/6J mice (left), but not with plasma exosomes from IgM-KO (right) (n=3). Figure 6 C57bl/6J Exosomes IgM-Ko Exosomes B Fractions C57bl/6J Exosomes Density(g/mL) A TSG 101 IgG 52kDa 50kDa 25kDa C1 C57bl/6J Exosomes + C1 7 1.12 8 1.16 9 1.24 IgM-Ko Exosomes + C1 7 1.12 8 1.16 9 1.24 IgM-Ko Exosomes C1 86kDa 58kDa 28kDa 52kDa C1s Activated Activated TSG 101 50 Figure 14. Analysis of plasma exosomes, isolated with ExoQuick from control and STZ-induced diabetic (7 weeks) C57BL/6J and IgM-KO mice. (A) DLS measurements showed no difference in diameter of plasma exosomes from C57bl/6J and IgM-KO in control or diabetic mice (n = 6 for each group). (B) The number of plasma exosomes was increased significantly in both C57bl/6J and IgM-KO diabetic mouse plasma compared to controls. Exosomes were isolated with ExoQuick from plasma from control ( n = 6) and STZ-induced diabetic ( n = 5)mice and measured by using SLS (kilocounts per second), *P < 0.05, ** P < 0.01 . (C) Western blot analysis of ExoQuick-isolated mouse plasma exosomes (top). Under loading condition of equal volume, more diabetic C57bl/6J plasma exosomes contained IgG (grey bar) than control plasma exosomes (white bar). No IgG expression was measured within IgM-KO control or diabetic plasma exosomes. (D) Retinal permeability was examined in control and diabetic C57bl/6J and IgM-KO (7weeks control and diabetes, n = 6). Diabetic C57bl/6J retina showed more vascular permeability than control retinas (top), but retinal vascular permeability in IgM-KO mice did not significantly change between control and diabetes states (bottom). Retinal fluorescence intensity is quantified through the use of fluorescein isothiocyanate-albumin, as shown in the graph at the bottom (control, n = 6; diabetes, n = 5). A ) m n ( r e t e m a i D 250 200 150 100 50 0 Control Diabetes Control Diabetes C57bl/6J IgM-KO B 4×1011 / L m s e c s e V i l 3×1011 2×1011 1×1011 0 C Control Diabetes IgG J 6 / l b 7 5 C ) % ( y t i s n e t n I 10 8 6 4 2 0 0.1 1 10 100 100010000 Size (d.nm) ) % ( y t i s n e t n I 12 10 8 6 4 2 0 0.1 1 10 100 1000 10000 Size (d.nm) ) % O K M g - I ( y t i s n e t n I 12 10 8 6 4 2 0 0.1 ) % ( y t i s n e t n I 12 10 8 6 4 2 0 0.1 1 10 100 1000 10000 Size (d.nm) 1 10 100 1000 10000 Size (d.nm) s t i n U y r t e m o t i s n e D G g I e s u o M 20000 15000 10000 5000 0 Control Diabetes ** * D J 6 / l b 7 5 C Control Diabetes Control Diabetes C57bl/6J C57bl/6J IgM-KO IgM-KO Control Diabetes Control Diabetes O K M g I 50kDa 24kDa *** Control Diabetes Control Diabetes C57bl/6J IgM-KO ) U F R ( y t i l i b a e m r e P r a u c s a V l 0.025 0.020 0.015 0.010 0.005 0.000 * * Control Diabetes Control Diabetes C57bl/6J IgM-Ko 51 2.4 DISCUSSION The complement system plays an integral role in both innate and adaptive immune response. Given the destructive power of complement activation, it is not surprising that aberrant complement system regulation has been shown to be involved in a wide range of diseases from nonalcoholic steatohepatitis to age-related macular degeneration (134–136). Recent studies have shown that complement proteins deposited in eyes of diabetic patients may contribute to the pathogenesis of DR. Mannose-binding lectin (MBL) and alternative complement pathways were shown to contribute to complement activation in ocular diseases including DR (2, 37); however, no studies have investigated the role of the classical pathway in the pathogenesis of DR due to the lack of detection of classical pathway proteins in the retina of DR patients. We hypothesize that the classical complement pathway may contribute to DR development via activation on the surface of plasma exosomes, instead of on the lumen of blood vessels. Proteomic studies have shown that classical complement proteins such as C1q, C3 and C4 are associated with isolated plasma exosomes (78,129). In this manuscript, we demonstrate that diabetic plasma exosomes activate the classical complement pathway, which leads to downstream retinal vascular damage. DLS and SLS measurements demonstrated that although the number of exosomes is elevated in the plasma of diabetic mice compared to control, the diameter of exosomes remains the same between these two groups. Current exosome quantification methods have limitations, for example, conventional flow cytometry cannot discriminate exosomes from background noise; enzyme-linked immuno-absorbent assay (ELISA) is limited by the available exosome markers; Nanoparticle Tracking Analysis (NTA) requires large sample volume and the reading is affected 52 by operators (100,138). Among them, the DLS/SLS measurement is the most user friendly while still yielding relatively accurate and consistent results. Importantly, these measurements require the smallest sample volume (100,138–141). A drawback associated with DLS size measurement is that the technique is more accurate with mono-disperse vesicles than with poly-disperse vesicles such as plasma exosomes (100,138–141). Moreover, there is a lack of quantifiable standards available for SLS measurement of exosomes. The present study analyzed isolated plasma exosomes using DLS and found that isolated plasma exosome samples are closely clustered around 100nm. This particle size is consistent with previous reported exosome size measurements and are in close agreement with Transmission Electron Microscopy (TEM) data. In addition, a standard curve for SLS measurement was generated using extruded lipid vesicles of known diameter and lipid composition to mimic exosomes. Both DLS and TEM measurements confirmed that extruded lipid vesicles were similar in diameter and morphology to isolated plasma exosomes. Lipid vesicles such as liposomes have been used as a model to mimic exosomes in previous studies, and to optimize exosome purification methods and diameter measurement (142,143). Combined DLS/SLS with extruded vesicles as a standard may serve as quantification method for plasma exosomes. In this study, various exosome isolation methods were used including polymeric-based exosome isolation (ExoQuick), differential ultracentrifugation isolation, and OptiPrep density gradient enrichment. Each isolation method has its own advantages and was used for different purposes. Ultracentrifugation isolation yields relatively purer exosomes than ExoQuick (65) and works well for exosome characterization, but requires greater sample volume. Therefore, ultracentrifugation-isolated rat plasma exosomes were used for immuno-gold staining to 53 visualize IgG association of plasma exosomes under TEM in this study. ExoQuick isolation allows less variation between groups due to sample handling or isolation process, which is beneficial when comparing exosomes from clinical samples such as between control and diabetic groups (65). The drawback is that ExoQuick co-precipitates other non-exosomal contaminants such as lipoproteins (LDL, HDL) and cellular organelles (ER) (36), which could mask the role of exosomes and confound the study (91,144), requiring OptiPrep density gradient centrifugation to further purify the exosome mixture and eliminate most lipoproteins and cellular organelles. Isolation of plasma exosomes by ultracentrifugation followed by OptiPrep density gradient centrifugation was conducted in parallel with ExoQuick-OptiPrep to confirm that exosome-induced complement activation is not preparation dependent. This once again demonstrates the drawbacks of each exosome isolation method, calling for more stringent verification of isolated exosomes in future studies and continual development in exosome isolation methods. Previously, Ogata, N. et al showed that increased levels of extracellular vesicles in diabetic plasma could contribute to acceleration of DR progression (106). Elevated level of serum immunoglobulin in diabetes was observed in clinical studies, and was suggested to contribute to complications (127,145,146). Consistent with these findings, we observed that there is an increased level of circulating exosomes in diabetes, and that exosomes associate with IgG and activate classical complement proteins, suggesting that an elevated number of IgG-laden plasma exosomes in diabetes may result in greater classical complement activation. To test this hypothesis, we turned to genetically modified mouse models that lack components of this system. C1q-deficient mice develop phenotypes similar to systemic lupus erythematosus or rheumatoid arthritis (147), making it difficult to separate the C1q-deficiency vs. diabetes-induced 54 damage. B-cell deficient mouse model (IgM-KO), on the other hand, does not have this complication. Moreover, IgM-KO mouse model specifically lacks circulating IgG with the rest of the complement proteins intact, helping us address the hypothesis that it is the IgG-laden exosomes that contribute to diabetes-induced complement activation. Based on these considerations, IgM-KO mouse model was used in the study. IgM-KO mice developed hyperglycemia after STZ induction at the same rate and severity as wild type C57BL/6J mice. Moreover, diabetic IgM-KO mice had increased number of exosomes compared to littermate controls, similar to diabetic C57BL/6J mice. Despite the increase in the number of exosomes in diabetic IgM-KO, the exosomes from IgM-KO mice were lacking IgG and had very low levels of C1 activation when compared to C57BL/6J exosomes. In biological fluids such as plasma, exosome populations are highly heterogeneous and originate from various cell types. Previously, Saunderson, S et al. reported that immunoglobulins are enriched in exosomes released by primary B-cells (148), and suggested that immunoglobulins present on the B cell plasma membrane could be shuttled into the exosomes (149). On the other hand, Blanc, L. et al. have demonstrated that reticulocyte-secreted exosomes have a natural affinity for immunoglobulins in blood (150). These findings suggest distinct possibilities for IgG association with exosomes: as primary secretion of IgG packaged inside the exosomes of IgG- producing B-cells, or as secondary association of exosomes potentially produced by multiple cell types with circulating immunoglobulins in blood. As complement complex C1 binding and activation requires surface presentation of IgG, only the exosomes with surface IgG would contribute to the activation. Indeed, we observed that it is only a fraction of the IgG-associated exosomes of specific density that activate complement complex C1. Previously, Diebolder et al. 55 (42) reported that the presentation of the Fc-Fc segment of IgG hexamers drives classical complement C1 activation. Based on this, we speculate that a sub-population of exosomes that activate the C1 complex may present the Fc motif which favors the activation of the classical complement pathway. Notably, the exosome sub-population that activates C1 in our study was found to be enriched with several exosomal markers: CD63, TSG101, and CD9. These markers were shown to be directly involved in different exosomal cargo sorting processes in the endosomal compartment (66,67). TSG101 is a component of the Endosomal Sorting Complexes Required for Transport (ESCRT) complex which participates in ESCRT-dependent exosome biogenesis (67). In contrast, tetraspanin family proteins CD63 and CD9 are responsible for ESCRT- independent exosome biogenesis (66). These two exosomal biogenesis mechanisms can act independently or simultaneously within an endosomal compartment, generating highly complex and heterogeneous exosomal populations (66). Exosomes found within fraction 9 are likely to be a mixed population generated by a combination of exosomal biogenesis mechanisms. These data are in agreement with a recent study showing that RPE-derived vesicles in fraction 9 contain complement components (13). Under normal conditions, low level of complement activation protects the eye from pathogenic infection. This process is tightly controlled by complement regulatory proteins which prevent complement-induced self-destruction (53). In diabetes, protein glycosylation and hyperglycemic byproducts greatly reduce the function of complement regulatory proteins in circulation and on cell membranes (116,151), suggesting that spontaneous alternative complement activation plays a role in MAC (C5b-9)-induced retinal vascular damage (44,56). Additionally, formation of cytolytic MAC leads to osmotic imbalance and ultimately leads to the lysis of cells. Extensive 56 deposition of MAC in the retinal endothelium of DR patients and in rat diabetes model could contribute to retinal endothelial cell death and resulting retinal vasculature leakage (57). To demonstrate that an increase in classical complement activation by IgG-laden exosomes under diabetic conditions may contribute retinal vascular damage and increased permeability in diabetic retina, we used IgM-KO mice. As discussed above, although IgM-KO develop STZ-induced diabetes and have increased level of exosomes in circulation, they lack IgG and classical complement activation by the exosomes. As expected, C57BL/6J mice had significantly increased retinal vascular leakage after 7 weeks of diabetes, however this increase in retinal vascular leakage was abolished in diabetic IgM-KO mice despite the same degree of hyperglycemia. In conclusion, this study demonstrates that diabetic plasma exosomes are a part of the circulating microenvironment responsible for the activation of the complement cascade, which in turn leads to the upregulation of pro-inflammatory cytokines and chemokines, resulting in vascular damage. Inhibition of exosome-induced complement pathway activation could prevent the advancement of DR pathogenesis. 2.5 RESEARCH DESIGN AND METHODS 2.5.1 Animal model All animal procedures were in compliance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and were approved and monitored by IACUC at Michigan State University. Eight to twelve weeks-old male C57Bl/6J mice and Ighmtm1cgn/J (002288) were purchased from Jackson Laboratory and diabetes was induced with daily intraperitoneal doses (55mg/kg) of streptozotocin (STZ) (Sigma Aldrich) in 100mM citric acid (pH=4.5) for five consecutive days (152). Body weight and blood glucose concentration were monitored biweekly 57 and blood glucose concentration was maintained in the 20 mmol/L range. Mice with seven to twelve weeks since onset of diabetes were used in our studies. 2.5.2 Retinal vascular permeability Retinal vascular permeability analysis was performed according to previously published procedure (153). 2.5.3 Western Blot Western Blot analysis was performed as previously described (154), with the following antibodies at dilution 1:1000: anti-CD63, anti-CD9, anti-ALIX, and anti-TSG101 (SBI, Cat.NO. EXOAB-CD63A- 1, EXOAB-CD9A-1, EXOAB-ALIX-1 and EXOAB-TSG101-1), anti-C1q (CompTech, Cat. No. A200), anti-C1s (Quidel, Cat. No. A302). IRDye Donkey anti-rabbit, anti-goat and anti-mouse was used as secondary antibodies (Rockland, Cat. No. 611-731-127) (LI-COR, Cat. No.925-32213) (Rockland, Cat. No. 610-731-124). In diabetic mouse experiments, exosomes isolated from equal volume (0.1mL) of control and diabetic plasma were loaded into the gel for comparison. Immuno-reactive bands were visualized by using the Odyssey digital imaging program. ImageJ software was used for blot quantification. 2.5.4 Exosome Isolation Blood was collected from inferior vena cava into EDTA tubes (SARSTEDT microvette-300), separated via centrifugation and either used immediately or aliquoted into 1.5mL micro-tubes and stored in -80°C. ExoQuick system (SBI) was used to purify plasma exosomes according to manufacturer protocol. Ultracentrifugation method for exosome extraction was conducted as previously reported (130). In short, 0.5mL of plasma was mixed with equal volume of PBS, and then centrifuged at 1000g 58 for 10 minutes at 4°C, the supernatant was saved and spun at 10,000g for 30 minutes at 4°C to remove cellular debris. After the supernatant was filtered through a 0.22um filter, exosomes were precipitated by spinning at 100,000g for 2 hours at 4°C in a SORVALL M120SE Micro- Ultracentrifuge (S55S-1155 Rotor, SORVALL). The exosome pellet was re-suspended with PBS, and then was washed by spinning at 100,000g or purified via OptiPrep Density gradient. 2.5.5 OptiPrep Density Gradient Exosome Purification Purification of exosomes by OptiPrep density gradient was done as previously described (155) with slight modifications. Initially, exosome mixture (0.08 mL) was overlaid onto the top of discontinuous gradient, and centrifugation was performed at 100,000g for 18 hours at 4°C with a SORVALL M120SE Micro-Ultracentrifuge (S55S-1155 Rotor, SORVALL). Twelve fractions (0.167mL/fraction) were collected, diluted with 0.833 mL of PBS, centrifuged for 2 hours at 4°C at 100,000g, washed with 1mL PBS, and then re-suspended in 30uL PBS. Exosomes were identified through exosomal markers CD63, CD9, TSG101, and ALIX via western blotting and visualized by electron microscopy. A control OptiPrep gradient containing 0.08mL buffer (0.25M sucrose/10mM Tris, pH at 7.5) was used to determine the density of each fraction (156). 2.5.6 Exosome Quantification To create a standard curve for quantification of plasma exosomes, artificial lipid vesicles were extruded using Avanti Mini-Extruder (Avanti Polar Lipids, Inc.) with 0.1um polycarbonate membrane. These lipid vesicles were created from 68.3% glycerophospholipid (1,2-dioleoyl-sn- glycero-3-phosphocholine (DOPC)), 22.7% sphingolipid (sphingomyelin (SM)), 4.8% glycolipid (1,2-dioleoyl-sn-glycero-3-phosphoglycerol (DOPG)), and 4.3% sterol lipids (cholesterol) as determined previously in order to mimic actual plasma exosome lipid composition and size (88). 59 The actual concentration of extruded lipid vesicles was calculated using vesicle surface area and composition, and a serial dilution was then performed. The extruded lipid vesicles and plasma exosomes were then analyzed with dynamic light scattering (DLS) and static light scattering (SLS) technologies via Zetasizer Nano NZ (Malvern Instruments Ltd, Malvern, UK), to determine both diameter and concentration of vesicles, respectively. 2.5.7 Transmission Electron Microscopy (TEM) TEM imaging of isolated exosomes was achieved as previously described (130). For immuno-gold labeling, immuno-gold goat anti-rat antibody was used (Sigma, G7035). Images were taken with JEOL 2200FS Ultra-High-Resolution Transmission Electron Microscope. 2.5.8 C1q Binding Assay C1q binding Assay was performed based on previously published procedures with modifications (157,158). Isolated human or mouse plasma exosomes (from 0.5 mL plasma) were re-suspended in 100uL HEPES buffer (150mM NaCl, 2mM CaCl2, 20mM HEPES, pH 7.0) and incubated with 2ug of C1q (CompTech. A099) for 30 minutes at 37°C. Complement activator, Heat Aggregated Gamma Globulin (HAGG) (Quidel, A144) was used as a positive control reaction and C1q alone in the buffer was served as a negative control. After the incubation, exosomes were isolated via ultracentrifugation, purified by OptiPrep density gradient, and analyzed by Western blot. 2.5.9 C1 Activation Assay The ability of exosomes to induce C1 activation was measured using previously published in vitro assay with modifications (132,159). Exosomes was isolated from 0.5mL of human or mouse plasma, re-suspended in C1 activation assay buffer (50nM triethanolamine-HCL, 145mM NaCl, 1mM CaCl2, pH 7.4) and incubated with C1 complex (0.25uM) (CompTech, A098) in the presence 60 or absence of C1 inhibitor (INHC1) (CompTech, A140) for 90 minutes at 37°C. After incubation, the reaction mixtures were placed on ice for 10 minutes, submitted to OptiPrep density gradient purification, and then activation of C1s was analyzed by Western blot. Complement activator, Heat Aggregated Gamma Globulin (HAGG) (Quidel, A144) was used as a positive control and C1 alone in the buffer as a negative control. 2.6 Statistical Analysis Student paired t-test was used to analyze data with two groups. For Multiple group comparisons one-way ANOVA with post-hoc analysis by Tukey’s range test (GraphPad Prim 7, GraphPad Software, San Diego, CA) was used. All values are expressed as mean ± Standard Deviation. P- values below 0.05 were considered significant. 61 Chapter 3. Exosome-induced classical complement activation leads to the retinal endothelial cells damage via MAC deposition Authors: Chao Huang1, Kiera P. Fisher1, Sandra S. Hammer1, Julia V. Busik1 3.1 ABSTRACT Several studies suggested that there is a link between MAC deposition in the retina and progression of Diabetic Retinopathy (DR). Our recent investigation demonstrated that circulating IgG-laden exosomes contribute to an increase in retinal vascular permeability in DR through activation of complement system. However, the mechanism through which exosome-induced complement activation contributes to retinal vascular cytolytic damage in DR is not well understood. In this study, we demonstrate that IgG-laden exosomes in rat plasma activate the classical complement pathway, and in vitro STZ-induced rat diabetic plasma results MAC deposition and cytolytic damage in human retinal endothelial cells (HRECs). Moreover, removal of the plasma exosomes reduced MAC deposition and abrogated cytolytic damage in HRECs. Together, the results of this study demonstrate that complement activation by IgG-laden exosomes could lead to MAC deposition and contribute endothelium damage and progression of DR. 3.2 INTRODUCTION Extensive experimental and clinical evidence supports the link between complement system activation and the pathogenesis of diabetic vascular complications, including diabetic retinopathy (DR) and atherosclerosis (3,44).The complement system is an effector for adaptive and innate immunity that is activated via three enzymatic cascades known as the classical, the mannose- 62 binding lectin (MBL) and alternative pathways(41,160). All three pathways eventually converge at the level of complement 3 (C3) and thereafter share a common sequence of C5 convertase formation and generation of membrane attack complex (MAC) (C5b-9). Formation of MAC results from the binding of C5b to complement proteins C6, C7, C8 and multiple molecules of C9. Once formed, MAC complex leads to osmotic imbalance and ultimately lysis of pathogens or cells. To prevent unintended damage to the host tissue by activated complement cascade, several complement regulatory proteins (CD55, CD46 and CD59) are anchored on the plasma membrane via a glycosylphosphatidylinositol. These regulatory proteins protect host cells from complement-induced self-cell damage (41). However, aberrant complement activation and impairment of complement regulatory proteins in pathological conditions, can lead to MAC formation on host cells (41,160). An increased level of MAC deposition found in the eyes of patients with DR when compared to eyes from non-diabetic subjects (56), is suggested to be the result of both the reduced levels of complement regulatory proteins, and continued activation of the alternative pathway (57). Moreover, glycosylation-induced impairment of functional activity of complement regulatory protein CD59 may also contribute to complement activation in pathologies such as diabetes (58). Increased MAC deposition in diabetic tissues is suggested to induce the release of growth factors that promote cell proliferation in the vascular wall of vessels and thus contributes to the development of vascular proliferative disease (161). However, whether complement activation and MAC deposition participate in retinal endothelium damage remains unclear. Intriguingly, complement components have been found to associate with extracellular vesicles, such as exosomes, in circulation (117,162). Exosomes are small extracellular vesicles measuring 63 40-200 nm in diameter (66), present in most of biological fluids including blood, urine, cerebrospinal fluid and ascites (67). Importantly, exosomes released from parental cells carry biological information such as nucleotides, proteins and lipids that exert various effects on target cells. In fact, it has been suggested that exosomes may serve as a novel cell-cell communication method due to the heterogeneity of the cargo they carry from cell to cell (67). Recently, we have reported that plasma exosomes activate the classical complement pathway and may contribute to MAC-induced retinal vascular permeability in diabetes (178). However, whether MAC induces cellular damage in the diabetic retinal vasculature remains unknown. In this study, we investigated the physiological mechanism in which diabetic plasma exosomes induce complement activation that leads to MAC deposition and cellular impairment in HRECs 3.3 RESULTS Immunoglobulins are associated with exosomes in rat plasma. Previously, we have demonstrated that immunoglobulins are associated with exosomes isolated via ExoQucik method in mouse plasma (178). ExoQucik method provides fast and efficient way to isolate exosomes, however, it is not specific for exosomes and can precipitate a wide range of extracellular vesicles and proteins that could potentially affect the results of the study. In this study, sufficient plasma volume from the rat model allows us to precipitate exosomes via ultracentrifugation method, allowing for better purity and more homogeneous exosome isolation. We observed that rat exosomes isolated by ultracentrifugation showed an enrichment of classical exosomal markers (CD63, ALIX, TSG101 and CD9) compared to total plasma (Figure.15A). In agreement with our previous study, immunoglobulins were also enriched in rat plasma exosomes (Figure. 15A) (178). Purity of the isolation was further confirmed by a 64 reduction of LDL lipoproteins (ApoE) and Calnexin markers (Figure. 15A, bottom) in the exosomal preparation. OptiPrep density gradient was used post sequential ultracentrifugation to demonstrate that the presence of IgG in exosomal fractions. We found that in fraction 6 to 10 were positive for exosomal markers (CD63 and ALIX), where IgG were also presence (Figure. 15B). After depleting exosomes from plasma, by repeating ultracentrifugation isolations, we found an associated reduction of IgG level in the rat plasma (Figure. 15C, middle lane). These data demonstrate that in rat plasma, immunoglobulins are associated with exosomes. Figure 15. Characterization and specificity of immunoglobulin binding to exosomes. (A) Western blot analysis of rat plasma and ultracentrifugation isolated exosomes (25ug protein loaded), isolated rat plasma exosomes showed an enrichment of exosomal markers (CD63, ALIX TSG101, CD9) compared to whole plasma. Specificity of exosome isolation is confirmed using lipoprotein marker (ApoE) and ER marker Calnexin. (B) Western blot analysis of exosomes with OptiPrep density gradient separation of exosomes post-ultracentrifugation isolation. Exosomal markers CD63, ALIX and CD9 as well as Immunoglobulin are present in fractions with 1.10-1.39 g/ml density (fractions 6 to 10). (C) Immunoglobulin is removed in exosome-depleted plasma (right lane) via combined ExoQuick and ultracentrifugation methods. Figure 1 OptiPrep Gradients 7 1.12 9 1.24 8 1.16 A Exosome Plasma IgG CD63 ALIX TSG101 CD9 ApoE Calnexin B Fractions # Density (g/mL) CD63 5 1.08 6 1.10 ALIX CD9 IgG 50kDa 25kDa 54kDa 95kDa 52kDa 24kDa 36kDa 75kDa 10 1.39 11 1.42 C Exosome depleted plasma Plasma Exosome 54kDa 95kDa 24kDa 50kDa ALIX TSG101 CD9 IgG 25kDa 95kDa 52kDa 24kDa 50kDa 25kDa Figure 1. Characterization and specificity of immunoglobulin binding to exosomes. (A) Western blot analysis of rat plasma and ultracentrifugation isolated exosomes with 25ug protein loading volume, isolated rat plasma exosomes showed an enrichment of exosomal markers (CD63, ALIX TSG101, CD9) compared to whole plasma. Specificity of exosome isolation is confirmed using lipoprotein marker (ApoE) and ER marker Calnexin. (B) Western blot analysis of exosomes with OptiPrep density gradient separation of exosomes post-ultracentrifugation isolation. Exosomal markers CD63, ALIX and CD9 as well as Immunoglobulin are present in fractions with 1.10-1.39 g/ml density (fractions 6 to 10). (C) Immunoglobulin is removed in exosome-depleted plasma (right lane) via combined ExoQuick and ultracentrifugation methods. 65 IgG laden exosomes activate the classical complement protein C1 complex. Immunoglobulin is a potent activator of the classical complement pathway. C1q is the first classical complement protein that associates with immunoglobulin (131). To examine whether IgG-laden rat plasma exosomes activate classical complement pathway, we performed a C1q binding assay (163). Rat plasma exosomes isolated by ultracentrifugation were used in these experiments to reduce the non-exosomal vesicle contamination. Results of the C1q binding assay showed that purified human C1q protein (CompTech, A099) binds with IgG laden rat plasma exosomes (Figure. 16A). OptiPrep density gradient centrifugation post C1q binding assay, demonstrated that human C1q binds to rat plasma exosomes in fraction 9 (Figure. 16B), in the absence of lipoprotein (ApoE) and ER marker (Calnexin) (Figure. 16B, bottom). Once C1q binds to the classical complement activator, auto-activation of serine proteases of C1r occurs. This in turn cleaves and activates another serine protease C1s in the C1 complex (44). To investigate if C1q binding to exosomes activates the C1 complex, we performed the C1 activation assay (159). Rat plasma exosomes isolated by ultracentrifugation were incubated with purified human C1 complex (CompTech, A098) and activation of C1 was measured by a cleaved form of C1s. Rat plasma exosomes activate C1 (Figure. 16C, middle lane), and C1 activity was reduced in the presence of human C1 inhibitor (C1-INH) (CompTech, A140) (Figure. 16C, right lane). OptiPrep density gradient showed that C1 activation occurred in the same fraction (fraction 9) that was also positive for exosomal markers (CD63 and ALIX) (Figure. 16D, left lane). These results suggest that IgG laden rat exosomes bind to C1q and activate the classical complement protein, C1, in plasma. 66 Figure 16. Western Blot analysis of complement activation in rat plasma exosomes isolated via ultracentrifugation and purified by OptiPrep density gradient. (A) Binding of human complement protein, C1q, to ultracentrifugation isolated rat exosomes. (B) C1q specifically bound to rat plasma exosomes with density at 1.24 g/mL (fraction 9). This fraction was negative for ApoE and Calnexin. (C) C1 activation assay showed that C1 was activated by ultracentrifugation-isolated rat plasma exosomes (middle-lane) and C1 activity was inhibited in the presence of C1 inhibitor (INHC1) (right-lane). (D) Incubating C1 complex with rat plasma exosomes followed by fractionation using OptiPrep density gradient showed that C1 complex bound to exosomes of 1.24g/mL density (fraction 9) and activated serine protease subcomponent C1s (left), and the activity of the C1s was inhibited by C1 inhibitor (INHC1) (right). C Exosomes Exosomes C1 Exosomes C1 + INHC1 ALIX CD63 C1S Activated Activated D Exosomes C1 8 9 1.24 Fractions 7 Density(g/mL) 1.12 1.16 C1S Activated Activated CD63 ALIX 30kDa 54kDa 86kDa 50kDa 28kDa Exosomes C1+INHC1 7 8 1.12 1.16 9 1.24 86kDa 58kDa 28kDa 54kDa 95kDa A Exosomes Exosomes C1q C1q CD63 IgG 30kDa 54kDa 50kDa Exosomes + C1q 7 1.12 8 1.16 9 1.24 30kDa 54kDa 50kDa 36kDa 75kDa B Fractions Density(g/mL) C1q CD63 IgG ApoE Calnexin Quantification of exosomes in rat control and diabetic artery plasma. Previously, we have demonstrated a new exosome quantification method by using combined dynamic and static light scattering (DLS and SLS) assay using Zetasizer Nano NZ (Malvern Instruments Ltd, Malvern, UK) (178). To reduce the technical variability caused by exosome isolation procedure, we used 67 ExoQuick exosome isolation method instead of ultracentrifugation. By using this exosome quantification method, we found that STZ-induced (7 weeks) diabetic rats have a higher number of extracellular vesicles in the artery plasma when compared to non-diabetic rats (Figure. 17A). This is consistent with our previous observation using plasma from diabetic mice (178). Interestingly, the diameter of the exosomes also increased in diabetic artery blood when compared to non-diabetic controls (Figure. 17B). These data suggest that diabetes not only increases the quantity of circulating arterial exosomes, but it also causes changes the physical properties of exosomes, as measured by calculating exosome diameter. Figure 17. Comparison of rat artery plasma exosomes isolated via ExoQuick. (A) The number of exosomes isolated from control (left) and STZ-induced diabetic rat plasma (7 weeks on set of disease) (right) was quantified by using SLS (kcps) reading combined with a standard curved generated with extruded vesicles (178). Number of plasma exosomes was increased significantly in diabetic rat plasma than in control. (B) DLS measurement of rat plasma exosomes showed an increase of diameter in diabetic exosomes (right) than in control (left). n = 10 for control and diabetes, *P < 0.05. Figure 3 A Rat Tail Plasma Exosomes Rat Artery Plasma Exosomes B Rat Artery Plasma Exosomes Rat Tail Plasma Exosomes l i / L m s e c s e v f o r e b m u N 6×1011 4×1011 2×1011 0 ** Control Diabetes ) m n ( r e t e m a D i 300 200 100 0 ** Control Diabetes Diabetic rat plasma exosomes contribute to HREC cytotoxicity via MAC deposition. During the Figure 3. Comparison of rat artery plasma exosomes isolated via ExoQuick. (A) The number of exosomes isolated from control (left) and STZ-induced diabetic rat plasma (7 weeks on set of disease )(right) was quantified by using SLS (kcps) reading combined with a standard curved generated with extruded vesicles (ref). Number of plasma exosomes was increased significantly in diabetic rat plasma than in control. (B) DLS measurement of rat plasma exosomes showed an increased of diameter in diabetic exosomes (right) than in control (left). n=10 for control and diabetes, *p<0.05. terminal complement cascade, complement proteins C5b, C6, C7, C8 are assembled on the cell membrane into C5b-8 complex that binds C9 proteins and intercalates into the plasma membrane, creating the MAC. To examine if plasma exosome-activated complement cascade 68 leads to MAC formation and cell lysis, we treated HREC with 20% of control or diabetic rat plasma in the presence or absence of exosomes. Cytotoxicity was measured by LDH assay, Trypan Blue Exclusion Assay, and MAC formation was examined by using immunocytochemistry. STZ-induced diabetic (7 weeks) and control rat tail-artery blood was collected into EDTA coated tubes, and plasma was isolated by centrifugation. To remove plasma exosomes, we combined ExoQuick with ultracentrifugation methods to maximize the exosome depletion. LDH Assay was used to quantify the cytotoxicity of the conditioned media. LDH assay showed a significant increase in HRECs cytotoxicity in the cells treated with diabetic rat plasma compared to the cells treated with control rat plasma (Figure. 18A). Removal of exosomes from the diabetic rat plasma significantly reduced cytotoxicity (Figure. 18A). Interestingly, addition of diabetic exosomes back into the exosome-removed plasma, resulted in increased cytotoxicity when compared with diabetic exosome-removed condition (Figure. 18A). Trypan blue assay showed similar results with substantial increase of cell death in HRECs treated with diabetic rat plasma compared to control plasma treated cells (Figure. 18B). This phenotype was reversed by removal of exosomes from diabetic plasma (Figure. 18B). To investigate whether complement terminal cascade was involved in the cytotoxicity, rat plasma treated HRECs were stained with anti-MAC antibody (AbCam, ab55811). A large number of HRECs treated with diabetic rat plasma detached from the culture plates within 6 hours after treatment. In 3 hours, remaining diabetic plasma treated HRECs became round, granular and stained positive with MAC (Figure. 18D). In contrast, when HRECs were treated with diabetic exosome-removed plasma, cells remained viable, attached and negative on MAC staining (Figure. 18F). Moreover, when diabetic exosomes were added back into the diabetic exosome-removed plasma condition, MAC deposition was observed in HRECs, 69 but this deposition was less than in the diabetic rat plasma condition (Figure. 18H). On the other hand, HRECs treated with rat control plasma remained viable, and showed negative MAC staining (Figure. 18 C, E and G). Overall, these data suggest that diabetic plasma exosomes contribute to MAC-deposition and cytolytic damage in HRECs. Figure 18. Diabetic rat plasma induced cytotoxicity and cell death in Human Retinal Endothelial Cells (HRECs). STZ induced diabetic (7 weeks) rat plasma was collected from tail artery and exosomes were removed via ExoQuick combined with Ultracentrifugation method. (A) LDH Assay; after 6 hours, diabetic rat plasma (white bar) treated HRECs showed a significant increase of cytotoxicity comparing to control plasma treated HRECs (grey bar). When exosomes were removed from diabetic plasma (white bar), cytotoxicity was reduced compared with diabetic plasma treated HRECs (white bar). After added diabetic exosomes back into the exosome- removed plasma condition, cytotoxicity showed a significantly increase when compared with diabetic exosome removed condition (white bars). (B)Trypan Blue staining of detached HRECs showed similar results with LDH measurement. *, # P < 0.05, ** P < 0.01. (C-H) HRECs were treated with rat plasma in different conditions for 3 hours, and rat MAC (red) deposition and cellular nuclei (DAPI) were measured by immunocytochemistry. Diabetic rat plasma induced MAC deposition in Human Retinal Endothelial Cells (HRECs)(D). MAC deposition was greatly reduced in HRECs treated with exosome removed diabetic rat plasma (F). Once diabetic exosomes were added back into the plasma, MAC deposition was observed (H), while all the controls remain unchanged (C, E and G). 70 MAC deposition in the retinal vascular of diabetic rats. In 7 weeks after STZ-induced diabetic rats, immuno-stained rat retina showed an increase of MAC deposition when compared to controls (Figure. 19). These experiments demonstrated that MAC deposition occurred in the eyes of diabetic rats. Figure 19. MAC in the retina of diabetic rats. Retinal sections from a control rat (A) showed less MAC immunostaining (red fluorescence) than in a STZ-induced diabetic rat (7 weeks) (B). (C) MAC negative staining of the diabetic rat retina section. DAPI (blue) was used to stain for cell nuclei. Figure 5 Control Diabetes Negative Control A B C MAC DAPI MAC DAPI MAC DAPI Figure 5. MAC in the retina of rats with 7-weeks duration of STZ-diabetes. Retinal sections from a control rat (A) showed less MAC immunostaining (red fluorescence) than in a diabetic rat (B). (C) MAC negative staining of the diabetic rat retina section. DAPI staining in blue fluorescence. 3.4 DISCUSSION Under normal physiological conditions, low level complement activation and non-cytolytic MAC deposition are thought to be beneficial, serving as a mechanism to remove opsonized cellular debris and pathogens (164), as well as maintaining the eyes immune privilege. However, under diseased conditions such as the ones present in DR, continued complement activation and reduced level of complement regulatory proteins are suggested to be associated with increased MAC formation in the retinal endothelium lumen (57). Moreover, irreversible MAC deposition on endothelial cells leads to release of mitogenic factors (165), further supporting the rationale that complement activation is involved in the advancement of DR. Furthermore, circulating plasma 71 exosomes have been shown to play a role in complement pathway activation leading to retinal vascular damage in vivo (178). No studies, however, have investigated the role that diabetic exosomes play in causing cytolytic damage of endothelium in vitro. In the current study, we demonstrate that diabetic exosomes induce complement activation and contribute to MAC deposition, which induces retinal endothelial cell damage. The results of this study suggest that exosome-induced complement activation followed by MAC deposition could provide a novel mechanism for increased retinal vascular permeability and DR pathogenesis. Previously, we have demonstrated that immunoglobulins are associated with exosomes in circulation and activate classical complement protein, C1 in both human and mouse model (178). In the current study, we demonstrated similar findings using rat plasma exosomes isolated by ultracentrifugation. These data suggest that activation of classical complement by IgG-laden exosomes occurs across species and is not dependent on the isolation method. Consistent with our previous finding, C1 activation occurs in specific exosomal fraction further supports the rationale that a sub-population of exosome may favors complement activation. As expected, rat plasma exosomes isolated by ultracentrifugation had reduced level of non-exosomal vesicles such as lipoproteins compared exosomes isolated by ExoQuick method. This once again demonstrated that ultracentrifugation is the gold standard for exosome isolation (65) and it is ideal for exosome characterization. However, ultracentrifugation method is not quantitative due to inherent variability of this multi-step process and is thus not suitable for comparing samples between control and diabetic groups. ExoQuick isolation method therefore, was the method of choice to compare the number of vesicles between control and diabetic exosomes. 72 We have recently reported, using a novel exosome quantification method in mice, that diabetes causes an increase in the level of extracellular vesicles found in venous blood (178). In agreement to this finding, our present study demonstrates that the level of extracellular vesicles is also increased in arterial blood of STZ-induced diabetic rat plasma when compared to controls. However, the diameter of the rat extracellular vesicles varies within different parts of the circulation. We speculate that exosome cargo may change as they pass from arterial to venous vascular beds. Moreover, other non-exosomal vesicles such as lipoproteins may be present in the ExoQuick isolated plasma exosomes(129), accounting for the differences in diameter since lipoproteins are enriched in blood, and have overlapping diameter with exosomes. Furthermore, we found that isolated rat plasma exosomes show a higher value of polydispersity when compared to mouse exosomes (data not shown). This suggests that rat plasma exosomes isolated by ExoQuick may contain various size of vesicles such as lipoproteins, which could contribute to the vesicle difference between control and diabetic condition. In physiological states, MAC depositions on the surface of cells are rapidly removed. Endothelial cells are targeted continuously by activated complement cascade, and MAC deposition is rapidly eliminated via endocytosis, mitigating the cells from cytolytic destruction(166). A similar mechanism was reported to also occur in retinal pigmented epithelial cells (RPEs) (167). Moreover, it is suggested that, in vitro, MAC-induced mitogenesis contributes to focal tissue repair or pathological cell proliferation (168). Other groups have shown that in diabetic eyes, extensive MAC deposition in the retinal vascular endothelium leads to the embedment of plasma membrane bearing MAC in adjacent basement membrane without causing cytolytic damage (57). These reports suggest a different finding than the one presented in this study since here we show 73 evidence suggesting that diabetic rat plasma induces robust MAC deposition and cytolytic damage in HRECs. These deleterious changes are not seen in the presence of control plasma. We speculate that in the early stage of the DR, increased non-cytolytic retinal vascular MAC deposition contributes to signal focal tissue repair. As the disease advances, retinal tissue repair process becomes deficient (169), and accumulation of MAC deposition may contribute to cytolytic damage and increased retinal permeability. Previously, Kim D et al. suggested that complement regulator proteins are homologously restricted and are less effective on complement targets from different species (51). Moreover, gal-(alpha 1-3)-gal epitopes on endothelial cells have a high affinity to natural antibodies from different species and favor classical complement-induced xeno-organ rejection(170). These reports suggest that in our study, surface expressed complement regulatory proteins on HRECs might be less effective in protecting cells from rat exosome-induced complement activation. Furthermore, HRECs might be sensitized by rat exosomes associated- and/or exosome un-associated- immunoglobulins, which favor classical complement activation. Additionally, in diabetic plasma, hyperglycemia is usually accompanied with an elevated level of inflammatory cytokines, and chemokines, this pro- inflammatory environment coupled with exosomes ability to carry proteins, nucleotides and lipids makes exosomes likely to facilitate cellular damage. Thus, we speculate that in diabetic rat plasma, there are other factors, besides exosome-induced complement deposition that may contribute to HRECs damage. Interestingly, when we deplete exosomes from the diabetic plasma, complement-induced MAC deposition and cytolysis of HREC was prevented. This suggests that plasma exosomes, in part, contribute to complement dependent retinal cellular injury. However, addition of diabetic plasma exosomes back into the diabetic exosome-free plasma environment, 74 the cytotoxic effect was not fully present, suggesting that exosome isolation methods employed in our study may have inhibited proteins that are important for cellular cytotoxicity. These results highlight the importance of continued investigation and development of exosome isolation techniques. In summary, this study demonstrates that diabetic rat plasma exosomes activate classical complement protein C1. Activation of the complement cascade in turn contributes to MAC deposition and cytolytic damage of the retinal endothelial cell. This finding may provide a novel mechanism of endothelial cell dysfunction and advancement of DR pathogenesis. 3.5 METHODS 3.5.1 Animal Studies All animal procedures were in compliance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals, and with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. Diabetes was induced by intraperitoneal injection with a single dose of streptozotocin (STZ) (65 mg/kg) in Sprague-Dawley rats (171) and the animals were maintained as previously described (178). 3.5.2 Cell Culture Primary human retinal endothelial cells (HRECs) were prepared from postmortem tissue obtained from National Disease Research Interchange, Philadelphia, PA, and EverSight Midwest Eye-Banks, Ann Arbor, MI. Primary HREC were isolated and cultured as previously described (172). Passages 3-6 were used in the experiments at 80-90% confluence in 10% FBS media before treatments. 75 3.5.3 Western Blot Western Blot analysis was performed as previously described, with the following antibodies at dilution 1:1000: anti-CD63, anti-CD9 and anti-TSG101 (SBI, Cat.NO.EXOAB-CD63A-1, EXOAB- CD9A-1 and EXOAB-TSG101-1), anti-ALIX (AbCam, Cat. No. ab117600), anti-C1q (CompTech, Cat. No. A200), anti-C1s (Quidel, Cat. No. A302). IRDye Donkey anti-rabbit or anti-goat was used as secondary antibodies (Rockland, Cat. No. 611-731-127) (LI-COR, Cat. No.925-32213). Immuno- reactive bands were visualized by using the Odyssey digital imaging program. ImageJ software was used for blot quantification. 3.5.4 Blood Sample Collection Blood was collected from inferior vena cava or from tail artery of the animal into tubes with EDTA (SARSTEDT microvette-300). Plasma was harvested via centrifugation and either used immediately or aliquoted into 1.5mL micro-tubes and stored in -80°C. 3.5.5 Exosome Isolation Sequential ultracentrifugation method for exosome extraction was conducted as previously reported(130). In short, 0.5mL of rat plasma was mixed with equal volume of PBS, and then a series of low speed centrifugations of the supernatant were conducted to remove cellular debris. After the supernatant was filtered via a 0.22um filter, exosomes were precipitated by spinning at 100,000g for 2 hours at 4°C in a SORVALL M120SE Micro-Ultracentrifuge (S55S-1155 Rotor, SORVALL). The exosome pellet was washed by re-suspension with PBS and spun down at 100,000g before further analysis. 76 3.5.6 Exosome Quantification Exosome quantification was conducted based on previously reported procedures by combining dynamic light scattering (DLS) and static light scattering (SLS) technologies via Zetasizer Nano NZ (Malvern Instruments Ltd, Malvern, UK) )(178). 3.5.7 OptiPrep Density Gradient Exosome Purification Discontinuous iodixanol gradient was used to further purify the exosome solution. Purification of exosome by OptiPrep density gradient was done as previously described (15). 3.5.8 C1q Binding Assay C1q binding Assay was performed based on previously published procedures with modifications (157,158). Isolated rat plasma exosomes (from 1mL plasma) were re-suspended in 100uL HEPES buffer (150mM NaCl, 2mM CaCl2, 20mM HEPES, pH 7.0) and incubated with 2ug of C1q (CompTech. A099) for 30 minutes at 37°C. After the incubation, exosomes were either isolated via ultracentrifugation, purified by OptiPrep density gradient and analyzed by Western blot. 3.5.9 C1 Activation Assay The ability of exosomes to induce C1 activation was measured using an in vitro assay as previously published with modifications (132,159). Exosomes were isolated from 1mL of rat plasma, re- suspended in C1 activation assay buffer (50nM triethanolamine-HCL, 145mM NaCl, 1mM CaCl2, pH 7.4) and incubated with C1 complex (0.25uM) (CompTech, A098) in the presence or absence of C1 inhibitor (INHC1) (CompTech, A140) for 90 minutes at 37°C. The reaction mixtures were incubated on ice for 10 minutes, submitted to OptiPrep density gradient purification, and activation of C1s was analyzed by Western blot. 77 3.6 LDH Assay Cytotoxicity of the cells was quantified using LDH assay kit following the manufacturer’s protocol (AbCam, ab102526). Briefly, 10^5 HRECs/100uL were plated onto a 0.1% gelatin coated 96-well plate, incubated at 37°C for 48 hours, and then treated with 20% of control or diabetic rat plasma at 37°C for 6 hours. In addition, to determine the contribution of exosomes vs. other plasma components exosome-removed control or diabetic plasma, and exosome- removed control or diabetic plasma with control or diabetic exosomes added back in was used. After the treatment, the media was collected into a 1.5 mL Eppendorf tube and spun down at 10,000g for 15 minutes at 4°C. Supernatants (50uL) were transferred into a 96 well plate and a mixed detection kit reagent (50uL) was added to each well alone with the NADH standard. The absorbance (450nm) was taken on a micro-plate reader in a kinetic mode every 3 minutes for 30 minutes at 37°C. Cytotoxicity = [(Experimental LDH)-(Negative control LDH)]/[(Positive control LDH)-(Negative control LDH)], LDH concentration from no cell conditions (background) was subtracted from each condition before the calculation of cytotoxicity. Experiments were performed in triplicates following the assay manufacturer’s recommendation in duplicates, and results are presented as mean ± SD. Control plasma treated HRECs was used for normalization to other conditions in LDH. 3.7 Immunocytochemistry Immunocytochemistry was performed as previously described using anti-C5b-9 antibody (AbCam, ab55811) at 1:100 in PBS with 1.5% BSA overnight at 4°C, followed by chicken anti-rabbit Alexa Fluor 594 secondary antibody (1:500) and DAPI (Sigma-Aldrich, St.Louis.MO) nuclei counterstaining. The slides were analyzed using Nikon TE2000 microscope equipped with 78 Photometric Cool-SNAP HQ2 camera. All images were taken with matched exposure time for experimental and control slides by using the MetaMorph imaging software (Molecular Devices, Downington, PA). 3.8 Statistics A Student paired t-test was used to analyze data with two groups. In experiments with multiple group comparisons one-way ANOVA with post-hoc analysis by Tukey’s range test (GraphPad Prim 7, GraphPad Software, San Diego, CA) was used. All values are expressed as mean ± Standard Deviation. P-values below 0.05 were considered significant. 79 Chapter 4. Summary and Future Perspectives Summary: In summary, the studies presented in this dissertation demonstrate for the first time that circulating extracellular vesicles such as exosomes, cargo immunoglobulins in circulation and activate the complement cascade. More importantly, this work aims to fill the knowledge gap that currently exists between the complement activation and DR progression. Furthermore, we explored a novel mechanism in which extracellular vesicles participate in retinal vascular damage in DR. First, I explored the role exosome-induced complement activation in diabetic retinal vascular damage in vivo. I demonstrated that immunoglobulins are associated with circulating mouse and human exosomes in plasma. After quantifying exosomes found in plasma, I found that there is an increased number of extracellular vesicles in the STZ-induced diabetic mouse model when compared to non-diabetic mice. IgG-laden plasma exosomes activate classical complement protein C1 via binding with its sub-component C1q. Furthermore, using an immunoglobulin deficient mouse model, reduced levels of exosome-induced classical complement activation were demonstrated when compared to wildtype mouse (C57bl/6J strain). These immunoglobulin deficient mice also display reduced levels of retinal vascular permeability. Next, I investigated the mechanism of retinal vascular damage by diabetic exosome-induced complement activation in vitro. In these series of experiments, we showed that immunoglobulins are also associated with plasma exosomes in STZ-induced diabetic rat model. Rat diabetic artery plasma induced MAC deposition and led to cytolytic damage in HRECs. More interestingly, removal of exosomes from the diabetic plasma significantly reduced MAC deposition and cell 80 death. Partial restoration of MAC deposition on HRECs can be achieved by adding back of diabetic exosomes in HRECs. Taken together, these findings strongly suggest that circulating exosomes activate the complement cascade, which contributes to the development of DR. Notably, this work also highlights the potential use of exosomes as biomarkers for DR prognosis and therapeutic strategy. However, further investigations would be beneficial to have a better understanding of the difference between diabetic and non-diabetic plasma exosomes. Future Perspectives: The following experiments are suggested for future study: 1. The role of other complement components associated with plasma exosomes. We have demonstrated that IgG laden plasma exosomes activate classical complement protein C1 and participate in MAC deposition on the retinal endothelium. However, there are other downstream complement components after C1 activation and are equally important in initiating the MAC deposition. Hence, it is of interest to investigate after exosome-induced C1 activation, what other downstream complement proteins are associated with exosomes. 2. Complement-induced anaphylatoxin secretion leads to inflammation Anaphylatoxin such as C3a and C5a are the byproducts released during complement activation and are pro-inflammatory. Several studies have demonstrated that C3a and C5a bind to their receptors on the cells and induce cytokine production such as IL-8, IL1b and RANTES (173,174). Hence, it is of interest to investigate whether retinal endothelial cells respond to anaphylatoxin-induced inflammatory activation. 81 3. Exosomes as a biomarker for DR progression. Exosomes secreted from cells contain large amount of biological information, which includes nucleotides, proteins and lipids. Exosome cargo changes based on secreted cells in response to normal and/or pathological conditions. Ubiquitous presence in most biological fluids makes the exosomes an ideal, noninvasive biomarker for prediction and diagnosis of asymptomatic diseases such as DR in the early stage. Previously reported observations confirmed that early treatment can greatly reduce an individual’s risk of severe visual loss by 57% (175). Therefore, a detailed study of nucleotide, proteomic and lipidomic content of the plasma exosomes in different stages of the DR could provide an in-depth profile for diagnosing DR patients. 4. Origin of the IgG laden plasma exosomes. In current study (Chapter 2), we suggested that B-cells may contribute to release of the IgG laden exosomes. Therefore, B-cell deficient (IgM-KO) mouse provides a good model to investigate the source of the IgG laden exosomes by comparing plasma exosome difference between C57bl/6J with IgM-KO mouse. Potential clinical applications: Exosomes are now recognized as a vehicle that package and deliver cellular information between cells. In addition, lipid membrane of the exosome protects the exosomal cargos from degradation. Moreover, molecules present on the exosome membrane allow specific uptake by the target cells (67). In recent years, there has been intense interest in engineering exosomes as a biological nano-platforms for drug delivery. For instance, when paclitaxel(PTX) is packaged in exosomes, it is more effective at inhibiting tumor growth in a murine lung carcinoma model than 82 direct introduction of PTX by itself (176). Despite multiple ongoing clinical trials to investigate the potential of extracellular vesicles/exosomes-based therapies, there are limitations of exosome as therapeutic agents. These includes an incomplete understanding of the role of exosomes in physiological as well as in pathological states and how they cross biological barriers, the absence of methods for the isolation of homogeneous exosome populations and sub-populations, and a need for drug loading methods(177). Hence, additional studies are needed to fully understand the biology of the exosome. In total, our study demonstrated for the first time a novel mechanism connecting complement deposition and the DR progression via IgG laden plasma exosomes. This study provides novel insights for future efforts to develop innovative and effective therapies for early assessment and treatment of DR. 83 REFERENCES 84 REFERENCES Fong DS, Aiello L, Gardner TW, King GL, Blankenship G, Cavallerano JD, et al. Retinopathy in Diabetes. Diabetes Care. 2004;27(SUPPL. 1). Rübsam A, Parikh S, Fort P. Role of Inflammation in Diabetic Retinopathy. Int J Mol Sci [Internet]. 2018;19(4):942. Available from: http://www.mdpi.com/1422-0067/19/4/942 Clark SJ, Bishop PN. The eye as a complement dysregulation hotspot. Seminars in Immunopathology. Seminars in Immunopathology; 2018. p. 65–74. 4. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006;3(11):2011–30. 5. Centers for Disease Control and Prevention. National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States. US Dep. Health Human Serv. 2017;(Cdc):2009–12. 1. 2. 3. 6. 7. 8. 9. Spanakis EK, Golden SH. Race/Ethnic Difference in Diabetes and Diabetic Complications. Curr Diab Rep. 2014;13(6):1–18. Das A, Stroud S, Rangasamy S, Mehta A. New treatments for diabetic retinopathy. Diabetes, Obes Andm. 2015;17(3):219–30. Eye T, Prevalence D. The prevalence of diabetic retinopathy among adults in the United States. Arch … [Internet]. 2004 [cited 2014 Oct 13];122. Available from: http://archopht.ama-assn.org/cgi/content/abstract/122/4/552 Klaassen I, Noorden CJF Van, Schlingemann RO. Progress in Retinal and Eye Research Molecular basis of the inner blood-retinal barrier and its breakdown in diabetic macular edema and other pathological conditions. Prog Retin Eye Res [Internet]. Elsevier Ltd; 2013;34:19–48. Available from: http://dx.doi.org/10.1016/j.preteyeres.2013.02.001 Funk RHW. Blood Supply of the Retina. Ophthalmic Res. 1997;29:320–5. 10. 11. Curtis TM, Gardiner TA, Stitt AW. Microvascular lesions of diabetic retinopathy: Clues towards understanding pathogenesis? Eye [Internet]. Nature Publishing Group; 2009;23(7):1496–508. Available from: http://dx.doi.org/10.1038/eye.2009.108 12. Antonetti D a, Klein R, Gardner TW. Diabetic retinopathy. N Engl J Med [Internet]. 2012 Mar 29;366(13):1227–39. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22455417 85 13. Kern TS. Contributions of inflammatory processes to the development of the early stages of diabetic retinopathy. Exp Diabetes Res. 2007;2007:95103. 14. Ding J, Wong TY. Current epidemiology of diabetic retinopathy and diabetic macular edema. Curr Diab Rep. 2012;12(4):346–54. 15. Zhang W, Liu H, Al-Shabrawey M, Caldwell RW, Caldwell RB. Inflammation and diabetic retinal microvascular complications. J Cardiovasc Dis Res [Internet]. 2011 Apr [cited 2014 Sep 7];2(2):96–103. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3144626&tool=pmcentrez& rendertype=abstract 16. Davidson JA, Ciulla TA, McGill JB, Kles KA, Anderson PW. How the diabetic eye loses vision. Endocrine. 2007;32(1):107–16. 17. White NH, Sun W, Cleary PA, Danis RP, Davis MD, Hainsworth DP, et al. Prolonged effect of intensive therapy on the risk of retinopathy complications in patients with type 1 diabetes mellitus: 10 years after the Diabetes Control and Complications Trial. Arch Ophthalmol [Internet]. 2008;126(12):1707–15. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&A N=19064853%5Cnhttp://lshtmsfx.hosted.exlibrisgroup.com/lshtm?sid=OVID:medline&id =pmid:19064853&id=doi:10.1001%2Farchopht.126.12.1707&issn=0003- 9950&isbn=&volume=126&issue=12&spage= 18. Holman RR, Paul SK, Bethel AA, Matthews DR, Neil AA. 10-Year Follow-Up of Intensive Glucose Control in Type 2 Diabetes. N Engl J Med [Internet]. 2008;359(15):1577–89. Available from: http://content.nejm.org/cgi/content/abstract/359/15/1577 19. The Action to Control Cardiovascular Risk in Diabetes Follow-On (ACCORDION) Eye Study Group. Persistent Effects of Intensive Glycemic Control on Retinopathy in Type 2 Diabetes in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Follow-On Study. Diabetes Care [Internet]. 2016;39(July):1089–100. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27289122 20. Boussageon R, Bejan-Angoulvant T, Saadatian-Elahi M, Lafont S, Bergeonneau C, Kassai B, et al. Effect of intensive glucose lowering treatment on all cause mortality, cardiovascular death, and microvascular events in type 2 diabetes: meta-analysis of randomised controlled trials. Bmj [Internet]. 2011;343(jul26 1):d4169–d4169. Available from: http://www.bmj.com/cgi/doi/10.1136/bmj.d4169 21. Deschler EK, Sun JK, Silva PS. Side-effects and complications of laser treatment in diabetic retinal disease. Semin Ophthalmol. 2014;29(5–6):290–300. 86 22. Tang J, Kern TS. Inflammation in diabetic retinopathy. Prog Retin Eye Res [Internet]. Elsevier Ltd; 2011 Sep [cited 2014 Sep 4];30(5):343–58. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3433044&tool=pmcentrez& rendertype=abstract 23. Osaadon P, Fagan XJ, Lifshitz T, Levy J. A review of anti-VEGF agents for proliferative diabetic retinopathy. Eye (Lond) [Internet]. Nature Publishing Group; 2014;28(5):510–20. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24525867 Scorcia V, Ventrice P, Aloe JF, Bruzzichesi D, Greco E, Leporini C, et al. Anti-vascular endothelial growth factor drugs safety and efficacy in ophthalmic diseases. J Pharmacol Pharmacother [Internet]. 2013;4(5):38. Available from: http://www.jpharmacol.com/text.asp?2013/4/5/38/120947 25. Calle MC, Fernandez ML. Inflammation and type 2 diabetes. Diabetes Metab [Internet]. Elsevier Masson SAS; 2012 Jun [cited 2014 Nov 16];38(3):183–91. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22252015 26. Mohr S, Xi A, Tang J, Kern TS. Caspase activation in retinas of diabetic and galactosemic mice and diabetic patients. Diabetes. 2002;51(4):1172–9. 27. Vincent JA, Mohr S. Inhibition of caspase-1/interleukin-1 beta signaling prevents degeneration of retinal capillaries in diabetes and galactosemia. Diabetes. 2007;56(1):224–30. Joussen AM, Poulaki V, Qin W, Kirchhof B, Mitsiades N, Wiegand SJ, 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 [Internet]. 2002;160(2):501–9. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0002944010648699 29. Antonetti DA, Barber AJ, Hollinger LA, Wolpert EB, Gardner TW. Vascular Endothelial Growth Factor Induces Rapid Phosphorylation of Tight Junction Proteins Occludin and Zonula Occluden 1: A POTENTIAL MECHANISM FOR VASCULAR PERMEABILITY IN DIABETIC RETINOPATHY AND TUMORS. J Biol Chem [Internet]. 1999;274(33):23463–7. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10438525%5Cnhttp://www.jbc.org/cgi/doi/10.10 74/jbc.274.33.23463 30. McLeod DS, Lefer DJ, Merges C, Lutty G a. 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. 24. 28. 87 Lutty G a, Cao J, McLeod DS. Relationship of polymorphonuclear leukocytes to capillary dropout in the human diabetic choroid. Am J Pathol. 1997;151(3):707–14. Joussen AM, Poulaki V, Le ML, Koizumi K, Esser C, Janicki H, et al. A central role for inflammation in the pathogenesis of diabetic retinopathy. FASEB J [Internet]. 2004;18(12):1450–2. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15231732 33. Miyamoto K, Khosrof S, Bursell SE, Rohan R, Murata T, Clermont a C, 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. Stitt AW, Jenkins AJ, Cooper ME. Advanced glycation end products and diabetic complications. Expert Opin Investig Drugs. 2002;11(9):1205–23. Stitt AW. The role of advanced glycation in the pathogenesis of diabetic retinopathy. Exp Mol Pathol. 2003;75(1):95–108. Li G, Tang J, Du Y, Lee CA, Kern TS. Beneficial effects of a novel RAGE inhibitor on early diabetic retinopathy and tactile allodynia. Mol Vis [Internet]. 2011;17(November):3156– 65. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3235538&tool=pmcentrez& rendertype=abstract 37. Walport MJ. Complement. First of two parts. N Engl J Med. 2001;344(14):1058–66. 38. Walport MJ. Complement , second of two parts. N Engl J Med. 2001;344(15):1140–4. 39. Laufer J, Katz Y, Passwell JH. Extrahepatic synthesis of complement proteins in inflammation. Mol Immunol. 2001;38(2–3):221–9. 40. Ghosh P, Sahoo R, Vaidya A, Chorev M, Halperin J a. Role of Complement and Complement Regulatory Proteins in the Complications of Diabetes. Endocr Rev [Internet]. 2015;36(June):er.2014-1099. Available from: http://press.endocrine.org/doi/10.1210/er.2014-1099 41. Merle NS, Church SE, Fremeaux-Bacchi V, Roumenina LT. Complement system part I - molecular mechanisms of activation and regulation. Front Immunol. 2015;6(JUN):1–30. 42. Diebolder C a, Beurskens FJ, de Jong RN, Koning RI, Strumane K, Lindorfer M a, et al. Complement is activated by IgG hexamers assembled at the cell surface. Science [Internet]. 2014;343(6176):1260–3. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24626930 31. 32. 34. 35. 36. 88 43. Nauta AJ, Trouw LA, Daha MR, Tijsma O, Nieuwland R, Schwaeble WJ, et al. Direct binding of C1q to apoptotic cells and cell blebs induces complement activation. Eur J Immunol. 2002;32(6):1726–36. 44. Yanai R, Thanos A, Connor KM. Complement Involvement in Neovascular Ocular Diseases. Current Topics in Innate Immunity II [Internet]. 2012. p. 161–83. Available from: http://link.springer.com/10.1007/978-1-4614-0106-3 45. Bally I, Rossi V, Lunardi T, Thielens NM, Gaboriaud C, Arlaud GJ. Identification of the C1q- binding sites of human C1r and C1s. A refined three-dimensional model of the C1 complex of complement. J Biol Chem. 2009;284(29):19340–8. 46. Brier S, Pflieger D, Le Mignon M, Bally I, Gaboriaud C, Arlaud GJ, et al. Mapping surface accessibility of the C1r/C1s tetramer by chemical modification and mass spectrometry provides new insights into assembly of the human C1 complex. J Biol Chem. 2010;285(42):32251–63. 47. Noris M, Remuzzi G. Overview of complement activation and regulation. Semin Nephrol. 2013;33(6):479–92. 48. Tegla CA, Cudrici C, Patel S, Trippe III R, Rus V, Niculescu F, et al. Membrane Attack By Complement : the Assembly and biology of terminal complement complexes. Immunol Res. 2011;51(1):45–60. 49. Zipfel PF, Skerka C. Complement regulators and inhibitory proteins. Nat Rev Immunol [Internet]. Nature Publishing Group; 2009;9(10):729–40. Available from: http://dx.doi.org/10.1038/nri2620 50. Meri S, Jarva H. Complement Regulatory Proteins. Encycl Life Sci. 2001;(April):1–7. 51. Kim DD, Song WC. Membrane complement regulatory proteins. Clin Immunol. 2006;118(2–3):127–36. Streilein JW. Immunologic privilege of the eye. Springer Semin Immunopathol. 1999;21(2):95–111. 52. 53. Sohn JH, Kaplan HJ, Suk HJ, Bora PS, Bora NS. Chronic low level complement activation within the eye is controlled by intraocular complement regulatory proteins. Investig Ophthalmol Vis Sci. 2000;41(11):3492–502. 54. Bora NS, Gobleman CL, Atkinson JP, Pepose XJS, Kaplan HJ. Differential Expression of the Complement Regulatory Proteins in the Human Eye. Invest Ophthalmol Vis Sci. 1993;34(13):3579–84. 89 55. Gao BB, Chen X, Timothy N, Aiello LP, Feener EP. Characterization of the vitreous proteome in diabetes without diabetic retinopathy and diabetes with proliferative diabetic retinopathy. J Proteome Res. 2008;7(6):2516–25. 56. Gerl V, Bohl J, Pitz S, Stoffelns B. Extensive deposits of complement C3d and C5b-9 in the choriocapillaris of eyes of patients with diabetic retinopathy. Investig Opthalmology Vis Sci [Internet]. 2002 [cited 2015 Jan 10];43(4):1104–8. Available from: http://www.iovs.org/content/43/4/1104.short 57. Zhang J, Gerhardinger C, Lorenzi M. Early Complement Activation and Decreased Levels of Glycosylphosphatidylinositol-Anchored Complement Inhibitors in Human and Experimental Diabetic Retinopathy. Diabetes. 2002;(7):3499–504. 58. Qin X, Goldfine A, Krumrei N, Grubissich L, Acosta J, Chorev M, et al. Glycation Inactivation of the Complement Regulatory Protein Cd59. Diabetes [Internet]. 2004;53:2653–61. Available from: http://elibrary.ru/item.asp?id=8406366 59. Xu D, Yi H, Yu S, Li X, Qiao Y, Deng W. Association of complement C5 gene polymorphisms with proliferative diabetic retinopathy of type 2 diabetes in a Chinese han population. PLoS One. 2016;11(3):1–10. 60. Trams EG, Lauter CJ, Norman Salem J, Heine U. Exfoliation of membrane ecto-enzymes in the form of micro-vesicles. BBA - Biomembr. 1981;645(1):63–70. 61. Pan BT, Johnstone RM. Fate of the transferrin receptor during maturation of sheep reticulocytes in vitro: selective externalization of the receptor. Cell. 1983;33(3):967–78. Johnstone RM, Bianchini A, Teng K. Reticulocyte Maturation and Exosome Release: Transferrin Receptor Containing Exosomes Shows Multiple Plasma Membrane Functions. Blood. 1989;74(5):1844–51. Johnstone RM, Adam M, Hammond JR, Orr L, Turbide C. Vesicle formation during reticulocyte maturation. Association of plasma membrane activities with released vesicles (exosomes). J Biol Chem. 1987;262(19):9412–20. 64. G Raposo, H W Nijman, W Stoorvogel, R Liejendekker, C V Harding, C J Melief and HJG. B Lymphocytes Secrete Antigen-presentingVesicles. J Exp Med. 1996;183(March):1161–72. Li P, Kaslan M, Lee SH, Yao J, Gao Z. Progress in exosome isolation techniques. Theranostics. 2017. p. 789–804. 66. Colombo M, Raposo G, Théry C. Biogenesis, Secretion, and Intercellular Interactions of Exosomes and Other Extracellular Vesicles. Annu Rev Cell Dev Biol [Internet]. 2014;30(1):255–89. Available from: 62. 63. 65. 90 http://www.annualreviews.org/doi/10.1146/annurev-cellbio-101512-122326 67. van Niel G, D’Angelo G RG. Shedding light on the cell biology of Extracellular Vesicles. Nat Rev Mol Cell Biol [Internet]. Nature Publishing Group; 2017;accepted. Available from: http://dx.doi.org/10.1038/nrm.2017.125 68. Hurley JH, Hanson PI. Membrane budding and scission by the ESCRT machinery: it’s all in the neck. Nat Rev Mol Cell Biol [Internet]. 2010 Aug [cited 2014 Jul 9];11(8):556–66. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2922035&tool=pmcentrez& rendertype=abstract 69. Colombo M, Moita C, van Niel G, Kowal J, Vigneron J, Benaroch P, et al. Analysis of ESCRT functions in exosome biogenesis, composition and secretion highlights the heterogeneity of extracellular vesicles. J Cell Sci [Internet]. 2013;126:5553–65. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24105262 70. Trajkovic K, Hsu C, Chiantia S, Rajendran L, Wenzel D, Wieland F, et al. Ceramide triggers budding of exosome vesicles into multivesicular endosomes. Science (80- ) [Internet]. 2008 Feb 29 [cited 2014 Jul 9];319(5867):1244–7. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18309083 71. Busik J V, Esselman WJ, Reid GE. Examining the role of lipid mediators in diabetic retinopathy. Clin Lipidol [Internet]. 2012 Dec 1;7(6):661–75. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3640872&tool=pmcentrez& rendertype=abstract 72. Andreu Z, Yáñez-Mó M. Tetraspanins in extracellular vesicle formation and function. Front Immunol [Internet]. 2014 [cited 2014 Oct 8];5(SEP):1–12. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4165315/ 73. van Niel G, Bergam P, Di Cicco A, Hurbain I, Lo Cicero A, Dingli F, et al. Apolipoprotein E Regulates Amyloid Formation within Endosomes of Pigment Cells. Cell Rep. 2015;13(1):43–51. 74. Chairoungdua A, Smith DL, Pochard P, Hull M, Caplan MJ. Exosome release of β-catenin: A novel mechanism that antagonizes Wnt signaling. J Cell Biol. 2010;190(6):1079–91. 75. Buschow SI, Nolte-’t Hoen ENM, van Niel G, Pols MS, ten Broeke T, Lauwen M, et al. MHC II In dendritic cells is targeted to lysosomes or t cell-induced exosomes via distinct multivesicular body pathways. Traffic. 2009;10(10):1528–42. 76. Kim DK, Lee J, Kim SR, Choi DS, Yoon YJ, Kim JH, et al. EVpedia: A community web portal for extracellular vesicles research. Bioinformatics. 2015;31(6):933–9. 91 77. Kalra H, Simpson RJ, Ji H, Aikawa E, Altevogt P, Askenase P, et al. Vesiclepedia: A Compendium for Extracellular Vesicles with Continuous Community Annotation. PLoS Biol. 2012;10(12):8–12. Simpson RJ, Kalra H, Mathivanan S. ExoCarta as a resource for exosomal research. J Extracell Vesicles. 2012;1(0):1–6. 79. Morelli AE, Larregina AT, Shufesky WJ, Sullivan MLG, Stolz DB, Papworth GD, et al. Endocytosis, intracellular sorting, and processing of exosomes by dendritic cells. Blood. 2004;104(10):3257–66. 80. Material S. Cell surface tetraspanin Tspan8 contributes to molecular pathways of exosome-induced endothelial cell activation. 81. Barrès C, Blanc L, Bette-Bobillo P, André S, Mamoun R, Gabius HJ, et al. Galectin-5 is bound onto the surface of rat reticulocyte exosomes and modulates vesicle uptake by macrophages. Blood. 2010;115(3):696–705. 82. Hoshino A, Costa-Silva B, Shen TL, Rodrigues G, Hashimoto A, Tesic Mark M, et al. Tumour exosome integrins determine organotropic metastasis. Nature [Internet]. Nature Publishing Group; 2015;527(7578):329–35. Available from: http://dx.doi.org/10.1038/nature15756 83. Denzer K, van Eijk M, Kleijmeer MJ, Jakobson E, de Groot C, J. Geuze H. Follicular Dendritic Cells Carry MHC Class II-Expressing Microvesicles at Their Surface. J Immunol [Internet]. 2000;165(3):1259–65. Available from: http://www.jimmunol.org/cgi/doi/10.4049/jimmunol.165.3.1259 84. Tian T, Zhu YL, Zhou YY, Liang GF, Wang YY, Hu FH, et al. Exosome uptake through clathrin-mediated endocytosis and macropinocytosis and mediating miR-21 delivery. J Biol Chem. 2014;289(32):22258–67. Feng D, Zhao WL, Ye YY, Bai XC, Liu RQ, Chang LF, et al. Cellular internalization of exosomes occurs through phagocytosis. Traffic. 2010;11(5):675–87. 86. Escrevente C, Keller S, Altevogt P, Costa J. Interaction and uptake of exosomes by ovarian cancer cells. BMC Cancer [Internet]. BioMed Central Ltd; 2011;11(1):108. Available from: http://www.biomedcentral.com/1471-2407/11/108 87. Kalra H, Drummen GPC, Mathivanan S. Focus on extracellular vesicles: Introducing the next small big thing. Int J Mol Sci. 2016;17(2). 78. 85. 92 88. 89. Lydic T a., Townsend S, Adda C, Collins C, Mathivanan S, Reid GE. Rapid and comprehensive “shotgun” lipidome profiling of colorectal cancer cell derived exosomes. Methods [Internet]. Elsevier Inc.; 2015; Available from: http://linkinghub.elsevier.com/retrieve/pii/S1046202315001607 Skotland T, Sandvig K, Llorente A. Lipids in exosomes: Current knowledge and the way forward. Prog Lipid Res [Internet]. The Authors; 2017;66:30–41. Available from: http://dx.doi.org/10.1016/j.plipres.2017.03.001 90. Valadi H, Ekström K, Bossios A, Sjöstrand M, Lee JJ, Lötvall JO. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol. 2007;9(6):654–9. 91. Zarovni N, Corrado A, Guazzi P, Zocco D, Lari E, Radano G, et al. Integrated isolation and quantitative analysis of exosome shuttled proteins and nucleic acids using immunocapture approaches. Methods [Internet]. Elsevier Inc.; 2015;87:46–58. Available from: http://dx.doi.org/10.1016/j.ymeth.2015.05.028 92. Channavajjhalaa SK, Rossatoa M, Morandini F, Castagna A, Pizzolo F, Bazzoni F, et al. Optimizing the purification and analysis of miRNAs from urinary exosomes. Clin Chem Lab Med. 2014;52(3):345–54. 93. Batrakova E V., Kim MS. Using exosomes, naturally-equipped nanocarriers, for drug delivery. J Control Release [Internet]. Elsevier B.V.; 2015;219:396–405. Available from: https://doi.org/10.1016/j.jconrel.2015.07.030 94. Zeringer E, Barta T, Li M, Vlassov A V. Strategies for isolation of exosomes. Cold Spring Harb Protoc. 2015;2015(4):319–23. 95. Witwer KW, Buzás EI, Bemis LT, Bora A, Lässer C, Lötvall J, et al. Standardization of sample collection, isolation and analysis methods in extracellular vesicle research. J Extracell vesicles [Internet]. 2013 Jan;2:1–25. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3760646&tool=pmcentrez& rendertype=abstract 96. Szatanek R, Baj-Krzyworzeka M, Zimoch J, Lekka M, Siedlar M, Baran J. The Methods of Choice for Extracellular Vesicles (EVs) Characterization. Int J Mol Sci [Internet]. 2017;18(6):1153. Available from: http://www.mdpi.com/1422-0067/18/6/1153 97. Nolan JP, Stoner SA. A trigger channel threshold artifact in nanoparticle analysis. Cytom Part A. 2013;83 A(3):301–5. 98. Caby M-P, Lankar D, Vincendeau-Scherrer C, Raposo G, Bonnerot C. Exosomal-like vesicles are present in human blood plasma. Int Immunol [Internet]. 2005 Jul [cited 2014 93 Nov 17];17(7):879–87. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15908444 99. Rupert DLM, Claudio V, Lässer C, Bally M. Methods for the physical characterization and quantification of extracellular vesicles in biological samples. Biochim Biophys Acta - Gen Subj [Internet]. Elsevier B.V.; 2017;1861(1):3164–79. Available from: http://dx.doi.org/10.1016/j.bbagen.2016.07.028 100. Palmieri V, Lucchetti D, Gatto I, Maiorana A, Marcantoni M, Maulucci G, et al. Dynamic light scattering for the characterization and counting of extracellular vesicles: a powerful noninvasive tool. J Nanoparticle Res. 2014;16(9). 101. Kesimer M, Gupta R. Physical characterization and profiling of airway epithelial derived exosomes using light scattering. Methods [Internet]. Elsevier Inc.; 2015;87:59–63. Available from: http://dx.doi.org/10.1016/j.ymeth.2015.03.013 102. Lötvall J, Hill AF, Hochberg F, Buzás EI, Di Vizio D, Gardiner C, et al. Minimal experimental requirements for definition of extracellular vesicles and their functions: a position statement from the International Society for Extracellular Vesicles. J Extracell vesicles [Internet]. 2014;3:26913. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4275645&tool=pmcentrez& rendertype=abstract 103. Cianciaruso C, Phelps EA, Pasquier M, Hamelin R, Demurtas D, Ahmed MA, et al. Primary human and rat beta cells release the intracellular autoantigens GAD65, IA-2 and proinsulin in exosomes together with cytokine-induced enhancers of immunity. Diabetes. 2016; 104. Deng Z, Poliakov A, Hardy RW, Clements R, Liu C, Liu Y, et al. Adipose tissue exosome-like vesicles mediate activation of macrophage-induced insulin resistance. Diabetes [Internet]. 2009 Nov [cited 2015 Jan 14];58(11):2498–505. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2768161&tool=pmcentrez& rendertype=abstract 105. Tokarz A, Szuścik I, Kuśnierz-cabala B, Kapusta M, Georgescu A, Stępień E. Extracellular vesicles participate in the transport of cytokines and angiogenic factors in diabetic patients with ocular complications. Folia Med Cracov. 2015;4(4):35–48. 106. Ogata N, Imaizumi M, Nomura S, Shozu A, Arichi M, Matsuoka M, et al. Increased levels of platelet-derived microparticles in patients with diabetic retinopathy. Diabetes Res Clin Pract. 2005;68(3):193–201. 107. Chahed S, Leroyer AS, Benzerroug M, Gaucher D, Georgescu A, Picaud S, et al. Increased vitreous shedding of microparticles in proliferative diabetic retinopathy stimulates endothelial proliferation. Diabetes. 2010;59(3):694–701. 94 108. Sreekumar PG, Kannan R, Kitamura M, Spee C, Barron E, Ryan SJ, et al. αB crystallin is apically secreted within exosomes by polarized human retinal pigment epithelium and provides neuroprotection to adjacent cells. PLoS One [Internet]. 2010 Jan [cited 2014 Jul 2];5(10):e12578. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2951891&tool=pmcentrez& rendertype=abstract 109. Hajrasouliha AR, Jiang G, Lu Q, Lu H, Kaplan HJ, Zhang H-G, et al. Exosomes from retinal astrocytes contain antiangiogenic components that inhibit laser-induced choroidal neovascularization. J Biol Chem [Internet]. 2013 Sep 27 [cited 2014 Aug 24];288(39):28058–67. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23926109 110. Müller G. Studying exosomes and microvesicles-For a better understanding of metabolic diseases? Biol Biomed Rep [Internet]. 2012 [cited 2015 Jan 14];1(1):27–45. Available from: http://www.biomedicalreports.org/index.php?journal=bbr&page=article&op=view&path %5B%5D=87 111. Ogata N, Nomura S, Shouzu A, Imaizumi M, Arichi M, Matsumura M. Elevation of monocyte-derived microparticles in patients with diabetic retinopathy. Diabetes Res Clin Pract. 2006;73(3):241–8. 112. Jha P, Bora PS, Sohn J, Kaplan HJ, Bora NS. Complement System and the Eye. Advances in experimental medicine and biology. 2006. p. 53–62. 113. Joussen a M, Murata T, Tsujikawa a, Kirchhof B, Bursell SE, Adamis a P. Leukocyte- mediated endothelial cell injury and death in the diabetic retina. Am J Pathol. 2001;158(1):147–52. 114. Bora NS, Jha P, Bora PS. The role of complement in ocular pathology. Semin Immunopathol. 2008;30(2):85–95. 115. García-Ramírez M, Canals F, Hernández C, Colomé N, Ferrer C, Carrasco E, et al. Proteomic analysis of human vitreous fluid by fluorescence-based difference gel electrophoresis (DIGE): A new strategy for identifying potential candidates in the pathogenesis of proliferative diabetic retinopathy. Diabetologia. 2007;50(6):1294–303. 116. Aleksandrovskii YA. Antithrombin III, C1 inhibitor, methylglyoxal, and polymorphonuclear leukocytes in the development of vascular complications in diabetes mellitus. Thromb Res. 1992;67(2):179–89. 95 117. Looze C, Yui D, Leung L, Ingham M, Kaler M, Yao X, et al. Proteomic profiling of human plasma exosomes identifies PPARγ as an exosome-associated protein. Biochem Biophys Res Commun [Internet]. Elsevier Inc.; 2009;378(3):433–8. Available from: http://dx.doi.org/10.1016/j.bbrc.2008.11.050 118. Kowal J, Arras G, Colombo M, Jouve M, Paul J, Primdal-bengtson B. Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. PNAS. 2016; 119. Klingeborn M, Dismuke WM, Skiba NP, Kelly U, Stamer WD, Bowes Rickman C. Directional Exosome Proteomes Reflect Polarity-Specific Functions in Retinal Pigmented Epithelium Monolayers. Sci Rep [Internet]. 2017;7(1):4901. Available from: http://www.nature.com/articles/s41598-017-05102-9 120. Caby M-P, Lankar D, Vincendeau-Scherrer C, Raposo G, Bonnerot C. Exosomal-like vesicles are present in human blood plasma. Int Immunol. 2005;17(7):879–87. 121. Muller L, Hong C-S, Stolz DB, Watkins SC, Whiteside TL. Isolation of biologically-active exosomes from human plasma. J Immunol Methods [Internet]. Elsevier B.V.; 2014 Sep [cited 2014 Nov 2];411:55–65. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24952243 122. Zubiri I, Posada-Ayala M, Sanz-Maroto A, Calvo E, Martin-Lorenzo M, Gonzalez-Calero L, et al. Diabetic nephropathy induces changes in the proteome of human urinary exosomes as revealed by label-free comparative analysis. J Proteomics [Internet]. Elsevier B.V.; 2014 Jan 16 [cited 2014 Dec 11];96:92–102. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24211404 123. Urbanelli L, Magini A, Buratta S, Brozzi A, Sagini K, Polchi A, et al. Signaling pathways in exosomes biogenesis, secretion and fate. Genes (Basel) [Internet]. 2013 Jan [cited 2014 Aug 4];4(2):152–70. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3899971&tool=pmcentrez& rendertype=abstract 124. Simpson RJ, Lim JW, Moritz RL, Mathivanan S. Exosomes: proteomic insights and diagnostic potential. Expert Rev Proteomics. 2009/06/06. 2009;6(3):267–83. 125. Hajrasouliha AR, Jiang G, Lu Q, Lu H, Kaplan HJ, Zhang HG, et al. Exosomes from retinal astrocytes contain antiangiogenic components that inhibit laser-induced choroidal neovascularization. J Biol Chem. 2013;288(39):28058–67. 126. Blanc L, Barres C, Bette-Bobillo P, Vidal M. Reticulocyte-secreted exosomes bind natural IgM antibodies: Involvement of a ROS-activatable endosomal phospholipase iPLA2. Blood. 2007;110(9):3407–16. 96 127. Rodriguez-Segade S, Camiña MF, Carnero A, Lorenzo MJ, Alban A, Quinteiro C, et al. High serum IgA concentrations in patients with diabetes mellitus: Agewise distribution and relation to chronic complications. Clin Chem [Internet]. 1996 [cited 2015 Jan 13];42(7):1064–7. Available from: http://www.clinchem.org/content/42/7/1064.short 128. Lobb RJ, Becker M, Wen SW, Wong CSF, Wiegmans AP, Leimgruber A, et al. Optimized exosome isolation protocol for cell culture supernatant and human plasma. J Extracell Vesicles. 2015;4(1):1–11. 129. Kalra H, Adda CG, Liem M, Ang C-S, Mechler A, Simpson RJ, et al. Comparative proteomics evaluation of plasma exosome isolation techniques and assessment of the stability of exosomes in normal human blood plasma. Proteomics [Internet]. 2013 Nov [cited 2014 Nov 22];13(22):3354–64. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24115447 130. Théry C, Amigorena S, Raposo G, Clayton A. Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr Protoc cell Biol. 2006;Chapter 3:1–29. 131. Zubler RH, Lange G, Lambert PH, Miescher P a. Detection of immune complexes in unheated sera by modified 125I-Clq binding test. Effect of heating on the binding of Clq by immune complexes and application of the test to systemic lupus erythematosus. J Immunol. 1976;116(1):232–5. 132. Biro A, Ling WL, Arlaud GJ. Complement protein c1q recognizes enzymatically modified low-density lipoprotein through unesterified fatty acids generated by cholesterol esterase. Biochemistry. 2010;49(10):2167–76. 133. Kitamura D, Roes J, Kühn R, Rajewsky K. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin mu chain gene. Nature. 1991;350:423–6. 134. Jean-François Girmens, Sahel J-A, Marazova K. Dry age-related macular degeneration: A currently unmet clinical need. Intractable Rare Dis Res. 2012;1(3):103–14. 135. Rensen SS, Slaats Y, Driessen A, Peutz-Kootstra CJ, Nijhuis J, Steffensen R, et al. Activation of the complement system in human nonalcoholic fatty liver disease. Hepatology. 2009;50(6):1809–17. 136. Couser WG, Johnson RJ. The etiology of glomerulonephritis: roles of infection and autoimmunity. Kidney Int [Internet]. Elsevier Masson SAS; 2014;86(5):905–14. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0085253815303975 97 137. Xu H, Chen M. Targeting the complement system for the management of retinal inflammatory and degenerative diseases. Eur J Pharmacol [Internet]. Elsevier; 2016;787:94–104. Available from: http://dx.doi.org/10.1016/j.ejphar.2016.03.001 138. Maas SLN, de Vrij J, van der Vlist EJ, Geragousian B, van Bloois L, Mastrobattista E, et al. Possibilities and limitations of current technologies for quantification of biological extracellular vesicles and synthetic mimics. J Control Release [Internet]. Elsevier B.V.; 2015;200:87–96. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0168365914008384 139. Van Der Pol E, Hoekstra AG, Sturk A, Otto C, Van Leeuwen TG, Nieuwland R. Optical and non-optical methods for detection and characterization of microparticles and exosomes. J Thromb Haemost. 2010;8(12):2596–607. 140. Filipe V, Hawe A, Jiskoot W. Critical evaluation of nanoparticle tracking analysis (NTA) by NanoSight for the measurement of nanoparticles and protein aggregates. Pharm Res. 2010;27(5):796–810. 141. Bootz A, Vogel V, Schubert D, Kreuter J. Comparison of scanning electron microscopy, dynamic light scattering and analytical ultracentrifugation for the sizing of poly(butyl cyanoacrylate) nanoparticles. Eur J Pharm Biopharm. 2004;57(2):369–75. 142. Jin AJ, Huster D, Gawrisch K, Nossal R. Light scattering characterization of extruded lipid vesicles. Eur Biophys J. 1999;28(3):187–99. 143. Lane RE, Korbie D, Anderson W, Vaidyanathan R, Trau M. Analysis of exosome purification methods using a model liposome system and tunable-resistive pulse sensing. Nature [Internet]. 2015;5:7639. Available from: http://www.nature.com/doifinder/10.1038/srep07639 144. Jorge I, Burillo E, Mesa R, Baila-Rueda L, Moreno M, Trevisan-Herraz M, et al. The human HDL proteome displays high inter-individual variability and is altered dynamically in response to angioplasty-induced atheroma plaque rupture. J Proteomics [Internet]. Elsevier B.V.; 2014;106:61–73. Available from: http://dx.doi.org/10.1016/j.jprot.2014.04.010 145. Zhang L, Li Y, Payne J, Srivastava S, Fan X, Fung J, et al. Presence of retinal pericyte- reactive autoantibodies in diabetic retinopathy patients. Sci Rep [Internet]. Nature Publishing Group; 2016;6(December 2015):20341. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26839120 146. Ardawi MS., Nasrat HA., Bahnassy A. Serum Immunoglobulin Concentrations in Diabeic Patients. Diabet Med [Internet]. 1994;11(November):384–7. Available from: http://dx.doi.org/10.1016/j.jenvp.2013.01.004%0Ahttp://dx.doi.org/10.1016/j.jenvman. 98 2015.05.022%0Ahttp://dx.doi.org/10.1016/j.paid.2011.03.037%0Ahttp://dx.doi.org/10.1 016/j.jenvp.2010.01.005 147. Mitchell DA, Pickering MC, Warren J, Fossati-Jimack L, Cortes-Hernandez J, Cook HT, et al. C1q Deficiency and Autoimmunity: The Effects of Genetic Background on Disease Expression. J Immunol [Internet]. 2002;168(5):2538–43. Available from: http://www.jimmunol.org/cgi/doi/10.4049/jimmunol.168.5.2538 148. Saunderson SC, Schuberth PC, Dunn AC, Miller L, Hock BD, MacKay P a, et al. Induction of exosome release in primary B cells stimulated via CD40 and the IL-4 receptor. J Immunol. 2008;180(12):8146–52. 149. Saunderson SC, Schuberth PC, Dunn a. C, Miller L, Hock BD, MacKay P a., et al. Induction of Exosome Release in Primary B Cells Stimulated via CD40 and the IL-4 Receptor. J Immunol [Internet]. 2008 Jun 3 [cited 2014 Dec 8];180(12):8146–52. Available from: http://www.jimmunol.org/cgi/doi/10.4049/jimmunol.180.12.8146 150. Blanc L, Barres C, Bette-Bobillo P, Vidal M. Reticulocyte-secreted exosomes bind natural IgM antibodies: involvement of a ROS-activatable endosomal phospholipase iPLA2. Blood [Internet]. 2007 Nov 1;110(9):3407–16. Available from: http://www.ncbi.nlm.nih.gov/pubmed/17666570 151. Qin A, Goldfine N. Glycation Inactivation of the Complement Regulatory Protein Cd59. Diabetes [Internet]. 2004;2653–61. Available from: http://elibrary.ru/item.asp?id=8406366 152. Rossini a a, Like a a, Chick WL, Appel MC, Cahill GF. Studies of streptozotocin-induced insulitis and diabetes. Proc Natl Acad Sci U S A. 1977;74(6):2485–9. 153. KADY N, Liu X, Lydic TA, Syed MH, NAVITSKAYA S, Wang Q, et al. ELOVL4-Mediated Production of Very Long Chain Ceramides Stabilizes Tight Junctions and Prevents Diabetes-Induced Retinal Vascular Permeability. Diabetes. 2018;67(April):1–42. 154. Opreanu M, Lydic T a, Reid GE, McSorley KM, Esselman WJ, Busik J V. Inhibition of cytokine signaling in human retinal endothelial cells through downregulation of sphingomyelinases by docosahexaenoic acid. Invest Ophthalmol Vis Sci [Internet]. 2010 Jun [cited 2014 Aug 15];51(6):3253–63. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2891477&tool=pmcentrez& rendertype=abstract 155. Tauro BJ, Greening DW, Mathias R a, Ji H, Mathivanan S, Scott AM, et al. Comparison of ultracentrifugation, density gradient separation, and immunoaffinity capture methods for isolating human colon cancer cell line LIM1863-derived exosomes. Methods [Internet]. 99 Elsevier http://www.ncbi.nlm.nih.gov/pubmed/22285593 Inc.; 2012 Mar [cited 2014 Oct 31];56(2):293–304. Available from: 156. Schröder M, Schäfer R, Friedl P. Spectrophotometric determination of iodixanol in subcellular fractions of mammalian cells. Anal Biochem. 1997;244(1):174–6. 157. Bradley AJ, Brooks DE, Norris-Jones R, Devine D V. C1q Binding to liposomes is surface charge dependent and is inhibited by peptides consisting of residues 14-26 of the human C1qA chain in a sequence independent manner. Biochim Biophys Acta - Biomembr. 1999;1418(1):19–30. 158. Nydegger UE, Lambert PH, Miesghr P a, Gerber DI. Circulating Immune Complexes in the Serum in Systemic Lupus Erythematosus and in Carriers of Hepatitis B Antigen. J Clin Invest. 1974;54(April):297–309. 159. Biró A, Thielens NM, Cervenák L, Prohászka Z, Füst G, Arlaud GJ. Modified low density lipoproteins differentially bind and activate the C1 complex of complement. Mol Immunol. 2007;44(6):1169–77. 160. Merle NS, Noe R, Halbwachs-Mecarelli L, Fremeaux-Bacchi V, Roumenina LT. Complement system part II: Role in immunity. Front Immunol. 2015;6(MAY):1–26. 161. Acosta J, Hettinga J, Flückiger R, Krumrei N, Goldfine a, Angarita L, et al. Molecular basis for a link between complement and the vascular complications of diabetes. Proc Natl Acad Sci U S A. 2000;97(10):5450–5. 162. Kowal J, Arras G, Colombo M, Jouve M, Morath JP, Primdal-Bengtson B, et al. Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. Proc Natl Acad Sci [Internet]. 2016;113(8):E968–77. Available from: http://www.pnas.org/lookup/doi/10.1073/pnas.1521230113 163. Arlaud GJ, Biro A, Ling WL. Enzymatically modified low-density lipoprotein is recognized by c1q and activates the classical complement pathway. J Lipids. 2011;2011:376092. 164. Whitmore SS, Sohn EH, Chirco KR, Drack A V., Stone EM, Tucker BA, et al. Complement activation and choriocapillaris loss in early AMD: Implications for pathophysiology and therapy. Prog Retin Eye Res [Internet]. Elsevier Ltd; 2015;45:1–29. Available from: http://dx.doi.org/10.1016/j.preteyeres.2014.11.005 165. Benzaquen LR. Terminal complement proteins C5b-9 release basic fibroblast growth factor and platelet-derived growth factor from endothelial cells. J Exp Med [Internet]. 1994;179(3):985–92. Available from: http://www.jem.org/cgi/doi/10.1084/jem.179.3.985 100 166. Jane-wit D, Surovtseva Y V., Qin L, Li G, Liu R, Clark P, et al. Complement membrane attack complexes activate noncanonical NF-κB by forming an Akt + NIK + signalosome on Rab5 + endosomes. Proc Natl Acad Sci [Internet]. 2015;112(31):9686–91. Available from: http://www.pnas.org/lookup/doi/10.1073/pnas.1503535112 167. Georgiannakis A, Burgoyne T, Lueck K, Futter C, Greenwood J, Moss SE. Retinal Pigment Epithelial Cells Mitigate the Effects of Complement Attack by Endocytosis of C5b-9. J Immunol [Internet]. 2015;195(7):3382–9. Available from: http://www.jimmunol.org/lookup/doi/10.4049/jimmunol.1500937 168. Halperin JA, Taratuska A, Nicholson-Weller A. Terminal complement complex C5b-9 stimulates mitogenesis in 3T3 cells. J Clin Invest. 1993;91(5):1974–8. 169. HARSHINI C, NAVITSKAYA S, O’REILLY S, GALLIMORE J, MIZE H, BELI E, et al. Role of Acid Sphingomyelinase in Shifting the Balance Between Proinflammatory and Reparative Bone Marrow Cells in Diabetic Retinopathy. Stem Cells. 2016;34:972–83. 170. Sandrin MS, Mckenzie IFC. Galα(1,3)Gal, the Major Xenoantigen(s) Recognised in Pigs by Human Natural Antibodies. Immunol Rev. 1994;141(1):169–90. 171. Opreanu M, Tikhonenko M, Bozack S, Lydic T a, Reid GE, McSorley KM, et al. The unconventional role of acid sphingomyelinase in regulation of retinal microangiopathy in diabetic human and animal models. Diabetes [Internet]. 2011 Sep [cited 2014 Jul 9];60(9):2370–8. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3161322&tool=pmcentrez& rendertype=abstract 172. Busik J V., Olson LK, Grant MB, Henry DN. Glucose-induced activation of glucose uptake in cells from the inner and outer blood-retinal barrier. Investig Ophthalmol Vis Sci. 2002;43(7):2356–63. 173. Monsinjon T, Gasque P, Chan P, Ischenko A, Brady JJ, Fontaine MC. Regulation by complement C3a and C5a anaphylatoxins of cytokine production in human umbilical vein endothelial cells. FASEB J [Internet]. 2003;17(9):1003–14. Available from: http://www.ncbi.nlm.nih.gov/pubmed/12773483 174. Cortright DN, Meade R, Waters SM, Chenard BL, Krause JE. C5a, but not C3a, increases VEGF secretion in ARPE-19 human retinal pigment epithelial cells. Curr Eye Res. 2009;34(1):57–61. 175. Ting DSW, Cheung GCM, Wong TY. Diabetic retinopathy: global prevalence, major risk factors, screening practices and public health challenges: a review. Clin Exp Ophthalmol. 2016;44(4):260–77. 101 176. Kim MS, Haney MJ, Zhao Y, Mahajan V, Deygen I, Klyachko NL, et al. Development of exosome-encapsulated paclitaxel to overcome MDR in cancer cells. Nanomedicine Nanotechnology, Biol Med [Internet]. Elsevier B.V.; 2016;12(3):655–64. Available from: http://dx.doi.org/10.1016/j.nano.2015.10.012 177. García-Manrique P, Matos M, Gutiérrez G, Pazos C, Blanco-López MC. Therapeutic biomaterials based on extracellular vesicles: classification of bio-engineering and mimetic preparation routes. J Extracell Vesicles [Internet]. Taylor & Francis; 2018;7(1):1422676. Available from: https://www.tandfonline.com/doi/full/10.1080/20013078.2017.1422676 178. Huang C, Fisher KP, Hammer SS, Navitskaya S, Blanchard GJ, Busik JV. Plasma exosomes contribute to microvascular damage in diabetic retinopathy(DR) by activating classical complement pathway. Diabetes, under-review. 102