MICRORESPIROMETRY TO STUDY CERAMIDE - INDUCED MITOCHONDRIAL DYSFUNCTION IN DIABETIC RETINOPATHY By Yan Levitsky A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Physiology Doctor of Philosophy 2020 ABSTRACT MICRORESPIROMETRY TO STUDY CERAMIDE - INDUCED MITOCHONDRIAL DYSFUNCTION IN DIABETIC RETINOPATHY By Yan Levitsky Diabetic retinopathy (DR) is a sight - threatening complication of diabetes mellitus and a leading cause of preventable vision loss worldwide. Classically regarded as a vascular disease, the clinically observable lesions and hallmark histopathological findin gs are found in the vascular compartment . The metabolic insults affecting retinal cells in DR are multifactorial and complex ; however, hyperglycemia, dyslipidemia, and chronic inflammation are thought to be major contributors. Diabetic dyslipidemia affects systemic and local lipid metabolism driv ing the pro - inflammatory and pro - apoptotic retinal cell changes. Sphingolipids are known to play a key role in cell functioning. Ceramides, t he central bioactive sphingolipid species , control cellular responses to c ytotoxic stressors. C eramides can be generated de novo or through sphingomyelinase pathways. I mportant in vivo and in vitro results have demonstrated that acid sphingomyelinase (ASMase) - induced ceramide generation is a major contributor to retinal barrier cell apoptosis in diabetes . Results from studies in ASMase knockout models have shown that these animals are resistant to an array of cytotoxic insults, confirming that ASMase - dependent ceramide generation is important for apoptosis execution . Rece nt reports have demonstrated mitochondrial ceramide accumulation in response to cytotoxic insults in animal and in vi tr o models . Limited observations of oxidative metabolism are reported in the literature as limited biological material is an impediment for comprehensive metabolic characterization of retinal cells in the context of DR. Removal of such obstacles lays the foundation for this work. Comprehensive metabolic examination of DR model systems requires highly sensitive, flexible, and accurate measurem ents of cell or organelle oxygen consumption , a functional measure of oxidative metabolism . The instruments which perform such a measurement are called respirometers. Though commercially available options exist they each present unique limitations which ca n be remediated by rational design of a dedicated microfluidics - based microrespirometer. The first part of this work focuses on the d evelopment of a sensitive and customizable method of measuring O 2 consumption rates by a variety of biological samples in m icroliter volumes without interference from the aerobic environment. The work demonstrates use of 3D printing utilizing photopolymer (VeroClear) to reproducibly form micron - scale microchannels. The photopolymer demonstrat ed low oxygen permeability , optical clarity , and, i n combination with optode - based O 2 sensing, produced a microrespirometer showing > 100x dynamic range for O 2 consumption rates. Measurements are demonstrated with solution - based, suspension - based , and adherent samples . T he role of ASM - dependent mitochondrial ceramide accumulation in diabetes - induced retinal pigment ephithelial cell damage is described next . Mitochondria isolated from diabetic rat retina s (7 weeks duration) showed a n increase in the ceramide - to - sphingomyelin ratio compar ed to controls whereas, the ceramide - to - sphingomyelin ratio was decreased in mitochondria isolated from ASM - knockout mouse retina s compared to wild - type littermates. Cellular ceramide was elevated in RPE cells derived from diabetic donor s compared to contr ol donor s , with a corresponding increase in IL - , IL - 6 , and ASM expression . RPE from diabetic donor s showed fragmented mitochondria and a decreased respiratory control ratio (RCR). Treatment of ARPE - 19 cells with high glucose resulted in a decrease in cit rate synthase functional parameters. These results are consistent with diabetes - induced increase in mitochondrial ceramide through an ASM - dependent pathway leading to impaired mitochondrial function in RPE cells. iv To all my personal giants, whose shoulders allowed me to see so far. v ACKNOWLEDGEMENTS This work would not be possible without the support and guidance from many people. I would first like to thank my advisors Dr. Julia V. Busik and Dr. Denis A. Proshlyakov for all their help and guidance in this endeavor. None of this would be possible without their advice. I felt welcome in both labs and am als o grateful for all the advice and support they provided in personal life as well. The challenges with which I was presented on this journey have been truly transformative and the lessons will not soon be forgotten. Apart from my advisors, I would like to e xtend my sincerest gratitude to my dissertation committee, Dr. Robert Wiseman, Dr. Lawrence ( Karl ) Olson, Dr. Jason Bazil and Dr. Louis Glazer. Their constant support, advice and encouragement was instrumental in this process and it was a rarity for me to walk away from a conversation without a list of topics to learn about. These challenging and insightful interactions have played an important role in shaping me as a scientist. I would also like to thank all of the past and present lab members in the Busik and Proshlyakov labs: Dr. Christopher John, Allison Stettler, Nathan Frantz, Emily Groth, Maggie Conway, Artem Muchnik, Adam Fillion, Dr. Chao Huang, Dr. Qi Wang, Dr. Nermin Kady, Svetlana Navitskaya, Kiera Fisher, Travan Gentles, David J. Pegouske, Phili p Kirschner and Delaney McFarland. I enjoyed working alongside each of them and especially enjoyed talking shop and learning from each one. Dr. Sandra S. Hammer, Svetlana Navitskaya, Kiera Fisher, Dr. Christopher John, and Allison Stettler deserve an extra special thank you for the tremendous amount of help they offered me at varying points in this process, both with the science and otherwise. vi I would also like to acknowledge my undergraduate assistant David J. Pegouske. He was instrumental in the success of this project and I was honored in taking part of his maturation into a capable and independent member of the lab. He will make an excellent physician one day. Finally, I want to extend my deepest appreciation to my friends and family. My mother, father and brother have been my support system throughout this program and my life. Despite us being hundreds of miles apart, I never felt like I was alone. I am especially grateful to my mother for abstaining from bodily injury to my person considering all the h olidays I did not spend with them. The constant encouragement I received from my friends here in Michigan kept me going in the dark times when I just wanted to give up. I am grateful to each one of you. This work would not have been possible without the f inancial support of the National Eye Institute of the National Institutes of Health, the Michigan State University DO/PhD Program, and the Michigan State University Graduate School. vii TABLE OF CONTENTS LIST OF FIGURES ................................ ................................ ................................ ................................ ............ x KEY TO ABBREVIATIONS ................................ ................................ ................................ .............................. xii Chapter 1: Introduction ................................ ................................ ................................ ................................ 1 1.1 Diabetes ................................ ................................ ................................ ................................ .............. 1 1.2 Diabetic Retinopathy ................................ ................................ ................................ .......................... 1 1.2.1 Background ................................ ................................ ................................ ................................ .. 1 1.2.2 Clinical Perspective ................................ ................................ ................................ ...................... 2 1.2.3 Treatment and Prevention ................................ ................................ ................................ ........... 2 1.3 Cellular Changes in Diabetic Retinopathy ................................ ................................ ........................... 4 1.4 Molecular Mechanisms ................................ ................................ ................................ ....................... 6 1.4.1 Hyperglyc emia ................................ ................................ ................................ ............................. 6 1.4.2 Dyslipidemia ................................ ................................ ................................ ................................ . 7 1.4.3 Sphingolipids and Ceramide ................................ ................................ ................................ ......... 7 1.4.4 Downstream Effects of Ceramide Accumulation ................................ ................................ ....... 10 1.4.5 Mitochondrial Ceramide ................................ ................................ ................................ ............ 11 1.5 Purpose and Scope ................................ ................................ ................................ ............................ 12 Chapter 2: Bioenergetics Studies by Respirometry ................................ ................................ .................... 1 3 2.1 Bioenergetics Background ................................ ................................ ................................ ................ 1 3 2.2 Respirometer Evolution ................................ ................................ ................................ .................... 1 4 2.2.1 Warburg Manometer ................................ ................................ ................................ ................. 1 4 2.2.2 Clark - Type Electrode Respirometer ................................ ................................ ........................... 1 6 2.2.3 Modern Iterations ................................ ................................ ................................ ...................... 1 8 2.3 Microrespirometer Design Elements ................................ ................................ ................................ 2 2 2.4 Microchannel Formation ................................ ................................ ................................ .................. 2 3 2.5 Oxygen Sensing ................................ ................................ ................................ ................................ . 2 5 2.5.1 Stern Volmer Rel ation ................................ ................................ ................................ ................ 2 5 2.5.2 Application to Biosensors ................................ ................................ ................................ .......... 2 7 2.6 Whole Cell Respirometry ................................ ................................ ................................ .................. 2 8 2.6.1 Basal Respiration ................................ ................................ ................................ ........................ 3 0 2.6.2 Leak Respiration ................................ ................................ ................................ ......................... 3 0 2.6.3 Maximal Respiration ................................ ................................ ................................ .................. 3 1 2.6.4 Background Oxygen Consumption Rate ................................ ................................ .................... 3 2 2.6.5 Additional Experimental Considerations ................................ ................................ ................... 3 2 2.7 Permeabilized Cell Respirometry ................................ ................................ ................................ ...... 3 3 2.7.1 Permeabilization ................................ ................................ ................................ ........................ 3 4 2.7.2 Localization of Dysfunctional Segments ................................ ................................ .................... 3 6 2.7.3 Leak state ................................ ................................ ................................ ................................ ... 3 7 2.7.4 Complex I ................................ ................................ ................................ ................................ ... 3 7 viii 2.7.5 Complex II ................................ ................................ ................................ ................................ .. 3 8 2.7.6 Complex III ................................ ................................ ................................ ................................ . 39 2.7.7 Complex IV ................................ ................................ ................................ ................................ . 39 2.7.8 Maximum Electron Transport Chain Turnover ................................ ................................ .......... 4 0 2.7.9 Phosphorylation System ................................ ................................ ................................ ............ 4 1 2.7.10 Upstream Substrate Delivery ................................ ................................ ................................ ... 4 1 2.7.11 Conclusion ................................ ................................ ................................ ................................ 4 2 Chapter 3: Micro - Respirometry of Whole Cells and Isolated Mitochondria. ................................ ............. 4 4 3.1 Introduction ................................ ................................ ................................ ................................ ...... 4 4 3.2 Methods ................................ ................................ ................................ ................................ ............ 4 5 3.2.1 Materials ................................ ................................ ................................ ................................ .... 4 5 3.2.2 Micro - respirometric Oxygen Sampling ................................ ................................ ...................... 4 6 3.2.3 Oxygen Permeability and Solubility ................................ ................................ ........................... 4 8 3.2.4 Cell Culture and Respiration Assays ................................ ................................ ........................... 49 3.2.5 Glucose Oxidase Assays ................................ ................................ ................................ ............. 49 3.2.6 Non - Adherent Samples ................................ ................................ ................................ .............. 5 0 3.2.7 Adherent Samples ................................ ................................ ................................ ...................... 5 0 3.2.8 Mitochondrial Isolation and Assay ................................ ................................ ............................. 5 1 3.2.9 Calibration and Data Analysis ................................ ................................ ................................ .... 5 2 3.3 Results ................................ ................................ ................................ ................................ ............... 5 2 3.3.1 3D Printed Chip and Oxygen Optode Geometry ................................ ................................ ........ 5 2 3.3.2 Sample Demand for Cellular Respiration ................................ ................................ ................... 5 3 3.3.3 Isolated Mitochondria ................................ ................................ ................................ ................ 5 6 3.3.4 R O2 of Cells in Suspension ................................ ................................ ................................ ........... 5 7 3.3.5 Variability in Homogeneous Samples ................................ ................................ ........................ 5 8 3.3.6 R O2 of Adherent Cell Samples ................................ ................................ ................................ ..... 59 3.4 Discussion ................................ ................................ ................................ ................................ .......... 6 1 3.5 Conclusion ................................ ................................ ................................ ................................ ......... 6 6 3.6 APPENDIX ................................ ................................ ................................ ................................ .............. 6 8 Chapter 4: Mitochondrial Ceramide Effects on the Retinal Pigment Epithelium in Diabetes .................... 7 1 4.1 Introduction ................................ ................................ ................................ ................................ ...... 7 1 4.2 Methods ................................ ................................ ................................ ................................ ............ 7 3 4.2.1 Rodents ................................ ................................ ................................ ................................ ...... 7 3 4.2.2 Cell Culture ................................ ................................ ................................ ................................ . 7 3 4.2.3 Mitochondrial Isolation ................................ ................................ ................................ .............. 7 4 4.2.4 Mass Spectrometry ................................ ................................ ................................ .................... 7 4 4.2.5 Immunocytochemistry and Mitochondrial Morphology ................................ ........................... 7 5 4.2.6 Quantitative Real - Time Polymerase Chain Reaction ................................ ................................ . 7 6 4.2.7 Western Blot Analysis ................................ ................................ ................................ ................ 7 6 4.2.8 Citrate Synthase Activity ................................ ................................ ................................ ............ 7 7 4.2.9 Microrespirometry ................................ ................................ ................................ ..................... 7 7 4.3 Results ................................ ................................ ................................ ................................ ............... 7 7 4.3.1 Diabetes Results in Retinal Mitochondrial Ceramide Accumulation ................................ ......... 7 7 4.3.2 Diabetes Results in Pro - Inflammatory Changes in Human Retinal Pigment Epithelial (RPE) C ells ................................ ................................ ................................ ................................ ............................ 79 ix 4.3.3 Diabetes Results in Mitochondrial Fragmentation in Human RPE Cells ................................ .... 8 0 4.3.4 Diabetes Induces Acid Sphingomyelinase (ASM) - Mediated Changes in Mitochondrial Function of Human RPE Cells ................................ ................................ ................................ ............................. 8 1 4.3.5 Mitochondrial ASM Contributes to Impaired Mitochondrial Function In Vitro ......................... 8 3 4.4 Discussion ................................ ................................ ................................ ................................ .......... 8 6 Chapter 5: Conclusions and Future Directions ................................ ................................ ........................... 9 2 5.1 Conclusions and Future Directions ................................ ................................ ................................ ... 9 2 REFERENCES ................................ ................................ ................................ ................................ ................ 9 4 x LIST OF FIGURES Figure 1.1. Schematic of the neuroretina and vascular supply . ................................ ................................ .... 5 Figure 1. 2 . Ceramide is the central hub of sphingolipid metabolism. ................................ .......................... 9 Figure 2.1. Schematic representation of oxidative phosphorylation machinery. ................................ ...... 1 3 Figure 2.2. Constant volume (Warburg) manometer. ................................ ................................ ................ 1 5 Figure 2.3. Schematic of a Clark - type electrode respirometer. ................................ ................................ .. 1 6 Figure 2.4. Schematic of exogenous substrate oxidation by whole cells or tissues. ................................ .. 29 Figure 2.5. Functional organization of metabolic processes controlling respiratory activity in isolated mitochondria or permeabilized cells. ................................ ................................ ................................ ......... 3 6 Figure 2.6. Tricarboxylic acid cycle turnover provides substrate for the electron transport chain. .......... 3 8 Figure 3.1. Schematic of the MfR for adherent and non - adherent samples. ................................ ............. 4 7 Figure 3.2. R O2 by ARPE - 19 cell suspension in the microrespirometer. ................................ ...................... 5 3 Figure 3.3. Interface mass transfer of O 2 under zero and maximal gradients. ................................ ........... 5 5 Figure 3.4. Classical states of mitochondrial R O2 in the MfR. ................................ ................................ ...... 5 6 Figure 3.5. Characteristic R O2 states of whole cell in the MfR compared to traditional oxygraph. ............ 5 7 Figure 3.6. Characterization of MfR using homogeneous model reaction. ................................ ................ 5 8 Figure 3.7. Adherent cell configuration of the MfR. ................................ ................................ ................... 59 Figure 3.8. Repetitive R O2 assessment of adhered ARPE - 19 c ells. ................................ .............................. 6 0 xi Figure 3.9. Reversible inhibition of respiration in the MfR. ................................ ................................ ........ 6 1 Figure 3.10. Oxygen permeability of selected polymers. ................................ ................................ ........... 69 Figure 3.11. Biocompatibility of the MfR. ................................ ................................ ................................ ... 69 Figure 3.12. Schematic of the MfR for adherent samples. ................................ ................................ ......... 7 0 Figure 4.1. Negative - ion high - resolution/accurate mass spectrometric quantification of sphingolipids in retinal mitochondria. ................................ ................................ ................................ ................................ .. 7 8 Figure 4.2. Diabetes - induced pro - inflammatory changes in human RPE. ................................ .................. 8 0 Figure 4.3. Structural analysis of human RPE mitochondria. ................................ ................................ ...... 8 1 Figure 4.4. Mic rorespirometric analysis of human RPE cells. ................................ ................................ ..... 8 2 Figure 4.5. Colocalization between ASM and mitochondrial markers. ................................ ...................... 8 4 Figure 4.6. Citrate synthase activity in ARPE - 19 cells. ................................ ................................ ................ 8 5 xii KEY TO ABBREVIATIONS DM Diabetes mellitus DR Diabetic retinopathy BRB Blood - retinal barrier FIELD Fenofibrate Intervention and Event Lowering Diabetes ACCORD Action to Control Cardiovascular Risk in Diabetes VEGF Vascular endothelial growth factor RPE Retinal pigment epithelial ROS Reactive oxygen specie s AGE Advanced glycation end products RAGE Receptors of advanced glycation end products GAPDH glyceraldehyde - 3 - phosphate dehydrogenase S1P Sphingosine - 1 - phosphate SM Sphingomyelin SMases Sphingomyelinases ASMase Acid sphingomyelinase NADH Nicotine adenine dinucleotide FADH 2 Flavin adenine dinucleotide xiii ETC Electron transport chain Pmf Proton motive force PDMS Polydimethylsiloxane CCCP Carbonyl cyanide m - chlorophenylhydrazone ATP Adenosine triphosphate ADP Adenosine diphosphate TCA Tricarboxylic acid mGPDH Mitochondrial glycerophosphate dehydrogenase DHODH Dihydroxyorotate dehydrogenase ETF Electron transferring flavoprotein Q Ubiquinone/ubiquinol pool TMPD P i Inorganic phosphate DPBS Du lb BSA Bovine serum albumin PS Polystyrene KCN Potassium cyanide PtOEP Platinum octaethylporphyrin MfR Microfluidic respirometer xiv PEEK Polyetheretherketone BREC Bovine retinal endothelial cells AA Antibiotic/Antimycotic GOx Glucose oxidase Glu Glucose PMMA Polymethylmethacrylate BB c Base cellular respiration buffer RB c Respiration buffer LB c Leak buffer IB c Inhibition buffer MIB Mitochondrial isolation buffer R O2 Oxygen consumption rates N 2 Nitrogen Glu - GOx Glu cose - glucose oxidase STZ Streptozotocin ASM - / - ASM knockout nESI Nano - electrospray ionization Cer Ceramide PBST Phosphate buffered saline, 1% Tween - 10 xv IL1ß Interleukin 1ß IL6 Interleukin 6 ICAM1 Intercellular adhesion molecule 1 VDAC Voltage - dependent anion channel IgG Immunoglobulin G MS Mass spectrometry RCR Respiratory control ratio DC Differential centrifugation UC Ultracentrifugation UDP - glucose Uridine diphosphate glucose 1 Chapt er 1: Introduction 1.1 Diabetes Diabetes mellitus (DM) is a chronic metabolic disorder characterized by elevations in serum glucose due to decreased production of insulin from the pancreas, type I diabetes, or impaired tissue response to serum insulin, termed type II diabetes. Correspond ing to the rise in obesity and metabolic syndrome, type II diabetes accounts for approximately 90 95% of all cases whereas type I diabetes accounts for approximately 5% [1] . Diabetes mellitus is a major health concern with an estimated 10.5% of the curr ent U.S. population affected by the disease [2] . If current incidence trends persist indefinitely, 1 in 3 persons could be diagnosed with diabetes by 2050 [1,2] . The clinical sequelae of diabetes are coarsely grouped into macro - and microvascular complicat ions. The former includes pathologic changes to large caliber vessels increasing the risk for cardiovascular disease and stroke, whereas the latter are characterized by changes to small vessels resulting in diabetic retinopathy, nephropathy, and neuropathy [3] . As the prevalence of DM continues to increase, so too are the complication rates expected to rise. Despite such bleak predictions, delivery of health care is a powerful tool to stem the tide of disease. For instance, the annual inciden ce of diabetic retinopathy decreased 77% from 1980 to 2007, changes which are attributed to improvements in disease detection and management [4,5] . 1.2 Diabetic Retinopathy 1.2.1 Background In this work, attention is focused on diabetic retinopathy (DR), a common complication of DM and the leading cause of blindness among working age adults [5] . The 10 - year incidence of DR has been estimated at 74% with an estimated 20 25% of diabetic patients developing sight - threatening macular edema in 2 the same time period [5] . Though secondary prevention measures and treatments are available, there are no known cures. 1.2.2 Clinical Perspective Diabetic retinopathy has long been considered a vascular disorder and clinical severity categorizations are based on observ ation of vascular lesions by fundoscopy. Signs of DR include microaneurysms, small red dots in superficial retinal layers signifying outpouching of a capillary wall. Dot and blot hemorrhages result from microaneurysm rupture whereas exudates accumulate due to leakage of serum through a faulty blood - retinal barrier (BRB). Local nerve - fiber layer infarctions are evident as cotton - wool spots and signal local regions of retinal non - perfusion. Other changes include venous loops and beading as well as remodeling of capillary beds, typically found adjacent to non - perfused regions [6] . Diabetic retinopathy is divided into non - proliferative and proliferative stages, with the former further divided into mild, moderate and severe [6] . Mild non - proliferative DR is chara cterized by the presence of at least one microaneurysm, whereas the moderate stage requires presence of hemorrhages, microaneurysms and hard exudates [6] . The severe stage follows the 4 - 2 - 1 rule; hemorrhages or microaneurysms occur in four quadrants, venou s beading in two quadrants and at least one quadrant containing microvascular abnormalities [6] . Progression to the proliferative stage is accompanied by neovascularization, edema, retinal fibrosis and vitreal hemorrhage inducing tractional detachments of the retina [6] . If left untreated, DR progression results in complete vision loss. 1.2.3 Treatment and Prevention Multidisciplinary secondary prevention of DR progression centers on effective and early DM management, typically in primary care settings. Cu rrent standard of care for DM patients focuses on control of serum glucose, lipids, and blood pressure [7 13] . Intensive glycemic control, defined as a hemoglobin A1 C of 3 approximately 7% or less, to reduce DM complication rates in type I and type II patien ts has been well supported by landmark clinical trials [10,13,14] . Long lasting protection of early intensive glycemic control has been demonstrated as well. Upon termination of the Diabetes Control and Complications Trial, both the intensive glycemic cont rol - and standard management cohorts were assigned to intensive glycemic control protocol and long - term follow up revealed the persistence of a protective effect in the early glycemic control cohort, indicating that predisposed tissues are irreversibly cha nged early in the course of DM onset [15] . Such results established the current standard of care which seeks to establish intensive glycemic control as early in the disease process as possible [15] . Despite the strong effect of glycemia normalization to pr event complication onset and progression, the intervention is not curative. In the decades since the glycemia hypothesis has been supported so fully, DR has come to be regarded as complex and multifactorial with disease onset and progression affected by lo cal and systemic factors such as dyslipidemia, blood pressure and inflammation [16,17] . Several observations implicate dyslipidemia in the progression of DR. Firstly, dyslipidemia is associated with formation of hard exudates which are strongly predictive of vision loss [7,16] . Furthermore, the Fenofibrate Intervention and Event Lowering Diabetes (FIELD) study demonstrated the reduced need for laser photocoagulation in DR patients treated with fenofibrate [9,12] whereas the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial showed that DR progression risk was decreased by about one third when fenofibrate was added to simvastatin compared to simvastatin alone [7,8,13] . Blood pressure management showe d similarly protective effects [13] ; however , follow up studies failed to show a significant association [18] . Medical management of DM remains the cornerstone of complication risk reduction. These measures, however, are not curative and efforts to develop specific, vision - saving therapies continue. The earliest effective treatment for DR was pan retinal photocoagulation [19,20] . This technique consists of using high intensity laser light to ablate retinal regions containing abnormal vessels [21] . The prop osed 4 mechanism supposes absorption of the incident laser light by the pigment epithelium resulting in rapid heating and destruction of the outer retina. Though destructive, retinal oxygenation is promoted through the choriocapillaris due to retinal thinnin g, as well as by a decrease in whole retina oxygen demands due to ablation of photoreceptors [21] . Additionally, elimination of hypoxic or ischemia regions of the neuroretina results in decreased secretion of pro - angiogenic factors, likely playing a role i n the protection from disease progression [5,21] . Retinal photocoagulation provided the first vision - preserving treatment indicated for late stage (proliferative) DR ; however, the treatment is aimed at preventing progression to frank vision loss and is acc ompanied by loss of peripheral visual fields, and decreased night vision [21] . Though photocoagulation is still in use today, recent advances in specific medical treatments for DR have focused on controlling the pathological neovascularization process by m odulating inflammatory and pro - angiogenic stimuli commonly found in the diabetic retina [17,22] . The most well studied has been vascular endothelial growth factor (VEGF), a pro - angiogenic growth factor secreted in response to ischemia or inflammation resul ting in pathogenic retinal vascularization [17] . Though photocoagulation remains a popular choice, aptamer - or immune - based therapies against VEGF as well as intravitreal corticosteroids are options for use in many clinical contexts [7] . 1.3 Cellular Chang es in Diabetic Retinopathy The retina is a multilayered neural tissue which transduces photons to electrical currents for transmission to, and interpretation in, the brain. The high metabolic demands of the neuroretina place high perfusion demands on the v ascular supply ; however, retinal architecture is constrained to maintain transparency as a light transducing tissue. Thus, vascular density in the light path of incoming photons is minimized while simultaneously maintaining optimal retinal perfusion. To co pe with such constraints, the retina has evolved a dual blood supply, ensuring adequate perfusion capacity while maintaining a relatively low vascular density in the optical path of incoming photons , as depicted in Figure 1.1 [23] . Lying outside of 5 the lig ht path, the choriocapillaris is a high flow, low arteriovenous pO 2 system found dorsal to the retinal pigment epithelium, supplying the outer one third of the retina. In contrast , the inner two - thirds of the retina is perfused by a series of vascular ple xuses characterized by low flow and high arteriovenous pO 2 [24] . These characteristics suggest that the metabolic demands of the inner retina are matched by a perfusion system operating near maximal capacity whereas perfusion through the choriocapillaris maintain s a high excess capacity for oxygen delivery to the outer retina. Exchange between the blood and retina is tight ly regulated at both vascular supplies, giving rise to two distinct blood - retinal barriers (BRB). The cellular components of the outer barrier include the retinal pigment epithelial (RPE) cells whereas the cellular components of the inner barrier include e pithelial cells and pericytes. Diabetes - induced changes to these components are well described in the literature, however, the cells comprising the inner BRB, retinal endothelial cells and pericytes, have received considerable attention. Indeed, endothelia l cell apoptosis, pericyte dropout and formation of acellular capillaries are the hallmark histopathological changes associated with DR whereas retinal vascular permeability is frequently used as an end point to study disease progression in animal models [ 25] . Figure 1.1 . Schematic of the neuroretina and vascular supply . Select retinal layers (outer nuclear, inner nuclear, ganglion cell layers) and accompanying dual v ascular supplies (choroid, retinal vessels) are depicted on the right. Cell types are shown on the left. The path of an incoming photon is depicted as a yellow arrow (light). RPE = retinal pigment epithelial. 6 Though classically assumed to be a vascular disease, diabetes affects all cells of the retina including the neural and glial cells as well as pericytes, endothelial and epithelial cells. In fact, the hemodynamics of the choroidal and inner retinal vas cular supplies (see above) suggest that the inner retinal metabolic demand is met with a sparse vascular supply operating near maximum capacity resulting in a high sensitivity to metabolic or hypoxic insults [23] . Indeed, neurodegeneration in post - mortem d iabetic eyes without evidence of vascular lesions suggests this to be an early event in DR [26] . The diabetes - induced neuroretinal changes include increased apoptotic marker expression and glial cell activation in the diabetic retina [26 28] . Increased rat es of apoptosis have been reported in the inner plexiform and inner nuclear layers of diabetic rats [29] moreover, ganglion cell apoptosis with associated thinning of the nerve fiber layer are consistently observed across rodent models of diabetes [27,30] . Though vascular dysfunction has been the center of focus in DR research, technological advancements have paved the way to unraveling the complex interplay between the vascular - and neuro - retina [25,31] . 1.4 Molecular Mechanisms The molecular events leadin g to DR progression are complex and varied ; however, it is now appreciated that DR is the product of hyperglycemia, dyslipidemia, and chronic inflammation [17,22,31] . In fact, key in vitro studies have shown that retinal endothelial cells generate reactive oxygen species (ROS) and activate inflammatory and apoptotic pathways in response to cytokine stimulation rather than hyperglycemia [32] . These key findings argue for vascular injury resu lting from a glucose - induced cytokine release attributed to neighboring cells rather than a direct glucose effect on retinal endothelial cells. 1.4.1 Hyperglycemia In cells which cannot downregulate glucose uptake, hyperglycemia is thought to increase glucose - related biochemical pathway flux leading to changes in cellular function [33] . Indeed, increased polyol pathway flux depletes cytosolic NADPH and glutathione, p redisposing cells to oxidative stress [34] . Hyperglycemia 7 also favors the non - enzymatic glycation of proteins which can be irreversibly modified to yield advanced glycation end products (AGE). As AGE are formed irreversibly, long - lived proteins are particu larly vulnerable to accruing a significant population of AGE modified proteins, resulting in changes to protein structure and function [35] . Additionally, AGE bind to receptors of AGE (RAGE) leading to expression of pro - inflammatory and pro - angiogenic gene s [36] , an effect that is seen in diabetes - induced diacylglycerol accumulation and protein kinase C activation [37] . Importantly, a unifying mechanism has been proposed to explain the disparate observations of hyperglycemia - induced cellular damage [33] . Ob servations of ROS emission from mitochondria at a high proton motive force suggest that hyperglycemia induces flux through oxidative phosphorylation result ing in increased proton motive force, ROS release, inhibition of glyceraldehyde - 3 - phosphate dehydroge nase (GAPDH) and glycolysis further favor ing glucose flux through collateral pathways [33,38] . Though hyperglycemia remains the key factor in DM complication rates, other factors are implicated in DR progression and are described below. 1.4.2 Dyslipidemia Diabetic dyslipidemia refers to changes in local and systemic metabolism of a diverse array of lipid classes including plasma triglycerides, cholesterol, and lipoproteins [16] . Plasma lipid and lipoprotein levels and composition are dramatically altered i n diabetes [39] and disease severity is associated with serum lipid markers [40,41] . Retinal specific lipid metabolism leads to a unique fatty acid profile which is significantly altered in diabetes, suggesting that local retinal lipid metabolism plays a k ey role in DR progression [42 44] . T he next section will focus on sphingolipid changes in diabetic retinopathy. 1.4.3 Sphingolipids and Ceramide Sphingolipids are a diverse class of bioactive lipids subserving a variety of cellular functions such as cell growth, proliferation, apoptosis, inflammation and others [45] . Ceramides are the central sphingolipid species, serving as an intersection between major sphingolipid metabolic pathways. The diversity of the 8 sph ingolipidome can be rationalized by considering the availability of multiple modification sites to the chemical structure of ceramides, as shown in Figure 1. 2 A [45] . Ceramides contain two fatty acids which vary in their chain lengths providing for several degrees of freedom in modulating ceramide biological and physicochemical properties [43] . Additionally, the 4 - 5 trans double bond can be reduced to yield dihydroceramide whereas modification of the 1 - hydroxyl position with various head groups yields cerami de - 1 - phosphate, sphingomyelin, or glycosylated ceramides [46] . Ceramide is the central hub of sphingolipid metabolism, connecting all sphingolipid metabolic pathways, allowing the interconversion of distinct sphingolipid species (Figure 1. 2 B) [47] . Total s phingolipid pools are controlled by the relative influx of free fatty acids versus their efflux from the sphingolipid metabolic network. Cellular free fatty acids are incorporated into the sphingolipid metabolic network by condensation with serine to give, in the case that the fatty acid is palmitate, 3 - ketodihydrosphingosine which is subsequently converted to ceramide in the de novo synthesis pathway [45] . Fatty acid efflux from the sphingolipid pool, on the other hand, is through sphingosine - 1 - phosphate ( S1P) hydrolysis by S1P lyase , yielding a fatty aldehyde which can be oxidized to a free fatty acid [45] . The interconversion of distinct sphingolipid species through ceramide suggests that it is the relative ratios of sphingolipid species which determine the final biological response. This feature of sphingolipid ceramide, a pro - apoptotic lipid, to sphingosine - 1 - phosphate, a pro - survival signal [48,49] . Ceramides are generated de novo or by interconversion from other sphingolipid species. A major source of ceramide generation during cellular stress responses is the sphingomyelinase - catalyzed hydrolysis of sphingomyelin (SM) to ceramide and phosphorylcho line [50] . Sphingomyelinases (SMases ) are a group of proteins differentiated by the pH optima at which they demonstrate maximal activity, referred to as acidic, neutral, and alkaline SMases. 9 Alkaline sphingomyelinase has restricted expression in humans, whereas acid and neutral SMases are ubiquitous [16] . Acid SMase (ASMas e) has received much attention for its role in the etiology of Neimann - (a) (b) Figure 1.2. Ceramide is the central hub of sphingolipid metabolism . (a) Chemical moiet ies of a generic ceramide which are sites of chemical transformation yielding a diverse sphingolipidome. Pi = inorganic phosphate, ChoP = phosphorylcholine, m = number of methylene groups in the acyl chain of the exchangeable fatty acid, n = number of meth ylene groups in the acyl chain of the sphingoid base, m may or may not equal n. (b) Sphingolipid metabolic network (green). Ceramide (yellow) provides the link for interconversion of sphingolipid species whereas control over total sphingolipid pools is pro vided by free fatty acid influx at serine palmitoyl transferase and efflux (grey) at sphingosine - 1 - phosphate (S1P) lyase. 10 Pick disease type A, a lysosomal storage disorder where infants present with hepatosplenomegaly, central nervous system involvement, and rarely survive past 2 - 3 years of age [51] . Gener ation of ASMase knockout mouse models led to the observation that ASMase deficiency provided significant protection from a host of apoptotic stressors such as hypoxia, ischemia - reperfusion, and radiation [52] . In fact, cells derived from Niemann - Pick patie nts show significant resistance to ionizing radiation - induced apoptosis [53] . These observations suggest that ASMase - dependent ceramide generation plays a fundamental role in cellular responses to stressors, such as those encountered in the diabetic retin a ( see above ). Consistent with this, the Busik lab has demonstrated protection from retinal ischemia - reperfusion injury in an ASM knockout mouse model compared to wild type controls [54] . Other work in the same laboratory demonstrated that the retina expre sses both neutral and acidic SMases ; however, only ASMase upregulation was observed in ischemia reperfusion injured eyes [54] . Immunoblotting for ASMase protein content in cell culture revealed high expression levels in human retinal endothelial cells wi th smaller but detectable levels in human retinal pigment epithelial and Müller cells. Correspondingly, ASMase enzyme activity was highest in retinal endothelial cells and retinal pigment epithelial cells, with lower levels detected in human Müller cells [ 54] . Activation of ASMase by cytokine treatment results in pro - inflammatory signaling whereas its inhibition leads to reduced inflammatory gene expression in human retinal endothelial cells [55] . These results support a central role for ASMase - dependent ce ramide generation in the increased apoptosis rates characteristic of DR. 1.4.4 Downstream Effects of Ceramide Accumulation Cellular ceramide accumulation is a well - accepted and ubiquitous feature of apoptosis. Though many mechanisms have been proposed to explain the effects of ceramide in the cell, the exact mechanism by which ceramide exerts its biological effects are not kno wn. These effects, furthermore, are likely to involve 11 several parallel effects at distinct sites in the cell depending on the physiological context of each system [56] . Ceramide was once thought to serve as second messenger but is now known to have signifi cant physicochemical effects on biological membranes as well [53,56] . Accumulation of ceramide stiffens membranes, playing a key role in generation of lipid rafts for cell signaling, vesicle budding/fusion, and cell migration [56,57] . Here the focus will b e on the role of mitochondrial ceramide accumulation in apoptotic cells. 1.4.5 Mitochondrial Ceramide Mitochondrial structure and function changes are thought to play an early and central role in DR onset and progression. Diabetes - induced changes to mitoc hondrial structur e and function have been demonstrated using in vivo and in vitro DR models [58 62] . Furthermore, hyperglycemia - induced ROS production has been proposed as a unifying mechanism to explain the disparate biochemical derangements evident in th e hyperglycemic retina and retinal cells [33,58] . While descriptions of diabetes - induced mitochondrial ceramide changes in retinal cells are lacking (see Chapter 3), mitochondria are known to contain several classes of sphingolipids, including sphingomyeli n (SM) and ceramide, as well as enzymes of sphingolipid metabolism [63] . Mitochondrial ceramide accumulates secondary to cytotoxic stressors such as cytokines, UV radiation, and ischemia - reperfusion injury via de novo and SMase - dependent pathways. Whereas exposure of neutrophils to bacterial toxins or UV irradiation of HeLa cells each result in mitochondrial ceramide accumulation in a sphingomyelinase - dependent manner, similar changes in mouse brain after ischemia reperfusion injury are instead attributed t o de novo synthesis [64 66] . Interestingly, over expression of a mitochondrially - targeted bacterial SMase in MCF7 breast cancer cells resulted in apoptosis whereas expression in other cellular compartments had no effect, suggesting that in situ SMase - depen d e nt mitochondrial ceramide accumulation plays an essential role in cellular apoptosis [67] . 12 The effects of mitochondrial ceramide generation depend on the specific site of generation as its poor water solubility prevents spontaneous intermembrane transfer [63] . In the mitochondrial outer membrane, ceramides are thought to facilitate formation of pores resulting in release of proapoptotic cytochrome c [68 72] . Whether ceramide spontaneously forms membrane channels or instead interacts with pro - or anti - apoptotic Bcl - 2 family proteins is still an open question. In the inner mitochondrial membrane, ceramide inhibits respiratory complexes, favoring ROS generation and oxidative stress - induced apoptosis [63,73] . It is noteworthy that these effects are not mutually exc lusive. Because ROS generation is a common early feature of cells exposed to apoptotic stimuli whereas outer membrane permeabilization initiates the execution of apoptosis, the levels of mitochondrial ceramide may fluctuate in a time - and pathway - dependent manner. 1.5 Purpose and Scope The critical role of ASMase - dependent ceramide generation in retinal cell damage in DR and the mounting evidence of mitochondrially targeted ceramide as a control point over cell fate provides the motivation for this work. Fu nctional characterizations of retinal bioenergetic changes are conspicuously missing in the literature due to lack of available comprehensive and flexible methodology. To address this need, the development of a novel 3D printed microrespirometer is describ ed in Chapter 3. Chapter 4 provides descriptions of diabetes - induced, ASMase - dependent mitochondrial ceramide accumulation in retinal pigment epithelial cells and the functional consequences thereof. The next chapter provides a description of respirometry and a framework for its use as a screening platform or as a tool to study oxidative metabolism in finer mechanistic detail. 13 Chapter 2: Bioenergetics Studies by Respirometry 2.1 Bioenergetics Background Bioenerge tics is the study of biological energy transduction, describing how biological oxidation of partially oxidized hydrocarbons ( e . g . carbohydrates from food) is efficiently coupled to ATP generation through electrochemical cellular machinery. The bulk of this machinery resides within the mitochondria, semi - autonomous cellular organelles thought to have originated from the endocytosis and retention of a free - living prokaryote by a eukaryote owing to a symbiotic relationship between the entities [74] . Mitochondria are double bilayered cellular organelles and house the necessary machinery for efficient pathways of biological energy transduction. Modern day understanding of the physicochemical mechanisms of mitochondrial energy transduction originates wit conceptual framework which earned Peter Mitchell the Nobel Prize for Chemistry in 1978 [75] . The Chemiosmotic theory posits that the coupling of ADP phosphorylation with that of the oxidation reactions occurs through gen eration of a proton gradient across the impermeable inner mitochondrial membrane, Figure 2.1 . Schematic representation of oxidativ e phosphorylation machinery . Oxidation of electron carriers by the ETC is coupled to proton extrusion (red arrows) from the mitochondrial matrix and generation of a proton motive force (pmf). Phosphorylation, driven by proton ingress, yields ATP. 14 as depicted in Figure 2.1 [76] . In this conceptual framework, electron carriers (NADH, FADH 2 , CoQ, and cyt c ), reduced through cellular catabolic processes, are oxidized in t he electron transport chain (ETC) resulting in extrusion of protons to the intermembrane space side of the mitochondrial inner membrane (oxidation). Turnover of the ETC, therefore, establishes a proton motive force (pmf) which is used for ATP synthesis (ph osphorylation). The electrochemical gradient induced by the pmf provides the free energy necessary not only for ATP synthesis, but also the transport of a variety of charged species undergoing electrogenic exchange at the impermeable inner mitochondrial me mbrane [77] . Terminal reduction of oxygen occurs at complex IV of the inner mitochondrial membrane. This site serves as a sink for reducing equivalents and oxygen, providing the driving force for the respiratory cascade and forming the basis of using respi rometry to study biological energy transduction [78,79] . The coupling of oxidative phosphorylation through the pmf, as well as the reliance of substrate oxidation on mitochondrial and extra - mitochondrial processes, allows for interrogation of a range of fu nctional segments of cellular metabolism by monitoring respiratory flux [80 82] . Classically, the mainstay of in vitro bioenergetics research is carried out by monitoring the consumption of oxygen by whole tissue, cells or organelle preparations [83] . The instruments used for these studies are respirometers and they provide the continuous monitoring of oxygen concentration needed to derive the rates of oxygen consumption by biological samples. A description of respirometry instrumentation and its applicatio n to biological systems follows. 2.2 Respirometer Evolution 2.2.1 Warburg Manometer Some of the earliest methods for measuring oxygen consumption are attributed to manometric methods, such as used by Otto Warburg to study the metabolism of healthy and tum or tissue leading to the discovery of the Warburg Effect [84] . An illustrated schematic of a Warburg Manometer is depicted in 15 Figure 2.2. The core operating principal of the Warburg constant volume manometer relies on measuring the changes in headspace gas pressure produced by a respiring biological sample in a sealed chamber. Oxidative metabolism results in consumption of O 2 and production of CO 2 [85] . If uncompensated, the CO 2 produ ction abrogates the loss of gas pressure due to O 2 depletion . T herefore, a potassium hydroxide - soaked tissue paper is included in the sealed vessel to scrub CO 2 . By removing the emitted CO 2 , headspace gas pressure changes are attributed entirely to changes in the partial pressure of O 2 . As the total gas in the headspace of the respirometer decreases due to O 2 consumption and CO 2 scrubbing , the height of the manometric fluid in the sample arm of the U shaped tube rises due to the pressure differential betwee n the sample headspace and the atmosphere. A graduated scale on the sample arm of the manometer is used to read the changes in gas volume directly. Measuring the change of O 2 as a function of time provides the requisite raw data necessary to calculate the rate of oxygen uptake by the biological material. Despite the crude methodology by modern standards, manometric methods such as those used by Warburg were precise , accurate, temperature controlled, parallelized for efficiency, and modified to measure both Figure 2.2 . Constant volume (Warburg) manometer. A respiring sample is maintained in a sealed flask connected to a manometric U shaped tube filled with an appropriate manometric fluid. A valve positioned between the sample and manometer opened (vertical, solid line) between trials to equilibrate the buf fer and headspace with the atmosphere and is closed (horizontal, dashed line) during respirometry. As the biological sample metabolizes available carbon sources, the headspace above the sample is depleted of oxygen and enriched in carbon dioxide. The heads pace is scrubbed free of emitted CO 2 by a potassium hydroxide - soaked tissue paper. The graduated scale quantifies the decrease in gas volume due to uptake of O 2 by the biological sample. 16 oxygen consumption and carbon dioxide release by thin tissue slices in vitro [84,86] . These and similar instruments allowed the study of whole animal or tissue metabolism well before the cellular, molecular, and biochemical underpi nnings of metabolic processes were described . The Warburg manometer , however, suffered from several drawbacks which limited its usefulness for bioenergetics in the decades to come. The time resolution of the instrument is poor as changes in headspace gas d epend on the respiratory rate of the sample and conditions of the assay, as well as the volume of the apparatus, and usually require long times between data sampling to accurately resolve headspace pressure differences. Sampling rates, therefore, fell well short of the ~1 Hz easily attainable by modern standards and detailed studies of metabolic changes on short times scales would prove technically challenging [81,87,88] . Additionally, the instrument required a fair amount of technical knowledge and constan t attention to operate, making data acquisition slow and difficult. 2.2.2 Clark - Typ e Electrode Respirometer The next technical breakthrough in respirometer design occurred with the advent of gas analyzers. The most relevant to this work is the Clark electrode, named after its inventor, Leland Clark, as first described [83] . Whereas manometric methods rel ied on inferring changes in partial pressures or volumes of the gaseous analytes by monitoring headspace gas properties, electrochemical detection permitted direct measurement of dissolved oxygen in the aqueous phase [83][89] . This fundamental shift Figure 2.3. Schematic of a Clark - type electrode respirometer . Sample chamber is separated from the Clark - type electrode by a semi - permeable membrane permitting oxygen diffusion. Continuous stirring is typically provided by a magnetic st ir bar and the jacket provides thermoregulation of the sample. The cap is used to partition the sample from atmospheric oxygen and an injection port is typically provided for titrations. 17 to directly measuring oxygen content in biological fluids represents the origin of biosensors and has withstood the test of time as it remains a popular instrument for metabolic studies until today [90] . Modern day respirometers ma ke use of advances in manufacturing, electronics, and materials science but retain the same fundamental detection method and overall design elements as the original Clark - type electrode respirometers [80,91] . A schematic of a generic Clark - type electrode r espirometer is shown in Figure 2.3. The electrode component is made of a platinum cathode and a silver anode, separated from the sample by an oxygen permeable membrane. The opposing face of the membrane is in contact with the sample in the sample chamber. Maintenance of the cathode at - 0.6 V produces the following chemistry on the cathode surface: Reduction of oxygen on the cathode results in a current that is proportional to the [O 2 ] in the sample if several conditions are satisfied. First, the requisite consumption of oxygen at the electrode surface induces concentration gradients which drive oxygen flux from the sample to the surface of the electrode. The greatest impedance to oxygen diffusion lies at the s emipermeable membrane, which is required to prevent fouling of the electrode surface by biological fluids and must be carefully chosen to retain sufficient response times and signal stabilities of the electrode. In fact, Clark tested several candidate memb rane materials including cellophane, dialysis membrane and condoms, showing that the response time and stability of the electrode depended strongly on the composition and thickness of the selected membrane [83] . Modern day instruments use highly optimized materials and film properties to ensure responsive, accurate and stable signals [81] . Next, to relate the current generated at the cathode to the bulk [O 2 ], Clark - type electrode respirometers require a stirred sample volume to eliminate significant oxygen gradients within the sample chamber . If left unstirred, these gradients would limit oxygen delivery to the cathode surface and the current would reflect mass transport kinetics instead of bulk [O 2 ] . Indeed, the landmark studies by Chance and Williams perfo rmed on isolated mitochondria used a vibrating Clark - 18 type electrode to achieve a similar effect allowing them to measure mitochondrial respiratory rates in a range of metabolic states and to determine ADP:O ratios [92,93] . Clark - type electrode respirometer s have been the gold standard tool for bioenergetics studies for many decades and modern - day respirometers have been optimized for the unique requirements of accurate respirometry in the molecular biology revolution (see below). 2.2.3 Modern Iterations Mu ch like electrochemical oxygen detection fundamentally revolutionized respirometry more than six decades ago, so too are modern oxygen detection techniques and manufacturing methods paving the way for innovative designs of modern - day respirometers. The mos t relevant developments of the last several decades are the optode - based chemical sensors which use chemical indicators and spectroscopic methods to detect a range of analytes in a variety of environments including in the gas phase, plant soils, seawater, and biological fluids [94,95] . Oxygen optodes measure the lifetime of the excited electronic state of chemical indicators and relate changes in excited state lifetimes to oxygen content. As oxygen is particularly efficient at quenching such excited states, the lifetime of the excited state indicator is highly sensitive to the oxygen content in the sample [95] . Theoretical and practical considerations of using oxygen optodes in respirometry are described more fully below. Key advantages to using oxygen optod es in modern day respirometers are as follows. As an optical method, optode oxygen sensors do not consume oxygen, increasing the sensitivity of the respirometer to slowly respiring samples and simplifying data analysis and interpretation compared to Clark - type electrode oxygen detection [95] . Furthermore, optodes have the unique advantage that the sensor thin film need not be physically accessible to the experimenter. As an optical technique, the detection instrumentation needs only an optically clear path to the thin film sensor, allowing for the design of sample chambers with minimal connections to potential sinks or sources of oxygen such as the ambient 19 atmosphere. These properties make oxygen optodes popular in microfluidic applications, where oxygen se nsors are incorporated into hypoxia incubators [96 98] as well as respirometer - like devices [88,99 104] . Microfluidic respirometers have the potential to offer the flexibility and sensitivity necessary for adaptation into standard laboratory practice ; howe ver , few examples exist and will be summarized here. Kelbauskas and colleagues developed a two - component microchamber device incorporating phosphorescent O 2 and pH optodes [101] . The device consist s of two separate components, the lid, a silicon wafer including micropocket arrays containing the sensor optodes, and a separate wafer serving as a cell growth substrate. After cell attachment, the two components are aligned and the microchambers hermetica lly sealed using mechanical pressure. The hermetic seal was validated by showing no intrachamber changes in measured [O 2 ] upon incubation of aerobic solutions in anaerobic environments. To further validate the hermetic seal, the group demonstrated linear o xygen concentration profiles until anoxia using respiring human esophageal epithelial cells. These results are consistent with the low apparent K m of complex IV for oxygen and minimal oxygen diffusion into the microchambers [81,101] . Non - negligible oxygen back diffusion ( discussed below ) would manifest as loss of the expected linearity of the oxygen concentration profile due to the dependence of back diffusion on the sample chamber oxygen content [87,88] . While this device focuses on multiplexing biosensors in a static microchamber, the small sample demand (<100 cells) and microliter volume scales can justify the microfluidic classification despite the lack of fluidic handling provisions. Importantly, this work shows that microchambers, housing < 100 cells, can be hermetically sealed rendering them impermeable to gas exchange with the atmosphere, a requisite for accurate respirometry. Using a similar approach, Pham and colleagues developed microchambers for single mitochondrion respirometry [99] . Employing p hotolithography and wet etching of oxygen impermeable glass wafers, they fabricated arrays of microwells into which oxygen optodes were deposited. A lid, manufactured from 20 polydimethylsiloxane (PDMS) and coated with Viton rubber to act as oxygen barrier, w as compressed against the microwell array with a piston to form the microchamber. The microchambers were hermetically sealed as oxygen readings were insensitive to alternating room air (21% O 2 ) with 100% O 2 gas after coating with Viton. Loading of mitochon dria into the wells was stochastic and no attempts at deterministic patterning were made. Nevertheless, the authors reported detection of mitochondrial - dependent respiration, which was sensitive to the uncoupler carbonyl cyanide m - chlorophenylhydrazone (CC CP). Though a higher respiratory rate was observed in wells containing two mitochondria compared to wells containing one mitochondrion, the upper and lower limits of detection and non - mitochondrial apparent oxygen consumption were not reported [99] . Kondr ashina and colleagues reported on the adaptation of commercially available microfluidic slides ( - slides) as microrespirometers by measuring oxygen consumption using soluble fluorescent O 2 sensors [88] . In this paradigm, the group detected oxygen consumpti on from 30,000 cells and further demonstrated that the apparent oxygen consumption was appropriately sensitive to mitochondria lly target ed drugs such as uncouplers and electron transport chain inhibitors. Despite the ease of use exhibited by this system, t he authors noted that oxygen back diffusion through the polymer of the - slide was significant as even the maximum cell density (90,000 cells per slide) failed to deoxygenate the microchannel. Nevertheless, the adapted - slides allowed repetitive assessmen t of respiratory rates by perfusion of fresh media at regular intervals [88] . Prefabricated and sealed microfluidic chips require injection of cell suspensions and subsequent incubation to facilitate attachment and spreading, a process working on the order of minutes to hours. To facilitate maintenance of buffering power and normoxia of the media , while minimizing water loss during the cell attachment phase, such devices utilize polymers with high CO 2 , O 2 and, preferably, low H 2 O permeabilities or solubilit ies. These requirements are in direct opposition to accurate respirometry, discussed below, but, importantly, may be remediated by use of surface coatings for high permeability 21 polymers [105] . Due to the mutual exclusivity of barrier property requirements for cell culture and accurate respirometry, strategies aimed at allowing standard cell culture methods with transient microchamber or microchannel formation are highly desirable. These approaches are described below and in Chapter 3. The most widely adapte d respirometer across a range of biomedical disciplines has no doubt become the SeaHorse Bioscience Extracellular Flux Analyzer [87,106,107] . The extracellular flux analyzer uses a standard 24 - or 96 - well format with a sensor cartridge equipped with fluore scent probes sensitive to O 2 and pH [106] . The sensor cartridge is lowered to within ~200 m of the well bottom, creating a microchamber on the order of ~7 L total volume [87,106] . Formation of transient microchambers using an actuated lid allows for seed ing, culturing and manipulation of cell samples using standard laboratory techniques and conditions whereas the small effective chamber volume during measurement provides high respirometer sensitivities, detecting respiratory activity of 1 10 g mitochon drial protein per well or ~10 5 cells [87,106] . The lid actuating mechanism is further employed to permit repetitive probing of organelle or cell samples by raising and lowering the cartridge to allow mixing of the small probed volume with the bulk media in the well [106] . Sample titrations, to mimic classic titration - based bioenergetics assays, are afforded by inclusion of four reagent delivery chambers which are loaded with appropriate stock solutions prior to measurement and programmatically controlled to deliver the reagents during the assay [106] . The enhanced sensitivity afforded by drastically reduced chamber volumes, high throughput nature of microplate - based assays coupled with multiplexed sensing and the high compatibility with standard cell culture techniques has resulted in widespread adopti on of routine bioenergetic characterizations of systems which have been difficult to study with less sensitive or lower throughput methodology [108] . Despite the substantial improvement in methodology , there is significant oxygen back diffusion into the mi crochamber which limits the sensitivity of the raw measurement and is corrected with a kinetic modelling scheme [87] . Furthermore, the microplate - based method requires adherence of 22 sample s to the well, therefore cells or organelles in suspension must be im mobilized prior to measurements. Finally, e xperimental protocols are limited to a maximum of four reagent additions per experiment, requiring preliminary experiments to determine optimum assay conditions as well as inclusion of several assay repetitions to fully characterize biological systems of interest [106] . The molecular biology revolution of the life sciences has brought novel demands upon the bioenergetics field. Modern day microrespirometers clearly have the potential to retain the properties of cur rent large volume apparatuses . Adoption of modern - day engineering breakthroughs will yield simple, low - cost, accurate , and flexible microrespirometers to afford routine bioenergetic characterization of physiologically relevant model systems. 2.3 Microrespi rometer Design Elements Respirometry involves continuous real - time detection of oxygen content of a given sample. The time - dependent oxygen concentration changes are then used to characterize the functional state of a given biological sample. Engineered fe atures and data analysis strategies are developed to ensure accurate assignment of respiratory rates to functional metabolic states. As an illustrative example, consider measurement of substrate - supported whole cell respiration. The apparent oxygen consump tion rate, defined here as the negative of the oxygen disappearance rate, is the sum of the rates of all processes affecting oxygen concentration in the bulk sample, summarized as: Where R sample is the portio n of the rate attributed to the biological sample, R el is the portion attributed to the electrode and R diff is a generic term to describe mass transfer with any potential sinks or sources of oxygen such as the walls of the container, stir bars, and sites o f interaction with the ambient atmosphere. Extraction of R sample from R app requires knowledge of R el and R diff , by direct intra - experiment measurement 23 or a priori characterization. The rate of oxygen consumption at the electrode and that of diffusive proc esses are both dependent on the oxygen content of the sample. This non - linearity requires complex modeling to offer accurate corrections [87] . Instead, engineering solutions can minimize the contribution of these terms thus providing for accurate and strai ghtforward data analysis and interpretation. These engineering elements are incorporated into the Oroboros O2k oxygraph, which utilizes an optimized Clark - type electrode and large sample chamber volumes to produce stable, accurate currents with small but non - zero R el . These significant improvements yield a large volume respirometer with high resolution of both oxygen concen tratio n s and oxygen consumption rates [109,110] . Further design improvements include careful material selection to minimize potential oxygen sources (R diff ) by using low O 2 solubility/permeability materials for sample chamber construction [80,110] . The result of such optimizations yields a large volume respirometer (~2 mL) capable of measuring respiratory activity of less than 1 million fibroblasts or endothelial cells [80] . The high resolution and high dynamic range permit accurate assessments of oxidative metabolism at the low oxygen concentrations typical of physiologically relevant systems [79,110,111] . As with all Clark - type electrode respirometers, the requisite stirring of the chamber contents provides a technical challenge to measuring natively adherent samples. It stands to reason that similar design elements, if incorporated into a microfluidic system, can effectively yield a useful microrespirometer offering orders o f magnitude increases in sensitivity with enhanced flexibility for experimental design using flow - through configurations. Approache s to implementing the two fundamental elements of a microfluidic respirometer , microchannel formation and analyte detection m ethods, are described below. 2.4 Microchannel Formation Formation of microchannels is the basis for development and manufacture of microfluidic devices. Though many techniques are available, the dominant form of microchannel formation remains casting 24 polym ethylsiloxane (PDMS) on photolithographic molds due to the ease of producing micron - scale features and capacity of PDMS to form strong bonds to glass [112] . Despite the relative ease - of - use, there are several key drawbacks. First, PDMS has high permeabilit y and solubility to water and gases and, furthermore, shows high rates of adsorption of hydrophobic reagents [96] . Additionally, while soft lithography is a powerful tool for embossing surface structures in PDMS, fabrication of three - dimensional geometries is much more complex. Finally, fabrication requires technical knowledge and specialized facilities, creating a high barrier to entry by non - specialists [96] . 3D printing has emerged as a viable alternative to standard micromanufacturing techniques owing to its rapid rise in popularity and technical innovations [113] . Additive manufacturing, or 3D printing, was [114] . 3D printing differs from traditional me thods in that parts are formed by successive layer addition, hence additive, as opposed to removal of stock material by milling, drilling, or cutting to form the final part. The mechanism of formation of each layer varies from fused deposition modeling, wh ere a continuous plastic filament is heated and extruded onto a build plate, to laser - stereolithography, where layers are successively photopolymerized from a liquid resin with a computer guided light source [114] . Resolution of 3D printers, while anisotr opic, has reached the micron scale, and 3D printed microfluidic devices are becoming popular owing to the complex geometries and lower barrier to entry afforded by the wide array of available 3D printing materials and technology [115] . Availability of mate rials has expanded greatly, and parts are available in a host of materials for varying applications requiring optical clarity, variable hardness or other physical specifications [115] . Importantly, the rapid expansion of polymer availabilities in recent ye ars has led to materials with unknown chemical and biological properties [116,117] . The rapid growth in popularity of 3D printing surely predicts a great variety of material choices in the coming years, nevertheless, current work employing uncharacterized 3D printed polymers adopts strategies, such as passive polymer overcoating, 25 to overcome such issues [114,115,117] . Microfluidic devices are designed to incorporate provisions for monitoring specific aspects of biological samples contained within the formed microchannels. The relevant measurement for microrespirometry is real - time detection of dissolved oxygen using optode - based sensing. Theoretical and practical aspects of thin film oxygen sensors are described next. 2.5 Oxygen Sensing Optode - based sensors are composed of three essential components, a chemical sensor, an immobilization matrix and optical excitation and detection instrumentation. Thin film sensors can be applied to a variety of surfaces and sampled using fluorescence intensity or fluorescence lifetime measurements [88,95,118,119] . 2.5.1 Stern Volmer Relation Thin film optodes operate on the principle of intermolecular collisional quenching described by the Stern Volmer relation [95,119,120] . The following is a brief description of the underlying physicochemical processes which lay the foundation for the development of thin fil m biosensors. Consider a chemical species (A) which can absorb a photon to yield its excited state (A*). Relaxation to the ground state can be represented by the following system of equations: Where (1) represents fluorescence or phosphorescence, (2) represents radiation - less decay and (3) represents bimolecular collisional quenching of A* by the quencher, Q. We can consider the time - 26 dependent excited state decay behavior by ass uming a pulse of photons producing a population of excited state molecules ([A*] 0 ) in the presence of excess Q at t = 0. The decay of [A*] can be described kinetically as: Assuming [Q] >> [A*], , and [A *] = [A*] 0 at t = 0, integration yields the pseudo - first order expression of the form: Setting , we can re - arrange (5) to give: In equation (5), k 1 and k 2 are the unimolecular ra te constants describing fluorescence and radiation - less decay respectively, k q is the bimolecular rate constant describing collisional quenching of A* by Q, and, in equation (6), is the lifetime of the excited state fluorophore in the presence of Q. In the case that the system lacks a quenching species, the relaxation pathways are limited to equations (1) and (2) and the corresponding rate expression for [A*] is: Which is integrated analogously to (4) to give: Where k 1 and k 2 are the same constants as in (5) whereas , is the excited state lifetime in the absence of quencher. The ratio: 27 Can be rearranged, to yield the Stern - Volmer relation: Plotting as a function of [Q] reveals a line with slope = 0 k q , from which the bimolecular rate constant can be extracted. 2.5.2 Application to Biosensors T he use of excited state quenching to construct oxygen biosensors relies on measurements of 0 and known quencher concentrations to construct a Stern - Volmer plot which serves as the calibration for oxygen sensing. Fluorometers are required to detect th e changes in excited state lifetimes as described above. For use as an oxygen biosensor, the measurement can be performed with steady - state or pulsed excitations providing intensity or lifetime information, respectively [95,119] . Though both methods are em ployed, lifetime measurements yield greater reproducibility for solid - state thin film sensors as they are independent of dye concentration and film thickness [95,118] . Used as biosensors, the bimolecular rate constant k q is irrelevant and the Stern - Volmer relationship is expressed as , where K SV , the Stern - Volmer quenching constant, is the fitting parameter. Lifetime measurements on unknown samples are interpolated to yield [Q], the analyte of interest. In the case of [O 2 0 is measured by exposing the sensor to a practically feasible anaerobic solution ([O 2 ] anaerobic = 0) and an identical aerobic solution of known [O 2 ] aerobic > 0, typica lly a buffered salt solution at a defined temperature, pressure and humidity [121] . Arbitrary [O 2 ] can be determined in unknown samples provided [O 2 ] anaerobic [O 2 ] [O 2 ] aerobic . The Stern - Volmer relationship derived above assumes homogeneity of the fluo rophore and quencher 28 system. While this assumption is valid for dissolved fluorophores in liquids , the case of optode - based biosensors diverges from such ideal behaviors. Optode - based thin films immobilize a fluorophore in a solid matrix and collisional qu enching relies on diffusion of the quencher through the solid film. This method of entrapment produces inherent heterogeneity of the fluorophore microenvironments resulting in varying quencher accessibilities and a curvilinear Stern - Volmer plot [122] . A co mmon model applied to describe thin - film oxygen sensors with pronounced curvature of the Stern - Volmer plot is the two - site model of the form: Where f 1 and f 2 are the fractional contributions from each site in the absence of quencher whereas K SV1 and K SV2 are the quenching constants for each site [119] . In systems with marked curvature of the Stern - Volmer plot and which require highly precise [O 2 ] determination s, multisite models such as (11) are of importance. Development of an accurate, sensitive, flexible, and inexpensive respirometer allows for comprehensive examination of cell or tissue metabolism. Diverse biological systems can be studied globally, as appr opriate for screening experiments, or in greater mechanistic detail. The next section will describe how respirometry can be used to screen for, and later to localize, metabolic dysfunction. 2 . 6 Whole Cell Respirometry Respirometry of whole cells or tissue slices represents a physiologically relevant system for the determination of metabolic dysfunction. This relevance stems from the dependence of oxygen consumption rates on the entirety of biological substrate oxidation, depicted in Figure 2 .1. A generic su bstrate (S 0 ), destined for full oxidation in mitochondria, is subject to transfer across membranes (1 and 3) and flux through metabolic pathways (2 and 4). Terminal oxidation occurs in the mitochondrial matrix 29 and electron transport chain (ETC) (5), housin g the site of oxygen consumption. Oxidation in the ETC is coupled to generation of a proton motive force (pmf) across the inner mitochondrial membrane (5), whose discharge drives phosphorylation of ADP at ATP synthase (7). The pmf can additionally be disch arged through proton leak pathways without contributing to phosphorylation (6). Alternate metabolic pathways divert intermediates for storage or synthesis (S 2 ) depending on the physiological demands on the system. Ironically, these dependencies are also th e origin of significant limitations to whole cell respirometry. As control over oxygen consumption is simultaneously shared by substrate oxidation, ATP turnover, and proton leak [123,124] , changes in substrate - supported (basal) respiratory activity may, at worst, show false - negatives or, at best, produce detectable changes without adequate information to localize the source. For example, release of mitochondrial cyt - ochrome c during apoptosis impairs transfer of reducing equivalents from Complex III to Complex IV of the ETC and is expected to impair basal respiratory activity. Measurement of basal respirat ory rates in whole cells or tissues, however, may fail to resolve such differences [125] . Figure 2 . 4 . Schematic of exogenous substrate oxidation by whole cells or tissues. Exogenous substrate (S 0 ) traverses the plasma membrane (1) with cytosolic metabolism (2) yielding an intermediate (S 1 ) which is transported into the mitochondrial matrix (3). Oxidation in the mitochondrial matrix (4) produces reducing equivalents (NADH, FADH 2 ) to couple oxidation with proton translocation to generate a proton motive force (5). The proton motive force is discharged through ATP synthase to drive ADP phosphorylation (7) or through a non - (6). Metabolic intermediates can be shuttled into other pathways, represented as (S 2 ) and (8). 30 Metabolic dysfunction in whole cells or tissue is therefore typically assessed using sequential titrations with membrane permeable molecules to shift control over oxy gen flux to different components of cellular metabolism. A typical titration experiment yield s estimates of basal respiration, proton leak supported respiration, maximum respiratory capacity, phosphorylation - linked respiration and non - mitochondrial respira tion [77] . In each case described below, it will be assumed that respiratory states of a treatment group will be compared to the corresponding respiratory states in an appropriate control group. 2 . 6 .1 Basal Respiration Basal respiration is assessed by providing carbon sources to whole cells or tissues which mimic in vivo extracellular substrate availability. The basal respiratory rate is not particularly useful without an estimate of the imposed metabolic load on the sys tem, as oxygen consumption is also under the control of ATP demand in this state. Nevertheless, providing a series of carbon sources can reveal changes in oxidation of lipids, carbohydrates, and amino acids if dysfunctions lead to substantial deficiencies in substrate delivery [77,126] . Detection of altered substrate supported respiration in one class of substrates, palmitate for example, suggests alterations to substrate oxidation pathways. The basal respiratory rate is additionally useful to estimate the proportion of respiration linked to ATP synthesis, which is calculated as the difference between basal respiration rate and the respiratory rate in the presence of an ATP synthase inhibitor [127,128] . A common substrate for whole cell respiration is glucos e ; however , pyruvate lactate, palmitate, and amino acids are provided as replacements or in combination [81,129,130] . 2.6.2 Leak Respiration Mitochondrial membranes are not completely impermeable to protons, and all mitochondria possess inherent proton le ak ((6) in Figure 2 .1) [127,131] . Assessing whole cells or tissues in the leak state involves titrating oligomycin into a respiring sample in the presence of saturating substrates. Oligomycin inhibits ATP synthase, increasing the pmf and decreasing oxygen consumption rates. The low respi ratory rate 31 observed with optimum oligomycin concentrations is controlled by the proton leak with a component consisting of pmf discharge through electrogenic ion exchange at the inner membrane [132] . Detection of increased respiratory rates in the presenc e of oligomycin suggests increased permeability of the inner mitochondrial membrane to protons. Without a corresponding increase in basal respiratory activity, increased proton leak results in diminished ATP generation through oxidative phosphorylation [12 7] . It is important to note here that , at the molecular level, the leak proton current may be due to influx of protons to the matrix without production of ATP (proton leak) or oxidation of reducing equivalents without extrusion of protons from the matrix ( proton slip). Proton leak through the inner membrane is ohmic and inhibition of ATP synthase results in slight hyperpolarization of the inner membrane , accelerati ng the leak current , and, in turn, the respiratory activity. Thus, oligomycin treated samples are a slight overestimate of the true proton leak. Though proton leak is thought to be the major contributor, respirometry alone cannot be used to resolve proton leak and proton slip mechanisms [131,133] . 2.6 .3 Maximal Respiration Maximal respiratory rate s are achieved for a given set of substrates by treatment with chemical uncouplers to dissipate the pmf, shifting control over respiration to substrate delivery and inherent turnover capacity of the ETC [77] . Maximal respiratory rates are typically attribu ted to the highest rates observed with uncoupler titration of oligomycin - inhibited samples. In the presence of oligomycin , however, maximal OCR is underestimated by up to 47%, an effect that is cell type dependent [134] . Accurate determination of maximal O CR in whole cells or tissues, therefore, requires uncoupler titration in the absence of oligomycin. 32 2.6.4 Background Oxygen Consumption Rate Determination of the apparent respiratory rate in the presence of ETC inhibitors, such as rotenone and antimycin A , is required to accurately calculate and assign respiratory rates to metabolic states [81] . Non - mitochondrial oxygen consuming processes originate from both the biological sample and the instrumentation. Addition of inhibitors to measure background rates allows for correction of experimental data and accurate numerical assignments to respiratory states. Such measures are necessary to avoid biasing flux control ratio calculations [123] . 2.6.5 Additional Experimental Considerations Lipophilic reagents, such as oligomycin and uncouplers, must each be titrated to achieve an optimum reagent to membrane ratio and a maximum observable response. Titration accounts for variations in cell density or membrane content and ensures that a consistent respiratory state is achieved in all samples. Chemical uncouplers are particularly important to titrate as they inhibit respiration at high concentrations [111] . If experimental instrumentation limits capability to intra - experimentally titrate compounds, preliminary studies s hould be performed to determine optimal reagent concentrations and, importantly, sensitivity of measured parameters to changes in reagent concentration. Raw respiratory rates reflect respiratory activity contained in the sample chamber and must be normaliz ed to obtain cell - specific or mitochondrial - specific parameter values. Normalization of data can therefore play a significant role in dysfunction localization. Raw respiratory rates can be normalized to i) total cell or tissue mass by cell counting, weighi ng or assaying total protein, ii) mitochondrial content by mtDNA content, immunoblotting for markers, measuring TCA cycle enzyme activity, or iii) a reference respiratory state to serve as internal standard [123,124,135] . Employing each strategy, whether i ndividually or in combination, aids in localizing changes to bioenergetic properties. Normalization to total sample mass or mitochondrial content can resolve mitochondrial content changes from changes to 33 mitochondrial - specific metabolism [127] . Normalizati on to a reference state is the basis for flux control analysis, a framework for analyzing steady state biochemical pathway flux [136] . Each normalization scheme offers benefits and limitations but, importantly, they need not be mutually exclusive. Flux con trol analysis is available with appropriate assignment of reference state, while total protein and mitochondrial markers can be estimated from a common detergent extract [137] . With these considerations in mind, information density can be increased greatly in the case of whole cell or tissue respirometry. While amenable as a screening tool, these procedures suffer from a paucity of mechanistic information. Respirometry can be used to study metabolism in finer detail, as will be described below . 2 . 7 Permeabi lized Cell Respirometry Detailed characterization of bioenergetic (dys)function in a disease model by respirometry requires removal of confounding factors, yielding access to mitochondria with plasma membrane impermeable reagents. Mechanistic bioenergetic characterizations have classically used suspensions of isolated mitochondria in isotonic buffers mimicking the composition of the cytosol. Indeed, much of what we know today about the fundamental mechanisms governing oxidative metabolism were discovered in isolated mitochondria [92,138] . Recently , however, the universal validity of this system has been questioned. Mitochondrial purification results in large losses to sample yield, potential enrichment or depletion of mitochondrial subpopulations, as well as changes to mitochondrial structure, thus limiting the applicability of this method in a wide range of biological systems [139,140] . In contrast, selective permeabilization of cells or tissues provid es a convenient method for direct physical access to mitochondria, mimicking that of the isolated mitochondria case [80,129,141] . In this paradigm, the mitochondria remain constrained within the cytoskeletal framework whereas soluble cytosolic contents are removed by dilution into the extracellular space. Free access to the cytosol allows for use of a variety 34 of substrates and mitochondrial - specific reagents without the limitations imposed by plasma membrane transport and cytosolic processing. 2.7 .1 Permea bilization Selective plasma membrane permeabilization uses detergents or bacterial toxins to interact with plasma membrane cholesterol resulting in pore generation [141,142] . Mitochondrial membranes are spared owing to limited cholesterol content [141,142] . Using respirometry to monitor plasma membrane permeabilization requires assessment of plasma membrane permeability to non - permeable small molecules and quality control assessments to ensure intactness of mitochondrial membranes in the presence of permeab ilizer. The most convenient set of reagents to monitor permeabilization progress are succinate and rotenone [81,135] . As a cell permeable complex I inhibitor, rotenone will maintain low respiratory rates until the normally impermeable succinate gains acces s to the cellular cytosolic space to stimulate respiration through complex II. This combination provides a means to monitor both time - and concentration - dependent effects of whole cell permeabilization. Optimal permeabilizer concentrations ensure maximal p lasma membrane permeabilization without affecting the mitochondrial membranes. Mitochondrial membrane integrity is validated to ensure appropriate permeabilization conditions. Respirometric characterization of outer membrane integrity relies on detection o f respiratory stimulation by addition of exogenous cytochrome c [135] . Stimulation of respiration by exogenous cytochrome c addition, preferably in the presence of saturating ADP or uncoupler, is diagnostic of outer membrane permeability. The inner membran e on the other hand can be probed by measuring a respiratory control ratio using optimal ADP and oligomycin concentrations, allowing direct comparison with known high - quality mitochondrial preparations. 35 It should be mentioned here that permeabilization of cells, and especially of tissues, introduces complications to experimental interpretation. First, permeabilized muscle fibers display a greater than ten - fold increase in the apparent K m for ADP compared to isolated mitochondria from the same tissue [143] . Similar patterns are seen with oxygen affinities in permeabilized muscle fibers, showing distortions in oxygen traces about the aerobic - anaerobic transition suggesting apparent K m for oxygen of 50 M whereas isolated mitochondrial oxygen affinities are est imate between 0.1 10 M [O 2 ] [81] . Finally, translation of findings from isolated mitochondria to permeabilized tissue is not always possible. This is demonstrated by failure to resolve acute changes to ADP - stimulated respiration in saponin - permeabilized liver from ground squirrels in torpor. Characterization of isolated mitochondria, however , showed the expected 60 - 70% decrease [144] . Finally, permeabilized myofibers from young adult and senescent rats showed modest changes to respiratory activity, hydro gen peroxide emission and mitochondrial permeability transition pore induction by calcium, whereas isolated mitochondria demonstrate greater than 50% decreases in maximum ADP - stimulated respiratory rates, with similar patterns in ROS emission and mitochond rial permeability transition pore induction by calcium [145] . In whole tissue preparations, apparent changes to binding affinities are typically attributed to long diffusion distances and, potentially, retention of diffusive barriers with retention of intr a cellular structure s [141,143] . In cases where substrate delivery may be diffusion controlled, these restrictions are partly remediated by titrating reagents to maximum effect. This is exemplified by utilization of hyperoxia and high ADP concentrations (>2 mM) to accurately determine maximum oxidative phosphorylation turnover in permeabilized myofibers [81] . Nevertheless, cautious interpretation of certain calculated parameters, such as binding constants, from respirometry studies using permeabilized tissue s is warranted. Whether using isolated mitochondria or permeabilized cells or tissues, direct access to mitochondria provides the basis for detailed mechanistic studies to localize dysfunctional metabolic segments in a biological system of interest. 36 2.7.2 Localization of Dysfunctional Segments T urnover of the ETC rel ies on substrate delivery and generat es a proton motive force to drive ADP phosphorylation. As the ETC is the site of oxygen consumption, respirometry is well suited to studying these components of oxidative metabolism. A schematic of the functional organization of these pathways is presented i n Figure 2 .2. Addition of exogenous substrates into the sample medium partially reconstitutes the tricarboxylic acid (TCA) cycle, generating reducing equivalents in the form of NADH and/or FADH 2 which enter the ETC. Electron flow also enters the ETC throug h dehydrogenases which do not depend on the TCA cycle, including mitochondrial glycerol phosphate dehydrogenase (mGPDH), di - hydro orotate dehydrogenase (DHODH) and electron - transferr ing flavoprotein (ETF) as examples [80,146] . Electron flows converge on the ubiquinone/ubiquinol pool (Q), which is oxidized by cytochrome c at complex III. Cytochrome c then provides reducing equivalents for oxygen reduction at complex IV. Complexes I, II I, and IV couple redox chemistry with proton translocation to generate a pmf which either drives ADP phosphorylation or is alternatively dissipated through non - productive leak pathways. Figure 2. 5 . Functional organization of metabolic processes controlling respiratory activity in isolated mitochondria or permeabilized cells. Substrate delivery involves oxidation of TCA cycle intermediates to produce electron carriers such as NADH and FADH 2 . Terminal oxidation of electron carriers occurs in the mitochondrial electron transport chain (oxidation), consuming oxygen and producing a proton motive force (pmf). Electron carriers enter the electron transport chain through a variety of enzymes and converge at the ubiquinone/ubiquinol pool (Q), driving flux through complexes III and IV. Cytochrome c couples oxidation of Q with reduction of oxygen at complex IV. The pmf is generated at complexes I, II, and III and dissipated through ATP synthesis (phosphorylation ) or non - productive leak pathways. G3P = glycerol - 3 - phosphate, mGPDH = mitochondrial glycerol phosphate dehydrogenase, DHODH = dihydroorotate dehydrogenase, ETF = electron - transferring flavoprotein. 37 Respirometry can be used to study specific segments of this system to localize metabolic dysfunctions. What follows is a description of respirometry - based assays to localize dysfunctional segments of oxidative metabolism. Equivalent to the whole cell/tissue strategy described above, it is assumed that comparisons will be mad e to appropriate biological control samples in an identical respiratory state. 2.7 .3 Leak state Addition of saturating substrates to respiring mitochondria or permeabilized samples yields a leak state, analogous to that for whole cell/tissue preparations described above. In the presence of saturating substrates, endogenous ADP is quickly phosphorylat ed and, in the absence of retained ATPase activity, minimizes ATP synthase turnover, favoring pmf accumulation [141] . The high pmf decreases respiratory activity and shifts respiratory control to the proton leak through the inner membrane [77,80] . As menti oned above, contamination with ATPases presents a confounding factor as small amounts of adenylates could drive ATP synthase turnover and, in turn, artificially elevate respiratory rates in the leak state. Retention of ATPase activity in permeabilized samp les is especially likely, therefore the leak state in these samples is evaluated by inhibiting ATP synthase with oligomycin in the presence of saturating substrates [141] . 2.7 .4 Complex I Complex I (NADH:ubiquinone oxidoreductase) serves as an entry poin t for NADH generated by activation of matrix dehydrogenases. Redox chemistry is coupled to proton extrusion at complex I, thus this complex contributes to pmf generation. Reconstitution of the TCA cycle forms the basis for determining complex I - linked resp iration [130] . As shown in Figure 2 .3, pyruvate added with malate provides NADH and acetyl - CoA from pyruvate dehydrogenase whereas malate oxidation by malate dehydrogenase provides NADH and oxaloacetate for condensation with acetyl - CoA. Glutamate may be us ed with malate or added to - - ketoglutarate dehydrogenase 38 [81,147] . Succinate accumulation is negligible due to efflux via the dicarboxylate transporter [130] . This substrate combination, therefore, results in generation of NADH but negligible FADH 2 thus entering the ETC predominantly through complex I [130] . The efficiency of respiration using pyruvate/malate or glutamate/malate is cell and tissue specific and preliminary experiments may b e used to evaluate appropriate combinations [147] . Inclusion of saturating ADP or optimized uncoupler is required to limit control over respiratory flux by the phosphorylation system. 2.7 .5 Complex II Complex II (succinate dehydrogenase) is at the interface of the mitochondrial ETC and TCA cycle. Using a FAD - cofactor, succinate oxidation at complex II produces fumarate for the TCA cycle and ubiquino l to drive ETC turnover but does not extrude protons to generate the pmf. Partial reconstitution of the TCA cycle with succinate results in s uccinate dehydrogenase - mediated FADH 2 generation (Figure 2 .3), Figure 2 . 6 . Tricarboxylic acid cycle turnover provides substrate for the electron transport chain. NADH is generated by pyruvate - , isocitrate - - ketoglutarate - , and malate dehydrogenases in the mitochondrial matrix. Succinate dehydrogenase (complex II) fuels ETC flu x through FADH 2 . Glutamate dehydrogenase produces NADH or NADPH. Metabolites and cofactors are shown in block lettering, enzymes displayed in italics. 39 therefore succinate can be added alone to support complex II - linked respiration but reverse electron transfer through complex I is eliminated by inclusion of the complex I inhibitor, rotenone [ 108,135,141] . As above, inclusion of ADP or uncouplers is warranted minimize flux control by the phosphorylation system. 2.7 .6 Complex III Complex III (coenzyme Q:cytochrome c oxidoreductase) catalyzes the oxidation of ubiquino l to ubiquino ne with the co ncomitant reduction of cytochrome c ( Figure 2 .2). Redox chemistry at complex III is coupled to extrusion of protons to generate a pmf. Electron flux through the ETC converges at the Q - pool, and therefore, complexes III and IV present the final common path for electron flow through the ETC [79,130] . The high flux capacities of complexes III and IV are well suited to maintain high ETC turnover in the presence of convergent electron flow from mixed substrates [130,148] . This functional organization introduces artifactual control over respiratory flux by substrate delivery when using complex - specific substrates alone. Using substrate mixtures to maximize flux through complexes III and IV is discussed below ; however, soluble complex III substrates, suc h as durohydroquinone, bypass the Q - pool and support respiratory flux directly through complexes III and IV [149] . Inclusion of malonate or rotenone, to inhibit complexes II and I respectively, favors electron flux toward molecular oxygen. 2.7 .7 Complex I V Complex IV (cytochrome c oxidase) is the terminal oxidase in the ETC, catalyzing the oxidation of cytochrome c , reduction of molecular oxygen, and proton extrusion to generate the pmf. Cytochrome c , the natural substrate for complex IV, cannot cross the outer membrane to supply exogenous reducing - tetramethylphenylenediamine (TMPD) is provided in micromolar amounts in the presence of millimolar ascorbate [149] . TMPD serves as electron donor to drive complex IV turnover whereas ascorbate serves 40 to reduce the oxidized TMPD to maintain steady state turnover [150] . Importantly, substantial background oxygen consumption is expected from TMPD/ascorbate alone which is furth ermore sensitive to temperature, pH and ionic strength of the sample media [81] . It is highly recommended that background oxygen consumption is carefully measured with this system to avoid systematic errors in data interpretation. As above, inclusion of AD P or uncoupler is necessary for assessment of complex IV turnover. 2.7 .8 Maximum Electron Transport Chain Turnover In the fully uncoupled state, oxygen consumption rates are controlled by the intrinsic ETC turnover capacity and substrate delivery. The conceptual framework of convergent electron flow predicts that substrate delivery may become limiting if restricted to complex - specific combinations, such as pyruvate/malate for complex I - linked respiration [151] . In vitro reconstitution of substrate delivery pathways therefore relies on combinations of substrates to maintain high ETC turnover [145] . Respiratory rates in t he presence of pyruvate, malate, and succinate for example, are higher than respiratory rates with pyruvate/malate or succinate/rotenone alone [130,145] . As substrate preference is a system - specific property, preliminary experiments comparing respiratory a ctivities supported by several substrate combinations provides the requisite information necessary for experimental design whereas detectable differences in these parameters could signify metabolic reprogramming in the system of interest. Inclusion of comp lex I - and complex II - linked substrates, such as pyruvate, malate, and succinate , as well as acylcarnitines to stimulate ß - oxidation , can ensure maximum substrate delivery [149] . Additional considerations for enhancing substrate supply lie in the modulatio n of post - translational modifications to enzymes in oxidative metabolic pathways. Treatment with dichloroacetate, for example, favors dephosphorylation and maximal activity of pyruvate dehydrogenase, further enhancing substrate supply to complex I [152] . 41 Using chemical uncouplers to stimulate maximum respiratory activity depolarizes the inner membrane , removing control over respiration by the phosphorylation system. Respirometry is used to study changes in the phosphorylation system by replacing chemical u ncouplers with saturating ADP to activate the whole of oxidative phosphorylation. 2.7 .9 Phosphorylation System The phosphorylation system consists of the adenine nucleotide translocator, phosphate transporter, and ATP synthase which function to import substrates for ATP synthase (ADP and P i ) and to export ATP to the cytosol to meet cellular energy demands [153] . Saturating ADP in the presence of saturating substrates induces maximal turnover of oxidative phosphorylation, composed of the ETC and A TP synthase coupled through the pmf (Figure 2.2) . Detectable changes in oxidative phosphorylation turnover can be attributed to ETC capacity or ATP synthase but, importantly, are easily resolved by subsequent titration with uncoupler [81] . Loss of detectab le differences upon uncoupler titration is consistent with dysfunction in the phosphorylation system whereas retention of differences suggests ETC changes , assuming mitochondrial content is controlled for . Use of ADP in permeabilized samples requires titra tion to maximum effect to identify and minimize potential diffusion limitations [143] . 2.7 .10 Upstream Substrate Delivery Detection of impaired oxidation of a specific combination of substrates aids greatly in localizing dysfunctional metabolic segments ; however, it is incapable of differentiating whether substrate supply or complex - specific changes drive the observed differences [135] . Direct addition of NADH to respiring mitochondria is ineffective at supporting respiration due to poor permeability of NA DH through the inner membrane. Interestingly, use of alamethicin, a bacterial pore - forming toxin, induces NADH permeability through the inner membrane to drive ETC turnover through complex I [135] . As all membranes are permeabilized in the presence of alam ethicin, exogenous cytochrome c is included to maintain ETC 42 turnover. Bypassing substrate delivery pathways to drive flux through the ETC directly, resolves ETC changes from those to upstream pathways including membrane transport and dehydrogenase activity [135] . 2.7 .1 1 Conclusion Measuring rates of oxygen consumption by biological material has been a cornerstone of metabolic research for decades. Respirometry is used to reveal fundamental mechanisms underlying bioenergetics and is currently applied to a b road range of biological systems to study cancer, neurodegeneration, and immune cell metabolism. Respirometer design has evolved to offer highly stable, accurate , and precise instruments capable of operating competently under demanding conditions whereas m iniaturization of sample chambers enable d highly sensitive and high throughput respirometry to be adapted by non - specialist laboratories for routine bioenergetics characterizations. Modern micromanufacturing methods such as 3D printing and the availability of a wide array of materials has enabled the formation of micron - scale features in complex 3D geometries. Combining optode - based oxygen sensing and 3D printing - based micromanufacturing can produce novel, flexible, and highly sensitive microrespirometers f or use in non - specialist laboratories. Respirometry is a powerful tool for bioenergetics studies as it allows for interrogation of a range of biological samples at varying levels of detail. Model systems such as whole cells or tissues are well suited to ra pid screening by respirometry owing to rapid sample preparation, high yields, and the high physiological relevance. Standardized titration protocols yield quantitative estimates of basal, leak, ATP - linked, maximal, and background respiratory rates. Further more, normalization of respiratory rates to measures of total cell/tissue mass or to measures of mitochondrial content resolves inherent metabolic changes from changes to cell/tissue specific mitochondrial content. 43 Isolated mitochondria or permeabilized sa mples provide the requisite access to segment oxidative metabolism into finer detail and efficiently localize changes in metabolism . Retention of mitochondrial membrane integrity allows for in vitro reconstitution of oxidative metabolism to detect changes in pathway - specific substrate delivery including substrate transport and oxidation, complex - specific turnover capacity, maximal ETC capacity, oxidative phosphorylation system capacity, and integrity of inner and outer mitochondrial membranes. The next chapter describes design and development of a novel 3D printed microrespirometer utilizing the design elements discussed above. 44 Chapter 3: Micro - Respirometry of Whole Cells and Isolated Mitochondria. 3. 1 Introduction Oxidation is the most common means of transducing hydrocarbons into energy. Aerobic organisms oxidize molecules in a stepwise manner while synthesizing adenosine triphosphate (ATP), the energy currency of the cell. M itochondria are the specialized organelles where oxidation is coupled to phosphorylation by capturing the energy of electron transport, via a chain of proteins to O 2 , to create a proton gradient across the inner mitochondrial membrane, which then fuels ADP phosphorylation. As the main ATP generation mechanism of the cell, mitochondria have been extensively studied for over a century. The rate of oxygen consumption, the final step of the mitochondrial electron transport chain, provides information about the activity of electron transport chain protein complexes, transporters and ATP synthase. Two major approaches, polarographic or fluorescence quenching, are used for the measurement of O 2 concentration in solution, which provide the necessary data for calcula tion of oxygen consumption rates [83,87,88] . Sophisticated titration protocols using varying substrate and inhibitor combinations were developed to glean information about specific segments of the oxidative phosphorylation machinery. Though t remendous prog ress has been made in understanding mitochondrial function using these approaches, there are several limitations of currently available methodology. As polarographic measurements are based on the current produced by reduction of O 2 on an electrode, oxygen consumption by the sample must be significantly higher than that on the electrode, dictating the high tissue demand of this approach. Fluorescence quenching methods do not impose this demand, however , existing approaches utilize op en configurations that require compensation for ingress of atmospheric oxygen due to diffusion. Finally, both polarographic and fluorescence quenching measurements are performed with static samples that only allow for cumulative, non - reversible titration p rotocols [80,106] . 45 An enclosed flow - through cell respirometer has the potential of improving the currently available methodology by decreasing sample demand, greatly enhancing flexibility, and revolutionizing experimental approaches. The advantages of th e flow - through approach were first demonstrated by Jekabsons and Nicholls [154] . Using oxygen electrodes to determine pre - and post - sample differentials in oxygen tension in a continuous medium steam, they monitored the respiration of primary cerebellar gr anule neuron cultures and determined the ATP supply and demand, proton leak, and mitochondrial respiratory capacity during chronic glutamate exposure. Although revolutionary, this approach required complex custom assembly and was not amenable to automation and scaling up for high - throughput measurements . Recent developments in microfluidics and 3D printing technology provide new opportunities in the development of custom instrumentation [155] . In this study, we used O 2 - impermeable 3D printing plastics to ma nufacture microchannels. Optical transparency of the plastic allowed us to sample an oxygen - sensitive fluorescence - based thin film deposited on the inner surface of the channel without exposing the sample to atmospheric O 2 . We show that adherent cells can be cultured directly on - chip and sampled over prolonged periods of time using repetitive and reversible stimulation of a given sample for observation of metabolic response. In addition to adherent cells, the experimental protocol can be adapted to isolated mitochondria and cell suspensions. Ease of production, flexibility in protocol design, and direct quantitative reporting of O 2 consumption rates make this system highly amenable to both precise individual measurements of traditional respirometry and paral lelization as needed for drug discovery and testing . 3. 2 Methods 3. 2 .1 Materials Dulbecco's phosphate buffered saline (DPBS; D8662), bovine serum albumin (BSA), oligomycin, carbonyl 46 cyanide m - chlorophenyl hydrazone (CCCP), polystyrene pellets (PS), KCN and general laboratory chemicals were of reagent or better grade from Sigma - Aldrich (St. Louis, MO) and were used as acquired. Platinum octaethylporphyrin (PtOEP) was from Frontier Scientific (Logan, Utah). Prefabricated materials were from McMaster - Carr (Elmhurst, IL). 3. 2 .2 Micro - respirometric Oxygen Sampling The microfluidic respiro meter (MfR) was designed in Autodesk Inventor (Autodesk Inc, San Francisco, CA) and printed on an Objet Connex 350 or J750 (Objet Geometries Inc, Billerica, MA) 3D printer in VeroClear, a polymethylmethacrylate - like clear resin (Objet Geometries Inc). MfR consisted of three parts (Figure 3. 1A): the manifold (top), the sensing chip (middle), and the compression base (bottom). The manifold provided fluidic interface to the chip and positioned sampling optical fiber immediately above the optode (below). Contin uity of the ports was achieved by compression of o - rings (size - 001, Buna - N, shore durometer 70A) between the chip and the manifold. The entire assembly was compressed vertically using four bolts, hand tight. Pressure was transferred to the chip and the o - rings via a stacked wave disk spring for better alignment of contact surfaces of the chip and the manifold without pressure points. Two interchangeable versions of the chip were produced: closed shell (Figure 3. 1B) and open shell (Figure 3. 1C) configurati ons for suspension and adherent samples, respectively. A 2.0 mm wide by 0.15 mm deep microchannel of the same geometry was manufactured by 3D printing in VeroClear in both cases. In the closed - shell configuration, the channel was formed by permanent adhes ion of the printed part to the flat glass substrate. Four parts of Loctite EA E - 30CL epoxy mixture were thinned with 1 - part chloroform and applied along the perimeter of the microchannel, ensuring no spillover into the channel. The chip was spun at 2100 rp m for 20 seconds and solvent evaporated for 5 minutes at room temperature. A glass cover, cleaned in acetone, was applied to the chip and cured overnight under mechanical pressure. 47 In the open - shell configuration, the 3D - printed part was split into two components (Figure 3. 1B). The out er section of the chip formed a well and was adhered to the glass the same way as for the closed - shell version. It ensured proper alignment on the central, oval section (insert) against the manifold. The microfluidic channel was formed by pressing the inse rt against the glass without additional adhesives. An elongated, 1.4 mm by 2.6 mm cross section soft seal (shore durometer 26A 28A) was co - printed in Tango+ 3D printing resin along the perimeter of the well to prevent leaks. The seal was positioned along the upper edge of the well (Figure 3. 1B) to provide simultaneous 2 - way contact with the outer edge of the insert and the bottom surface of the manifold. In both configurations, the microchannel - containing surface was oriented upwards on the printer bed to prevent support material deposition in the microchannel. Bulk supporting plastic was removed using a brush with medium - soft plastic bristles followed by overnight soaking in a stirred solution of 2 M potassium hydroxide in water saturated with sodium bicarbonate as a mild abrasive. Cleaned parts were washed with deionized water and air dried before further use. Glass - contacting surfaces of chips in both de Figure 3. 1 : Schematic of the MfR for adherent and non - adherent samples. (A) An exploded view of MfR. Solution flow is shown in blue. Optical fiber and O 2 sensor are shown in red. O - rings are shown in black. Vertical lines show alignment points. (B) Closed shell configuration for suspension me asurements, showing a 3D - printed chip adhered to glass (cyan). (C) Open shell configuration for adherent measurments. Integral seal is shown in black. 48 The O 2 optodes were deposited by casting or spin coating methods. For drop casting, a ~220 mg/mL stock solution of PS in bromobenzene or chloroform was pre pared at room temperature overnight. The working PS/PtOEP solution was prepared by 1:4 (v/v) dilution of PS stock with a 2 mg/mL solution of PtOEP in the same solvent. One microliter of the PS/PtOEP mixture was deposited into the center of the microchannel and dried under a stream of warm air (T < 50 °C). For spin coating, PtOEP (~ 1 mg/mL) and 12.5 25% (m/m) PS in bromobenzene or chloroform solution was applied to the chip followed by spinning at 1000 rpm for 2 minutes. Remaining solvent was removed unde r reduced pressure (0.2 bar) overnight. The measurements are based on the reversible quenching of the luminescence intensity and decay time of PtOEP by oxygen modelled by the Stern Volmer equation 25 . The measurements were performed using a NeoFox GT phase fluorimeter with NeoFox Viewer software (Ocean Optics, Dunedin, FL). Fluorescence lifetime of the PtOEP sensor was acquired through the MfR wall by an unterminated optical fiber (ThorLabs Inc., Newton, New Jersey) at a minimal sensor to fiber distance (Fig ure 3. 1D, see Results). The optical fiber was coupled to a bifurcated optical fiber of the fluorimeter by a bare fiber terminator (ThorLabs). 3. 2 .3 Oxygen Permeability and Solubility Oxygen permeability was measured following ASTM D - 3985 using a Mocon Oxtran 2/21 (Minneapolis, MN) as previously described [156,157] . VeroClear discs were printed at 0.5 mm (N = 2) and 0.2 mm (N =2) thicknesses in two separate batches. Measured thicknesses were within 8% of the nominal thicknesses (0.184 ± 0.008 mm; 0.185 ± 0.010 mm; 0.477 ± 0.003 mm; 0.479 ± 0.009 mm), thus nominal values were used for calculations. Polyetheretherketone (PEEK) film (76.2 µm thick) was cut to size and measured under identic al conditions (N = 4). The thin film was conditioned for 1 hr. Measurements were performed at 23°C and 37°C using dry 100% O 2 at a flow rate of 20 sccm against 10 sccm flow of 98% N 2 , 2% H 2 as a carrier gas. 49 Assessment of oxygen solubility was performed u sing air - equilibrated and anaerobic water in ambient air and in a glovebox (Pas - Labs Inc., Lansing, Michigan) purged with nitrogen (<10 ppm O 2 ). Anaerobic water was prepared using a Schlenk line over 7 cycles between 0.026 atm vacuum and 1 atm Ar gas with agitation. Solutions were delivered to the chip from a gas - tight syringe through a minimal length of PEEK tubing. 3. 2 .4 Cell Culture and Respiration Assays ARPE - 19 cells (ATCC, CRL - 2302, passage 31) were grown on 75 cm2 polystyrene flasks (T75 flask; Cor ning Corning Cellgro, Manassas, VA), 50% F - 12 supplement (Life Technologies, Grand Island, NY), 1% antibiotic/antimycotic mix (Thermo Fisher Scientific, Wal tham, MA) with 10% fetal bovine serum (Atlas Biologicals, Fort Collins, CO). Bovine Retinal Endothelial Cells (BREC) were isolated and cultured as previously described in 10% Fetal Bovine Serum (FBS) Complete Media with 1% Antibiotic/Antimycotic (AA) (Gibc o; ThermoFisher; Waltham, MA) [43] . Passages 4 - 8 were used for all experiments. At 100% confluence, cells were trypsinized (0.25% trypsin - EDTA) (Thermo Fisher Scientific) and counted using the Trypan blue exclusion cell viability method (Sigma Aldrich). Ce lls were either plated on the MfR well or stored, in suspension, on ice for adherent or non - adherent measurements, respectively. 100% confluence, cells were trypsinized, counted and either plated in the MfR well or stored on ice prior to measurements. 3. 2 .5 Glucose Oxidase Assays Stock solutions of glucose oxidase (GOx) and glucose were prepared in 50 mM potassium phosphate buffer, pH 7.5, and diluted in the same buffer as necessary. Assays were prepared by mixing 1:1 (v/v) of glucose and GOx stock solutio ns to a final concentration of 75 mM glucose and varying concentrations of GOx as indicated. The microchannel was flushed with blank buffer between trials. 50 3. 2 .6 Non - Adherent Samples ARPE - 19 cell suspensions were stored in buffer containing 148 mM NaCl, 5 mM KCl, 0.81 mM MgSO 4 , 0.83 mM Na 2 HPO 4 , 0.14 mM KH 2 PO 4 , 1 mM CaCl 2 , 25 mM NaHCO 3 , 15 mM glucose at pH 7.5 [132] . Aliquots of the cell suspension stock were mixed with various buffers ( vide supra ) immediately prior to loading 2.0 - 33×10 3 cells/µL into the MfR using a 20 µL pipette. The microchannel was flushed with PBS between trials. A high - resolution respirometer, Oxygraph - 2k (Oroboros Instruments Corp., Innsbruck, Austria), was used as a reference following standard protocols at 23° C [80,110] . 3. 2 .7 Adherent Samples A removable, 3 mm in diameter by 3 mm in height seeding mask was used in preparation of the adherent cell respirometry. The seeding mask was printed in VeroClear and overcoated with polymethylmethacrylate (PMMA) by dip coa ting into a 5% (m/m) solution of PMMA in methylethylketone and drying overnight at 50° C. The open - shell well (Figure 3.1C) and the seeding mask were sterilized by washing with 70% ethanol in water followed by 15 min of UV irradiation prior to seeding. ARP E - 19 cells were loaded into the mask at densities of 3.5 7.0×10 3 cells/mm 2 and cultured overnight. Photomicrographs were acquired using an inverted light microscope equipped with an AmScope 0.3 megapixel color CMOS camera (United Scope, Irvine, CA) prior to measurements. Surface density was calculated using ImageJ software by converting the RGB photomicrograph into 16 - bit greyscale and applying a median filter (2 pixel radius). Find maxima was used to segment cells (noise tolerance of 5 8) excluding edg e particles. Respiratory activity was measured under stationary medium conditions. Medium flow of 10 - 20 µL/min for re - oxygenation and reagent replenishment was controlled by a syringe infusion pump (Model 22, 51 Harvard Apparatus, MA). Stationary conditions were assured by a manual diversion valve that isolated the chip from the pump. The base cellular respirometry buffer (BB C ) consisted of DPBS with calcium and magnesium (Sigma - Aldrich) supplemented with 0.2% BSA. Respiration buffer (RB C ) consisted of BB C supplemented with 10 mM glucose, 10 mM lactate, and 1 mM pyruvate. Leak buffer (LB C ) contained RB C with 2.5 µM oligomycin. Uncoupling buffer (UB C ) contained LB C with 5 µM CCCP. Inhibition buffer (IB C ) contained UB C with 5 mM KCN [77] . 3. 2 .8 Mitochondrial Isolation and Assay ARPE - 19 cells were trypsinized, resuspended in ice cold mitochondrial isolation buffer (MIB) (200 mM sucrose, 50 mM mannitol, 5 mM MOPS, 1 mM EGTA, 5 mM K2HPO4, pH 7.5) and homogenized using a Teflon pestle (20 strokes at 3000 RPM). Hom ogenate was centrifuged at 600 g for 10 min and supernatant was collected. Pellet was resuspended in 3 mL of fresh ice - cold MIB, homogenized and centrifuged two more times collecting supernatant every time. The pooled supernatants were centrifuged at 7,000 g for 10 min, discarding supernatant. The pellets were resuspended in 200 µL of MIB, pooled and centrifuged again for 10 min at 7000 g, repeating three times. Final pellet was resuspended in a minimal volume of ice - cold MIB and stored on ice prior to use. All steps were performed at 4° C [158] . Isolated mitochondria were stored on ice prior to measurements. The base mitochondrial respiration buffer consisted of 130 mM KCl, 20 mM Tris, 10 mM EGTA, 1 mM sodium phosphate at pH 7.5. Mitochondrial leak buffer contained 5 mM pyruvate, 0.5 mM malate, 10 mM succinate in base buffer. Leak buffer was supplemented with 0.24 mM ADP for ADP - stimulated respiration. Inhibited rates were determined by addition of 5 mM KCN and ADP to leak buffer. Total mitochondrial protein was determined using Bradford assay (BioRad). 52 3. 2 .9 Calibration and Data Analysis Calibration was performed daily using air equilibrated buffer as the aerobic standard (245 µM) and fresh ~1 mM sodium dithionite solution in the same buffer as anae robic standard [121,159] . The Stern - Volmer relation was used to convert fluorescence lifetime to O 2 concentration [118,160,161] . Oxygen consumption rates (R O2 ) were determined by linear fitting of steady state phases. For enzyme assays, background activity was corrected by subtracting R O2 of the buffer blank. For mitochondrial or cellular measurements, the R O2 of inhibited samples was subtracted from all others. Polynomial background drift corrections were applied to O 2 traces in some instances. Data analys is was performed using IgorPro (Wavemetrics Inc., Portland, OR) software. 3. 3 Results 3. 3 .1 3D Printed Chip and Oxygen Optode Geometry Figure 3. 1 shows a schematic of the device consisting of 3D printed flat chip with O 2 sensor film deposited in the microchannel. Optically transparent VeroClear allowed sampling of the sensor by the optical fiber (Figure 3. 1D). Patency of the microchannels was ensured by low - pressure perfusion prior to use. Distance between the optode and the optical fiber was 1.75 mm in the non - adherent configuration and 0.55 mm of view with diameters of 2.48 mm and 1.47 mm for the non - adherent and ad herent configurations, respectively. Drop - coated optode, with a limiting area of 1.8 mm 2 , was probed by an oversized field of view in the non - adherent configuration. In the adherent configuration, a section of an oversized optode was sampled by a limiting 1.7 mm 2 field of view. Nominal dimensions of non - adherent microchannels were 0.25 mm deep by 1.20 mm wide. Average printed microchannel width was 1.09 ± 0.117 mm (N = 3). Average channel depth was 0.235 ± 0.013 mm 53 (N = 3) and 0.145 ± 0.017 mm (N = 3) befor e and after polishing, respectively. Nominal and measured dimension of the adherent configuration were 0.150 mm vs 0.131 ± 0.018 mm (N = 3) in depth and 2.00 mm vs 2.23 ± 0.006 mm (N = 3) in width. 3. 3 .2 Sample Demand for Cellular Respiration ARPE - 19 cells were chosen for their robust respiratory capacity and high rate of proliferation, yielding abundant sample for instrument characterization . A representative oxygen concentra tion trace using suspensions of ARPE - 19 cells is shown in Figure 3.2A and 3.2B. Buffer blanks were used for system equilibration (arrowheads) until stable baseline was observed. Aliquots of ARPE - 19 cells were pre - mixed with respiration buffer (hollow arrow s; 10 mM glucose, 10 mM lactate, 1 mM pyruvate) or inhibition buffer (solid arrows; respiration buffer + 5 mM KCN) and loaded into the microchannel. Separate aliquots were prepared from the same suspension stock and sequentially sampled. After acquiring ce llular R O2 for 3 - 7 min, the microchannel was purged using blank buffer at 2.25 mL/min for 3 - 5 min, resulting in rapid re - oxygenation of the microchannel and no evidence Figure 3.2. R O2 by ARPE - 19 cell suspension in the microrespirometer. (A) Detailed view of ARPE - 1 9 substrate supported and KCN inhibited respiration (12.5×10 3 cells/µL). (B) Reproducibility of alternating injections of respiring and inhibited cells. Arrows and arrowheads indicate cell suspension and blank measurements, respectively. Hollow and solid s ymbols indicate respiration buffer (RB C : 10 mM glucose, 10 mM lactate, 1 mM pyruvate) and inhibition (IB C : RB C + 5 mM KCN) buffer, respectively. Gray arrowhead shows anaerobic calibration. Regions used for determining R O2 are shown by red highlights. (C) B asal R O2 of ARPE - 19 cells with various densities, including linear fit of data (dashed line, y = 0.02 + 4.22 × 10 - 5 x, R 2 = 0.98). Shaded area mean of inhibited cells ± SEM (N=5). R O2 are expressed as mean ± SEM (N=3 - 5). 54 of residual activity in the subsequent blank measurement. Cellular R O2 with inhibition buffer was not distinguishable from blank baseline. Injection of anaerobic standard (gray arrowhead) was used for calibration. Transient artifacts were seen immediately after sample loading, including O 2 consumption during handling, and were excluded from the steady - state R O2 assessment. The microfluidic respirometer sensitivity was characterized using varying cell suspension densities. Figure 3. 2C shows that R O2 scales linearly with cell density (R 2 = 0.98, linear regression) across an order of magnitude change in R O2 . Linear interpolation of observed R O2 , and its errors, showed the lower limit of detection of approximately 500 cells. Variances increased with increasing cell density, effectively limi ting the upper bound of meaningful R O2 determinations. Examination of cell suspension delivery into the MfR by light microscopy revealed non - homogeneous cell distribution at cell densities higher than 17×10 3 cells/µL, likely contributing to the increased v ariability in dense suspension. Oxygen diffusion limits the sensitivity of respirometers and leads to ambiguity in data interpretation. As glass is impermeable to O 2 , diffusion of atmospheric O 2 from the open ports along the major axis of the microchannel and diffusion through the VeroClear walls are the two remaining routes in the microchannel. The first mechanism is unlikely to contribute significantly to O 2 ingress due to the diffusion distance (5.6 mm) and respiration by cells occupying the entire chann el, effectively isolating the cells in the sampled volume. Thus, O 2 diffusion through VeroClear is the most probably source of O 2 ingress. The room temperature (23°C) barrier properties of VeroClear (0.125 ± 0.007 barrer) were comparable to that of PEEK (0 .143 ± 0.001 barrer). Permeability of VeroClear increased 1.8 - fold to 0.218 ± 0.006 barrer at 37°C. Since O 2 permeability at both temperatures was comparable and this study focuses on the fitness of the current approach for respirometry over details such a s metabolic response to temperature, further analysis was performed at ambient temperature. Since the ingress of O 2 into the sample during measurement may lead to a significant underestimation of 55 R O2 , we assessed the bi - directional mass transfer of dissol ved O 2 between the microchannel and the enclosing materials, including plastic and capillaries [79,110,162] . Figure 3. 3 illustrates changes in dissolved O 2 at the minimal and maximal concentration gradients in aerobically (left) and anaerobically (right) c onditioned open - shell chip. To eliminated potential contribution of proteins adsorbed to the chip surfaces, measurement was performed on a fresh chip that was not used in other studies. Sequential sampling of aerobic water in the air - equilibrated chip (Figure 3. 3A) showed no distinguishable changes in the solution [ O 2 ], as expected for zero pO 2 gradient between the plastic and the solution. Rapid replacement of aerobic water with anaerobic water to impose a maximal gradient on the same chip resulted in pronounced ingress of O 2 from plastic into the solution (Figure 3 . 3B). The observed ingress is attributed entirely to the immediate vicinity of the optode and not the upstream components because measurements were performed in the stationary solution. The average ingress rate decreased from 0.04 M/sec to 0.01 M/sec bet ween the first and the third measurement, suggesting that while a substantial Figure 3.3 . Interface mass transfer of O 2 under zero and maximal gradients. Left: Sequential measurements in a fresh, closed - shell chip that remained in the air since printing. Measurements using air - equilibrated (A) and anaerobic (B) water were performed in stationa ry solution and were alternated with 10 - min wash and replenishment phases, shown by the breaks over time axis. Changeover from (A) to (B) included a 20 - min purge with anaerobic water. Right : The same chip was transferred into a glovebox and conditioned for 48 h with N 2 stream over microchannel. Measurements were performed as in the left panel using anaerobic (C), air - equilibrated (D), and again anaerobic (E) water. In all cases, O 2 concertation is expressed as a change from the start of each stationary meas urement. Changes in the concentration indicate ingress (B, E) and egress (D) of O 2 into and from the solution, respectively. 56 amount of O 2 may be dissolved in the thin layer of air - equilibrated plastic, contribution of mass transfer from deeper layers is small, in agreement with direct permeability meas urements. This conclusion was further supported by the opposite trend that was observed when the same chip was pre - conditioned in anaerobic (N 2 ) atmosphere for 48 hours. Initial measurement of anaerobic water in the N 2 - conditioned chip (Figure 3. 3C) showed no ingress of O 2 into the sample. A noticeable loss of O 2 was observed when air - equilibrated water was subsequently measured in the same N 2 - conditioned chip, attributable to the mass transfer of O 2 from the solution into plastic. The O 2 transferred into t he plastic from the aerobic solution can be available for re - entry into the solution until it diffuses further from the interface. This was, indeed, observed when the perfusing solution was reverted to anaerobic water. Figure 3. 3E shows a noticeable, but r apidly diminishing, re - entry of O 2 from transiently oxygenated plastic into the solution. 3. 3 .3 Isolated Mitochondria To demonstrate the suitability of MfR for metabolic studies, R O2 of mitochondria isolated from ARPE - 19 cells was investigated under conditions mimicking standard respirometric measurements (Figure 3. 4). In Figure 3.4 . Classical states of mitochondrial R O2 in the MfR. Red trace indicates inhibition by KCN. Green leak buffer, black leak buffer + 0.24 mM ADP, blue leak buffer + 0.24 mM ADP after complete ADP phosphorylation. Traces were overlaid at t=0 and averaged over the steady state (solid lines). Dashed lines show ± SD. N=2 for inhibition and state II, N=4 for state III and IV. 57 the presence of saturating substrates (State II), mitochondrial respiration is limited by a protonmotive force and the observed R O2 is controlled by ATP hydrolysis and proton leak. Addition of ADP activates ATP synthase, decreasing the proton motive force and increasing the observed R O2 (State III). Upon complete phosphorylation of ADP, the increase in proton motive force again limits the observed R O2 (State IV) to levels similar to the leak state. Background R O2 is determined from KCN - inhibited samples. 3. 3 .4 R O2 of Cells in Suspen sion Similar substrate - inhibitor modulation of whole cell suspensions in the microchannel is shown in Figure 3. 5. Substrate - inhibitor modulation of R O2 by the same cell line in a high - resolution respirometer is shown for direct comparison. Inhibition of A TP synthase and a corresponding increase in the inner membrane potential by oligomycin caused the expected decrease in cell specific R O2 relative to basal R O2 in both approaches (Leak). Maximal cell - specific R O2 , achieved by dissipating the membrane potential using 5 µM CCC P (mitochondrial uncoupler) in the presence of substrates, were 0.025 ± 0.002 and 0.028 ± 0.004 nmol O2 /10 6 cells/sec in the oxygraph and in microfluidic respirometer, respectively (Uncoupled). The increased variability of R O2 in the MfR was attributed to ma nual handling of high - density cellular suspensions. Figure 3. 5 . Characteristic R O2 states of whole cell in the MfR compared to traditional oxygraph. Specific R O2 are inhibition corrected, normalized to cell density and expressed as mean ± SD. N=4 for oxygraph, N=4 - 7 for MfR. 58 3. 3 .5 Variability in Homogeneous Samples To characterize the sensitivity and reproducibility of the MfR independent of the variability associated with suspensions, measurements were performed using a homogeneous glucose - glucose oxidase (Glu - GOx) enzyme system to mimic cellular R O2 . Figure 3. 6A shows a representative trace of sequential Glu - GOx measurements. Inter - trial R O2 reproducibility was high and little residual Glu - GOx activity was detected following washes, which would appear as a loss of O 2 during blank measurements. Figure 3. 6B shows averaged traces of multiplicate measurements of R O2 at [GOx] of 0, 0.09, 4.85 and 333.5 µg/mL with 75 mM Glu. At low [GOx] the R O2 (blue) is statistically indistinguishable from the blank (red). At intermediate [GOx], there is a linear decrease in [O 2 ] (black), and at high [GOx] there is a loss of linearity in the [O 2 ] decrease (gre en). The loss of linearity at the upper limit of MfR sensitivity corresponds to the sensor response time. Maximum apparent R O2 was 5.22 ± 0.43 µM/sec (N = 3), observed following injection of anaerobic standard. R O2 between of 0.03 and 2.5 µM/sec showed a l inear dependence on [GOx] (Figure 3. 6C). The Glu - GOx model was also used to assess inter - instrument variability of the MfR (Figure 3. 6D). R O2 were measured in parallel, on two separate MfR and fluorimeters, using common Glu - GOx stocks. The R O2 Figure 3. 6 . Characterization of MfR using homogeneous model reaction. (A) Raw trace of Glu - GOx R O2 assay. Gray and black arrows show buffer blank and 4.85 µg/mL GOx + 75mM Glu in 50mM KPi (pH 7.5), respectively. (B) O 2 consumption traces by select GOx concentrations. Solid lines indicate overlaid and averaged blank and reaction trials. Dashed lines rep resent ± SD (N = 5 - 6). (C) Dynamic range of the MfR using Glu - GOx R O2 expressed as mean ± SD, N = 3 - 5. (D) Inter - MfR reproducibility using Glu - GOx expressed as mean ± SD, N = 6 - 7. 59 values obtai - 7), were not statistically distinguishable (p = 0.69, t - test). 3. 3 .6 R O2 of Adherent Cell Samples Adherent cell culture models were used to increase reproducibitliy, decrease sample demand, and permit transient sample stimulation using repetitive measurements of a given biological sample. As cell adhesion required extended incubation in proximity to a material with unknown biocompatibility, gross toxicity was assessed by culturing ARPE - 19 cells on - chip, without a seedi ng mask, for several days (Figure 3. 7A and 3. 7B). Photomicrographs, obtained daily, showed no visible differences between cells in the central and peripheral areas, indicating that proximity to polymer did not affect cell growth. After overnight incubation ARPE - 19 cells exhibited an adherent cell morphology. Subsequent proliferation over the three - day incubation period resulted in the cobblestone appearance typical of this cell type [163] . Cell surface density quantitation over time is presented in Figure 3 . 7B, demonstrating an initial increase followed by a plateau at day two in culture. The biocompatibility of the MfR instrument was further confirmed by a long - term continuous measurement of adherent cells, which showed no changes in respiration over 24 hou rs (Figure 3.11 in Appendix A). Following cell culture under standard conditions, the removable insert was used to form a microchannel immediately prior to R O2 measurements. Figure 3.7: Adherent cell configuration of the MfR. Photomicrographs (A) and quantitation (B) of ARPE - 19 cell growth and proliferation on cell culture well of the MfR. Scale bars = 100 µm. 60 Figure 3. 8 shows an example of R O2 measurements in adherent ARPE - 19 cells. Under a constant flow of buffer (Figure 3. 8A, arrows), the microchannel remains f ully oxygenated as O 2 delivery is faster than consumption. Assessment of R O2 was achieved by arresting buffer flow (Figure 3. 8A, arrowheads), resulting in a linear decrease of dissolved O 2 over time. Resumption of flow (Figure 3. 8A, arrows), leads to re - ox ygenation of the microchannel. Analysis showed a mean basal R O2 of 0.030 ± 0.002 (± 8.4%) µM/sec and a mean inhibited R O2 of 0.011 ± 0.002 (± 14.6%) µM/sec (N = 4) as shown in Figure 3. 8B. The continuous medium flow utilized in the microchannel allows for fast and complete exchange of buffer for reversible activation/inhibition cycles using a single cell sample. This capability is illustrated in Figure 3. 9A and 3. 9B. Basal R O2 is inhibited in the presence of the reversible inhibitor KCN. Upon subsequent pe rfusion with KCN - free buffer (RB C ), 99.6% of basal activity is recovered within 880 seconds at 10 µL/min flow. Finally, R O2 of adherent ARPE - 19 cells was examined using the same sequence of modulation as described above for cell suspensions (Figure 3. 5). Figure 3. 9C shows basal, oligomycin - inhibited, CCCP - uncoupled and KCN - inhibited R O2 obtained from 3.4×10 3 cells. The patterns of stimulation and inhibition were recapitulated with adherent cells and, importantly, showed significantly decreased variance of R O2 Figure 3.8 . Repetitive R O2 assessment of adhered ARPE - 19 cells. (A) Representative O 2 trace showing R O2 of 2800 ARPE - 19 cells. Arrowheads no buffer flow, arrows buffer flow, black respiration buffer, gray inhibition buffer, white arrowheads buffer exchange artifact. (B) Q uantitation of basal (RB C ) and inhibited (IB C ) R O2 obtained from (A), expressed as mean ± SD (N=4). 61 using the adherent configuration compared to cell suspensions. Significant and reproducible decreases were noted in R O2 after cell adhesion, as discussed below. 3. 4 Discussion Measurement of dissolved O 2 content provides valuable information on metabolic activity of plant, bacterial, animal and human samples. Sophisticated systems are commercially available for respirometric measurements; however, the advantages of each are balanced by unique limitations such as high sample demand, operating cost, and limited adaptability to sample type [81,106,164] . In this study we combined the advantages of high - resolution 3D printing with fluorometric O 2 detection to produce a simple, versatile, and highly sensitive m ethod for micro - scale repetitive respirometry. Measurements on cell suspensions demonstrated the high sensitivity of the MfR, detecting R O2 in as few as several hundred cells (Figure 3. 2). This represents three orders of magnitude higher sensitivity than l arge volume respirometers, and approximately 10 times the sensitivity of plate - based respirometry [81,106] . Such sensitivities are afforded by the small volumes, tight control over O 2 ingress, and short distances between the optode and respiring cells. Sus pension measurements showed minimal gain from Figure 3.9 . Reversible inhibition of respiration in the MfR. (A) Trace of reversible respiratory inhibition by KCN in adherent ARPE - 19 cells cultured on MfR. Arrows buffer flow, arrowheads static buffer, black RB C , gray IB c , white arrowhead buffer exchange artifact. (B) R O2 quantitation of the results shown in (A). Black circles respiration buffer, gray circles inhibition buffer. (C) Classical states of whole cell R O2 in the adherent compared to suspension MfR configuration. Respiration was in glucose, lactate, pyruvate (10, 10, 1 mM), leak was by addition of oligomycin (2.5 µM), and uncoupled R O2 was by addition of CCCP (5 µM). Squares adherent cells, gray faded circles suspension (from Figure 5). R O2 were normalized to uncoupled respiration and expressed as mean ± SD (N = 3). 62 increasing cell density as higher absolute R O2 were offset by decreased reproducibility. Sensitivity of adherent cell measurements is proportional to surface cell density and inversely proportional to the chann el depth, but is independent of channel width or length, assuming a uniform cell monolayer and sensor width greater than channel depth. Therefore, reduction in channel depth is beneficial until shear stress and O 2 ingress become liming factors (below). The phase - based fluorescence lifetime detection of O 2 by PtOEP in a polystyrene matrix is independent of film thickness and fluorophore concentration, enhancing reproducibility (Figure 3. 6D and Figure 3. 8) [94] . In F igure 3. 5 we showed R O2 modulation by resp iratory effectors, measured using a traditional oxygraph and the MfR. The expected trends were recapitulated in the MfR with some notable exceptions. The magnitude of inhibition caused by oligomycin was smaller in the MfR whereas the magnitude of stimulati on by CCCP was larger. Additionally, MfR cell suspensions showed high R O2 variances compared to polarographic R O2 , where cell densities were 2 - 6 - fold lower than in MfR. Both higher variances and changes in modulation efficiencies are likely to arise from t he higher cell densities in the MfR. Though the chemical composition of assay buffers were the same, the lipophilic modulators, oligomycin and CCCP, require optimization of modulator to membrane ratios in order to achieve optimal membrane concentrations wi thout side effects [165 168] . Despite these differences, we observed excellent agreement of cell - specific uncoupled R O2 between the oxygraph - 2k and microfluidic respirometer, demonstrating the utility of microfluidic respirometry in classical respirometric assays. Uncoupled R O2 was chosen as a normalization between instruments because it represents near - maximal respiratory capacity without the limiting effects of metabolic load. High variances in cell suspension measurements are attributed to micropipetting of concentrated cell stocks and cell sedimentation during sample loading, which can be minimized by incorporating hydrodynamic focusing [169,170] . We reasoned, however, that the high variability between independent cell suspension trials can be remediated by immobilization of the sample, which also improves 63 physiological relevance due cell - cell and cell - substrate contacts in natively adherent cell types. Repeated probing of a given sample in the adherent configuration not only improved variability of R O2 d etermination (Figure 3. 8) but further reduced sample demand relative to cell suspensions, requiring delivery of homogeneous medium between samplings. Importantly, ARPE - 19 cells had lower metabolic rate in the native adherent state compared to the same cell s in suspension. Such changes in metabolic rate upon resuspension of natively adherent cells is expected and highlight the versatility of the microfluidic respirometer, allowing for precise tuning of experimental design to the research question. The inter rupted - flow approach in the adherent microfluidic respirometer configuration enables development of novel measurement strategies [154] . First, cell samples are kept at the desired [O 2 ] (near saturation in this study) because cellular R O2 is measured for a short period before medium replenishment. This allows the microfluidic respirometer to sustain prolonged experiments without inducing metabolic changes associated with hypoxic responses (Figures 3. 8, 3. 9A and 3.10 ). Continuous buffer exchange can mimic cla ssical titration - based protocols and is further amenable to addition and removal of metabolic stimuli to study reversibility of metabolic switches (Figure 3. 9). As an example, we demonstrated reversible respiratory inhibition by KCN (Figure 3. 9A and 3. 9B) and observed the kinetics of R O2 recovery prior to reaching the near - complete pre - inhibition activity. These transient metabolic states are attributed either to partial washout of the inhibitor or, alternatively, to cellular recovery from the metabolic ins ult. Regardless of the cause, these observations open intriguing possibilities for the investigation of time - dependent metabolic changes affecting R O2 in real - time and the opportunity to resolve transient metabolic states during stimulation. Isolation of t he sample from the atmospheric environment is particularly important for microrespirometry due to the high surface area to volume ratios inherent in microfluidics. In such regimes, surface exchange of O 2 can lead to relatively rapid changes in bulk O 2 concentrations in the medium. This property is used widely for the development of microfluidic hypoxia incubators which focus on controlling 64 and rapidly changing dissolved oxygen concentrations within microchannels [161,171] . These devices, however, are f undamentally different from respirometers. Whereas control over dissolved oxygen concentration in a microchannel requires highly permeable polymers, microrespirometers require low permeability barriers because real - time detection of bulk analyte and the ki netic analysis of its dynamics are greatly hindered in the presence of extraneous analyte sources or sinks [161,172] . Ingress of atmospheric O 2 into the microchannel can adversely affect results, decreasing instrument sensitivities and causing non - linear r esponses due to the accumulation of concentration gradients and diffusion according [87,88,105] . The non - linearity of steady - state R O2 is particularly problematic for multi - phasic processes, such as in the transition from ADP - dependent to ADP - limited respiratory states of isolated mitochondria (Figure 3. 4). For example, ADP - limited respiration could be under - estimated when samples reach lower [O 2 ], affecting calculated parameters such as respiratory control and ADP:O ratios [93,110,173] . Excel lent barrier properties of VeroClear play a critical role in preventing interference from atmospheric O 2 . However, VeroClear has some O 2 buffering capacity due to a limited solubility of O 2 in the plastic. Oxygen transfer between the solution and the plast ic results in small changes in the solution O 2 following rapid imposition of concentration gradient of over 200 M. Even at the maximal gradient, the observed mass transfer was significantly slower than the R O2 reported elsewhere in this study. Since mass transfer is directly proportional to the difference of concentration at the boundary of the sample and all biological measurements were performed with 5 to 10 times smaller gradients, effective contribution of mass transfer is negligible. It is important t o note that the maximal gradient conditions were designed to test the limits of O 2 mass transfer at the boundary and do not represent a biologically relevant model. A well - designed applied experiment should be performed at relatively constant [O 2 ] at physi ological pO 2 , including hypoxic conditions. In such cases, pre - conditioning of the MfR by gas stream or a flow of a solution is the best strategy. 65 The results shown in Figure 3. 3A clearly demonstrated that there is no chemical O 2 scavenging by the resin it self. This shows that a small, but consistent, non - zero initial R O2 drift with static blank solutions (Figures 3. 2A, 3. 2B and 3. 6A) may originate from partial retention of proteins or organelles from previous measurements. Such background drifts are much s maller than mitochondrial R O2 and are corrected with standard respiratory inhibition controls in differential measurements. While reduction in the depth of the channel enhances sensitivity in adherent cell measurements, it also increases shear stress on ce lls during perfusion. This physiologically relevant stimulus is lacking in traditional cell culture models [172,174,175] . Physiological shear stress can range from 10 - 2 to 10 1 dyne/cm 2 [175,176] . In the adherent MfR configuration, ARPE - 19 cells experience d an estimated shear stress of 0.27 dyne/cm 2 , assuming a rectangular channel of nominal dimensions, flow rate of 10 µL/min and dynamic viscosity of 0.94 cP [172,175,177] . As epithelial cells experience fluid flow velocities closer to that of the interstiti um in vivo , the low, intermittent shear imposed by the MfR is likely comparable to physiologically relevant conditions and can be further adjusted as needed [174] . Optical transparency and O 2 barrier properties of VeroClear are ideally suited for MfR, but its biocompatibility requires further studies. Cell suspension samples are exposed to VeroClear for 10 min in MfR , in contrast to adherent cells which are cultured in the proximity of the polymer for many hours. We used several strategies to prevent or reduce potential cytotoxic effects. First, the cells were seeded as a tight cluster in the center of the well with g lass base and polymer walls, separated from the bulk media by a seeding mask. The seeding mask was coated with polymethylmethacrylate to act as a barrier film between VeroClear and cells. This impeded mass transfer from the perimeter of the well and seques tered the cell cluster from interaction with potentially toxic leachates [116] . We have successfully cultured human epithelial (ARPE - 19, Figure 3. 7A) and bovine retinal endothelial cells ( Figure 3.1 1 ) in the well without the use of a mask, although further investigation of cytotoxicity is needed for printed materials. During adherent cell measurements, the polystyrene matrix of the PtOEP optode is unlikely to 66 interfere with cellular metabolism, or suffer from bio - fouling, and provides an additional barrier against potential leaching of toxic compounds from the underlying printed polymer [116] . 3D printing technology now enables the production of micron scale features in an array of polymers. High resolution and flexibility permitted the manufacture of most components with adequate precision and reproducibility without specialized tooling or casting. An alternative to 3D printing, micro - milling, creates a rougher surface that is prone to trapping air bubbles, which can affect oxygen measurements. In contrast, 3D printed parts have smooth surfaces when properly oriented with respect to the supporting plastics, improving the reproducibility of the measurements. A more powerful and flexible open shell design (Figure 3. 1C) would be very difficult to implement usin g a traditional manufacturing techniques. While 3D printing has been a major enabling technology for this study, yielding hundreds of chambers of varying design, methodologies can vary widely from a single prototype chip to industrial production. Common to all methodologies is the concept of an isolated microchannel with an aspect ratio amenable to bulk analyte measurement without interference from mass transfer. The Z - dimension in 3D printing typically has higher resolution than the X - and Y - dimensions a microns [178,179] . This is sufficient for optimal MfR channel depths of 70 - 150 m because shear stress scales linearly with flow rate and as the inverse square of the channel height. Further reduction of height woul d necessitate large reductions in flow rates to control shear of adherent cells. The increased cell aggregation would negatively affect reproducibility in suspension measurements. 3. 5 Conclusion This work describes the development of microfluidics - based r espirometry for studies on biological energy transduction. Taking advantage of remote sensing in an isolated microchannel, this simple, yet versatile, method can detect O 2 consumption by minute amounts of sample, ranging from soluble enzyme systems to cell or organelle suspensions and adherent samples. We demonstrated performance of this analytical 67 tool in the context of eukaryotic respiration, although it can be employed for measurements of bacterial and plant metabolism. A combination of low oxygen permea bility with flexible configuration allow for direct, uncompensated data acquisition, amenable for automation and development of fundamentally new experimental protocols for use in a wide array of basic and applied biomedical fields. 68 APPENDIX 69 3.6 Appendix Polymer Permeability (Barrer) Reported Poly(dimethylsiloxane) (PDMS) 610 Polyethylene (PE) 23 Poly(tetrafluoroethylene) (PTFE) 4.2 Poly(methyl methacrylate) (PMMA) 0.09 Polyetheretherketone (PEEK) 0.13 Measured PEEK 0.143 ± 0.001 VeroClear 0.125 ± 0.007 0.218 ± 0.006 Figure 3.10: Oxygen permeability of selected polymers. Barrer = 10 - 10 cm 3 cm cm - 2 s - 1 cmHg - 1 , mean ± - - measured at 37° C. Figure 3.11 . Biocompatibility of the MfR. Photomicrographs of BRECs grown on the MfR before (top) and after (bottom) ~25 hours of continuous measurement (right). The experiment was reproduced multiple times, with the continuous measurements longer than 24 hours without apparent loss of activity. 70 Figure 3.12 . Schematic of the MfR for adherent samples . Cells are cultured on the glass base of the well, open to the incubator atmosphere (Adhesion). Assembly and Measurement refer to sequential application of MF chip and manifold immediately prior to R O2 determination. 71 Chapter 4: Mitochondrial Ceramide Effects on the Retinal Pigment Epithelium in Diabetes 4. 1 Introduction Diabetic retinopathy is the leading cause of blindness among working - age adults, representing a large socioeconomic burden on society. To date, medical and surgical treatment options have been revolutionary; however, indications for treatment rely on advanced markers of disease. Development of effective treatments f or early stages of the disease require elucidation of the underlying biochemical pathophysiology. The retina is composed of a highly ordered and bioenergetically active neural tissue, perfused by two independent vasculatures. The retinal and choroidal vessels, supplying the inner and outer retina, respectively, regulate molecular exchange across the in ner and outer blood retinal barriers. Breakdown of these barriers results in clinically observable lesions, such as microaneurysms and hemorrhages, ultimately leading to retinal hypoxia or ischemia and disease progression [59] . The inner blood retinal barr ier, consisting of non - fenestrated retinal endothelial cells and pericytes, has been the focus of many studies, but the outer barrier has received comparatively less attention [55,59,61] . The choriocapillaris, a vascular layer supplying circulation to the outer third of the retina, consists of a fenestrated endothelium ) . The RPE provides a barrier function, with expression of tight junction proteins and regulation of transcellular water, ion, and metabolite transport by polarized expression of transporters [180,181] . Apart from regulating the osmotic and ionic balance of the outer retina, the RPE plays a key role in vision by phagocytosing shed photoreceptor outer segments and recycling r etinoids for the visual cycle [180,181] . Therefore, RPE dysfunction can contribute to the hypoxic conditions common in DR, and to the fluid and ion fluxes thought to cause diabetic macular edema [22,182,183] . Diabetic retinopathy is a neurovascular compli cation of diabetes resulting from chronic exposure to 72 hyperglycemia, dyslipidemia, and inflammation. Diabetic dyslipidemia leads to changes in systemic and local lipid metabolism that drive the pro - inflammatory and pro - apoptotic cellular changes typical of diabetic retinopathy (DR) [16] . Ratios of key sphingolipid species, such as ceramide and sphingosine - 1 - phosphate, are a major factor in sphingolipid metabolism and play key roles in cell fate [48] . These ratios their importance in determining cell growth, proliferation, and apoptosis. In particular, ceramides are bioactive sphingolipid species which regulate cell stress responses [53,184] . Ceramides can be synthesized de novo from serine and palmitate or salvaged from other sphingolipid species, depending on the physiological state of the tissue [47] . Structurally, ceramide is composed of a sphingoid base and an exchangeable fatty acid. The chain length of the latter determines the biological effect of the ceramid e. While short - chain ceramides (<20 carbons) are associated with pro - apoptotic effects, long - chain ceramides (>20 carbons) exert a protective effect on cells [43] . Ceramide is produced from sphingomyelin by hydrolysis of the phosphocholine head group, and the enzymes which catalyze this reaction, the sphingomyelinases, are distinguished by the pH at which they show optimum catalytic activity [47] . Acid sphingomyelinase (ASM) catalyzes sphingomyelin hydrolysis in lysosomes and at the plasma membrane, showing relative specificity for producing short - chain ceramides [54,185] . The ASM - knockout mouse has been well characterized as an animal model of Neimann Pick disease, showing remarkable resistance to cellular toxicity stemming from a variety of stressors such as hypoxia, radiation, and ischemia - reperfusion injury [52 54] . Specifically, the ASM - knockout mouse is resistant to retinal ischemia - reperfusion injury, confirming the central role of ceramide generation in the response to cell stress [54] . Studies in animal and cell culture models of DR have shown that it is the ASM, rather than the neutral SM, that is increased in the retina and retinal cells [54] . Moreover, inhibition of ceramide synthase, the central enzyme of the de novo ceramide produ ction pathway, had no effect on cytokine - induced pro - inflammatory changes in the retina and retinal cells [55] , further supporting the central role of ASM in 73 ceramide - mediated retinal pathology. Reports of direct effects of ceramide on mitochondrial struct ure and function [46,66,71,73,186,187] prompted us to consider whether diabetes - induced ASM upregulation might lead to mitochondrial ceramide accumulation and, in turn, to structural and functional changes. Overall changes in sphingolipid levels have been documented in the diabetic retina, but elevated ceramide levels were not evident. Instead, decreases in ceramide species were compensated with increases in hexosylceramides, consistent with an increase in ceramide glycosylation in diabetes [188] . In the cu rrent study, we examine diabetes - induced changes in retinal mitochondria - specific ceramide and demonstrated that changes in mitochondrial network structure and function occur in an ASM - dependent manner, in contrast to the sphingolipid changes in the whole retina. 4. 2 Methods 4. 2 .1 Rodents All animal procedures complied with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Procedures received prior approval by IACUC at Michigan State University, approval #Busik08/17 - 151 - 00, 28/08/2017. Diabetes was induced in male Sprague Dawley rats (237 283 g) with a single intraperitoneal injection of streptozotocin (STZ) (65 mg/kg) (Sigma Aldrich, St. Louis, MO, USA) dissolved in 100 mM citric acid (pH = 4.5) (29). Bod y weights and blood glucose were monitored biweekly. Blood glucose concentration was maintained in the 20 mM range. Rats were used 7 weeks after diabetes induction. C57BL/6J ASM - deficient (ASM ) male mice and littermate wild - type controls at 6 8 weeks of age were used in the study. 4. 2 .2 Cell Culture Primary human RPE were isolated according to standard procedures and cellular phenotype was 74 confirmed by staining for ZO - 1 and RPE65 markers [189] . ARPE - 19 (ATCC CRL - 2302) cells were grown in v / v ) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37 °C in 95% relative humidity and 5% CO 2 . Primary human RPE cells were used at Passages 4 8. 4. 2 . 3 Mitochondrial Isolation Mitochondria were isolated according to previously described protocols with minor modifications [190,191] . Briefly, cells were resuspended in ice - cold mitochondrial isolation buffer (mIB) and lysed for 20 s with a Scilogex D160 ho mogenizer (Scilogex, Rocky Hill, USA) equipped with a 5 mm diameter probe operated at 18,000 rpm. The homogenate was brought to 30 mL with fresh mIB and centrifuged at 1000× g for 10 min at 4 °C. The supernatant was reserved, and the pellet was homogenized and centrifuged as above. The pooled supernatants were centrifuged at 8000× g for 15 min, and the mitochondrial pellet was washed with fresh mIB and subjected to further processing as indicated. Where required, half of the isolated mitochondrial sample wa s further purified via sucrose step - density gradient ultracentrifugation without modifications using a Sorvall M120 SE Micro - Ultracentrifuge (S55S - 1155, ThermoFisher Scientific, Waltham, USA) [190] . 4. 2 .4 Mass Spectrometry Mitochondria on dry ice were sub jected to lipid extraction with chloroform, methanol, and water as previously described [192] . Dri ed lipid extracts were washed with 10 mM ammonium bicarbonate solution to remove salts and buffer contaminants, and then dried under a vacuum and resuspended in methanol by normalizing volumes to total mitochondrial protein. Immediately before analysis, mitochondrial lipids were diluted 5 - fold by drying aliquots in a speed - vac centrifuge and resuspending in five volumes of isopropanol/methanol/chloroform (4:2:1 , v : v : v ) containing 20 mM ammonium formate . Lipids were analyzed by high - resolution/accurate mass spectrometry and tandem mass spectrometry in positive - and 75 negative - ionization modes on an LTQ - Orbitrap Velos mass spectrometer (Thermo Scientific, Waltham, MA, USA) [192] . A TriVersa Nanomate (Advion , Ithaca, NY, USA) functioned as a nano - electrospray ionization source and autosampler. The nESI spray voltage was held at 2.4 kV and nESI gas pressure was 0.3 psi. The 96 well sample plate (Eppendorf, Hamburg, Germany) was held at 12 °C. Sphingolipid spec ies were quantified as their formate adducts in negative - ionization mode against spiked synthetic sphingolipid internal standards of Cer(30:1) and SM(30:1) (Avanti Polar Lipids, Alabaster, AL, USA) at 250 femtomole/microliter [193] . Sphingolipid structures were confirmed by higher - energy collisional dissociation MS/MS in positive ionization mode. Each mass spectrum was subjected to offline mass recalibration using Thermo Xcalibur software to correct for any instrumental drift in mass calibration. Lipid peak s were subjected to isotope correction, identified, and quantified against sphingolipid internal standards using LIMSA software [194] as previously described [192] . 4. 2 .5 Immunocytochemistry and Mitochondrial Morphology Cells were washed three times with PBS and fixed for 15 min at room temperature with Histochoice fixative (Sigma, cat no. H2904). Cells were permeabilized with 0.1% Triton X - 100 in PBS for 20 min and blocked with 1.5% BSA, 1% Tween - 10 in PBS (PBST) overnight at 4 °C. Blocked samples were in cubated with anti - ceramide antibody (Sigma, cat no. 8104) at a 1:100 dilution at 4 °C overnight. After three washes with PBST, cells were incubated with anti - mouse secondary antibody conjugated to Alexafluor 488 at a 1:100 dilution. Cells were counterstain ed with DAPI and imaged on a Nikon Eclipse TE2000 (Nikon Instruments Inc., Melville, NY, USA) equipped with a Photometrics CoolSNAP HQ2 camera (Photometrics, Tucson, AZ, USA). Fluorescence intensity was quantified with ImageJ software (version 1.53a, Natio nal Institutes of Health, Bethesda, MD, USA). For mitochondrial morphology determination, primary human RPE were grown on coverslips and stained with 50 nM MitoTracker Green at 37 °C for 30 min. After washing with PBS, cells were imaged on a Ziess LSM880 m icroscope (Zeiss, Oberkochen, Germany). 76 Mitochondrial length, as a marker of fragmentation, was determined by measuring the major axis of from three fields of view were selected from each sample to represent the cellular population. 3D animations of the z - stacked images were created use the 3D projection command in ImageJ, setting layer 4. 2 .6 Quantitative R eal - Time Polymerase Chain Reaction Total cellular RNA extraction and RT - PCR were performed as previously described [54] . Human gene - - - 6), intercellular adhesion molecule (ICAM1), and vascular endothelial growth factor (VEGF) were used to measure gene expression. Results were normalized to cyclophilin A. 4. 2 .7 Western Blot Analysis Protein extraction and Western blots were carried out using the NuPAGE system as pr eviously described [55] . Fractions of mitochondrial isolates were normalized by suspension volume and quantitated relative to voltage - dependent anion channel (VDAC) intensity. Primary antibodies against LAMP - 1 (SC - 20011, Santa Cruz, Dallas, TX, USA), VDAC (PAI - 954A, Invitrogen, Waltham, MA, USA), and ASM were used at 1:1000 dilution. Anti - ASM antibody was a generous gift from Richard Kolesnick. Secondary antibodies against rabbit IgG (926 - 68073, Odyssey, Lincoln, NE, USA) and mouse IgG (610 - 731 - 124, Rocklan d Immunochemicals, Limerick, PA, USA) were used at 1:10,000 dilution. Bands were imaged on a LiCor Odyssey imaging system. Densitometric analysis was performed in ImageJ software after splitting the RGB image into individual channels and a background subtr action using a rolling ball radius of 16.3 pixels. 77 4. 2 .8 Citrate Synthase Activity Enzymatic activity was measured using a citrate synthase activity assay kit (Sigma, cat no. CS0720) hase activity was normalized to total protein content, measured using the Bradford assay (BioRad, Hercules, CA, USA). 4. 2 .9 Microrespirometry Cellular respirometry was measured as previously described [195] . Briefly, cells were seeded on - chip at a density of 860 cells/mm 2 and cultured overnight under standard cell culture conditions at 37 °C, 95% relative humidity, 5% CO 2 . Microrespirometer chips were assembled immediately before the measurement and the cells were perfused with the basal respiration buffer , which consisted of DPBS with calcium and magnesium supplemented with 10 mM glucose, 10 mM lactate, 1 mM pyruvate, and 0.2% bovine serum albumin. Respiratory control ratio [77] was determined from sequential respiration measurements in the leak buffer, co - chlorophenylhydrazone (CCCP). Measurements in inhibition buffer, consisting of uncoupling buffe r supplemented with 5 mM potassium cyanide, were used for correction of non - respiratory oxygen consumption. Perfusion was controlled using a syringe infusion pump (KD Scientific, Holliston, MA, USA) . Activity determinations were performed under stationary buffer conditions for 5 4. 3 Results 4. 3 .1 Diabetes Results in Retinal Mitochondrial Ceramide Accumulation We used an streptozotocin (STZ) - in duced diabetic rat model to determine whether upregulation of ASM expression and activity in cells comprising the inner and outer blood retinal barriers (BRBs) [54] 78 contributed to the mitochondrial ceramide accumulation that further leads to cell damage [196] . Mitochondria were isolated from control and diabetic rat retinas using differential centrifugation protocols, followed by lipid extraction using c hloroform, methanol, and water [192] and Orbitrap high - resolution/accurate mass mass spectrometry (MS) and MS/MS analysis. Samples were normalized based Figure 4.1. Negative - ion high - resolution/accurate mass spectrometric quantification of sphingolipids in retinal mitochondria. Mass spectra of control ( A ) and diabetic ( B ) rat retinal mitochondrial sphingolipids after 7 weeks of diabetes. Full - scan MS spectra are shown from mitochondrial lipids analyzed by negative - ionizatio n mode direct - infusion nano - ESI mass spectrometry. Sphingolipids are shown under magnifications indicated at the top of each panel. Abundant non - C ) Quantification of total sph ingomyelin (SM), ceramide (Cer), and the Cer/SM ratio based on mass spectrometry analysis of mitochondria from control and diabetic rat retinas (left panel), and wild type (WT) and acid sphingomyelinase knock out (ASMKO) mouse retinas (right panel). * p < 0.05, n = 3. 79 on total mitochondrial protein, and sphin - golipid peaks were compared to synthetic sphingolipid internal standards incorporated in each run. As presented in Figure 4.1A,B, diabetes resulted in a decrease in endogenous levels of mito - chondrial ceramide and sphin - gomyelin, consistent with previous reports [188] . To quantify changes in sphingolipid composition, total detected sphin - golipid abundances were summed, and sphingolipid species were expressed as a percentage of total sphingolipids. This approach revealed significant increases in relative ceramide l evels and decreases in the relative sphingomyelin levels in retinal mitochondria isolated from STZ - induced diabetic rat retinas (7 - week duration) compared to controls (Figure 4.1C, left). To test the role of ASM in the control of mitochondrial ceramide mor e directly, sphingolipid profiles of mitochondria prepared from ASM - knockout mice were similarly analyzed. In contrast to the diabetes - induced increase in the ceramide - to - sphingomyelin ratio, depletion of ASM resulted in lower relative levels of ceramide v ersus sphingomyelin compared to wild - type controls (Figure 4.1C, right), confirming that ASM plays an important role in mitochondrial sphingolipid dynamics. 4. 3 .2 Diabetes Results in Pro - Inflammatory Changes in Human Retinal Pigment Epithelial (RPE) Cells Whole - retina preparations, as shown in Figure 4.1, lack the RPE layer, a site of diabetes - induced ASM upregulation [54] . We therefore sought to determine sep - arately whether RPE cells demonstrated similar diabetes - induced changes. Results of fluorescent c eramide staining in control - and diabetic - derived cultured human RPE cells are presented in Figure 4.2A and demonstrated an average 2.7 - fold increase in cellular ceramide staining of diabetic - derived RPE cells compared to controls (Figure 4.2B). Analysis o f inflammatory gene expression expression in diabetic - derived RPE cells compared to controls (Figure 4.2C), which was consistent with the 80 increases in ceramide observed by immunohistochemistry. Furthermore, in vitro treatment of control - derived RPE cells with 25 mM glucose for 72 h led to significant increases in ASM, VEGF, and ICAM1 mRNA compared to untreated control - derived RPE cells (Figure 4.2D), supporting their roles in hyperglycemic response. 4. 3 .3 Diabetes Results in Mitochondrial Fragmentation in Human RPE Cells As mitochondria are known to accumulate ceramide [63] , and we demonstrated that diabetes changes the ceramide - to - sphingomyelin ratio in retinal mitochondria (Figure 4.1), we next sought to determine whether structural and functional changes coul d be detected in mitochondria isolated from control - or diabetic - derived RPE cells. Figure 4.3A demonstrates staining with MitoTracker Green, used to reveal the expected reticular mitochondrial network in the control RPE cells. This network appeared to be disrupted in the diabetic - derived RPE cells, which had predominantly round and fragmented mitochondria. Quantitation of morphological features revealed that the average mitochondrial length in diabetic - Figure 4.2. Diabetes - induced pro - inflammatory changes in human RPE ( A ) Representative images of control - and diabetic - derived retinal pigment epithelial (RPE) cells showing ceramide (green) and B ) Quantitation of ceramide - staining fluorescence intensity from panel (A). n = 9, error bars = S.D., * p < 0.05; ( C ) Inflamm atory gene expression in diabetic - derived RPE (black bars) compared to control (white bars); ( D ) Upregulation of inflammatory gene expression in control RPE treated with 25 mM glucose for 72 h (black bars) compared to untreated cells (white bars). * p < 0. 01, n = 6. 81 n = 3), whereas c ontrol - n = 3) in length, (Figure 4.3B). 4. 3 .4 Diabetes Induces Ac id Sphingomyelinase (ASM) - Mediated Changes in Mitochondrial Function of Human RPE Cells To determine whether the structural changes of mitochondria were correlated with detectable functional differences, we used micro - respirometry to examine oxidative phos phorylation in control - and diabetic - derived RPE cells [195] . In this approach, the flow of oxygenated medium over adherent cells is intermittently stopped and respiration leads to a steady consumption of oxygen, seen as periodic downward slopes in the O 2 concentration traces (Figure 4.4A). Ensuing resumption of flow reoxygenates the sample and the measurement is repeated. Following growth to confluency, RPE cells were transferred into the micro - respirometer and perfused with a medium containing glucose, l actate, and pyruvate as substrates, supplemented with the ATP - synthase inhibitor oligomycin (Figure 4.4A, leak). Respiratory activity in this state is limited by a high proton motive force and predominantly represents proton leakage through the inner mitoc hondrial membrane [81] . No substantial differences between sample groups were observed in this state, suggesting a lack of diabetes - induced changes to inner mitochondrial membrane Figure 4.3. Structural analysis of human RPE mitochondria. ( A ) Mitochondrial morphology determined by MitoTracker Green staining of control - and diabetic - derived RPE. Inset = skeletonized (green lines) binary mask (purple) of deconvoluted photomicrographs highlighting mitochondrial morphology. Scale B ) Quantitation of average mitochondrial length. n = 3, * p < 0.05. 82 proton permeability. Next, the maximal respirator y rate was assessed by dissipation of the proton motive force with the chemical uncoupler, carbonyl cyanide m - chlorophenylhydrazone (CCCP, uncoupled). In this state, control over respiration is shifted to substrate delivery pathways and innate turnover cap acity of the electron transport chain. Dissipation of the proton motive force with CCCP resulted in an increase in oxygen consumption rates of RPE cells over that observed with oligomycin alone. While the expected increase in respiration due to uncoupling was observed in control RPE cells and in diabetic cells with desipramine pretreatment, little response to the uncoupler was observed in the diabetic cells without desipramine pretreatment. The latter observation indicated that mitochondria in resting diabe tic RPE cells operate close to the maximal respiratory activity, which is limited by electron transport chain turnover or substrate delivery. Subsequent perfusion with potassium cyanide resulted in complete inhibition of mitochondria - dependent oxygen consu mption, evident in all experimental groups and used to correct for non - mitochondrial oxygen - consuming processes. The relative changes in the oxygen consumption between three conditions were used to calculate a respiratory control ratio (RCR, Equation (1)), a quantitative measure of mitochondrial fitness [77] . Figure 4.4. Microrespirometric analysis of human RPE cells. ( A ) Representative oxygen concentration traces of control (black), diabetic (blue), and desipramine - pretreated diabetic (red) RPE cells. Respirometry was performed in the presence of oligomycin (leak), carbonyl cyanide m - chlorophenylhydrazone (CCCP, uncoupled) , and potassium cyanide (KCN, inhibited). ( B ) Respiratory control ratios of control (white circles), diabetic (black circles), and desipramine - pretreated diabetic groups (gray circles). * p < 0.05, n = 3 4. 83 ( 1 ) As shown in Figure 4.4B, diabetic - derived RPE cel ls displayed a significantly decreased RCR compared to control - derived RPE cells (1.41 ± 0.27 vs. 3.78 ± 0.59). This difference was abolished by perfusion of diabetic - derived RPE cells with 15 µM desipramine, an ASM inhibitor, [197,198] for 1 h, which increased the RCR to 5.00 ± 1.78. 4. 3 .5 Mitochondrial ASM Contributes to Impaired Mitochondrial Function In Vitro Accumulation of ceramide at the expense of sphingomyelin with concomitant change in ceramide/sphingomyelin ratio ( see above ) suggests that it is the result of sphingomyelin hydrolysis, a reaction catalyzed by ASM. This raises the question of whether ceramide accumulation in mitochondria is due to the activity of mitochondrial ASM or to the transport of ceramide from remote sites . The presence of intrinsic ASM in mitochondrial membranes is demonstrated in Figure 4.5. In addition to mitochondria, ASM is known to be present in the lysosomes and the plasma membrane. Although a high degree of separation between the mitochondrial and p lasma membrane fractions is easily achievable by standard methodology, the lysosomes and mitochondria and are much harder to separate due to very similar size, shape, and density characteristics [191,199 201] . To conclusively localize ASM to mitochondrial membranes, we obtained mitochondrial preparations with increasing levels of purity from human RPE cells and subjected them to immunoblotting for (i) ASM; (ii) the mitochondrial outer membrane marker voltage - dependent anion channel (VDAC); and (iii) the lys osomal membrane marker lysosomal - associated membrane protein 1 (LAMP - 1). As shown in Figure 4.5A, mitochondria prepared by standard differential centrifugation protocols showed the presence of VDAC and LAMP - 1 (crude), indicating co - purification of lysosome s in mitochondrial preparations. However, further purification by successive centrifugations at 8000× g (pure) resulted in significant depletion of the lysosomal marker LAMP - 1 relative to the 84 mitochondrial marker VDAC. Each fraction was sampled at varying concentrations, ensuring fidelity of optical density quantitation, which revealed that extra purification yielded a two - fold depletion of LAMP - 1 (Figure 4.5B, left panel). The same samples were then probed for the level of ASM, which is known to localize t o the lysosome. If most of the ASM shown in Figure 4.5 originated from the lysosomal compartment, the ratio of ASM to VDAC would follow that of LAMP - 1. The experimental results demonstrated only a small ASM depletion in the purified mitochondrial sample, a nd optical density quantitation revealed an apparent enrich - ment of ASM when normalized to VDAC density (Figure 4.5B, right panel). These results were replicated in mitochondria isolated from ARPE - 19 using a similar protocol (Figures 4.5C). In this case, the mitochondr ial fractions were examined by immunoblotting after initial differential centrifugation (D.C.) and again after a second sucrose - step density Figure 4.5. Colocalization between ASM and mitochondrial markers. ( A ) Western blot analysis of human RPE cell mitochondria at varying levels of purity and ( B ) associated optical density quantitation; ( C ) Western blot analysis of ARPE - 19 cell mitochondria at varying purity levels and associated optical density quantitation ( D ). LAMP - 1 = lysosome associated membrane protein 1, VDAC = voltage dependent anion channel, crude = one 8000× g centrifugation, pure = three 8000× g centrifugations, D.C. = differential centrifugation, U.C. = sucrose step - density ultracentrifugation. 85 ultracentrifugation (U.C.). Consistent with the results from the human RPE cells (Figure 4.5B), the ultracentrifugation resulted i n a two - fold depletion of LAMP - 1 normalized to VDAC (Figure 4.5D, left panel) compared to the differential centrifugation preparation of mitochondria. ASM enrichment was also observed after ultracentrifugation when normalized to VDAC in ARPE - 19 samples (Fi gure 4.5D, right panel). To examine the consequences of increased mitochondrial ASM expression and the resulting mitochondrial ceramide accumulation, we assessed changes in citrate synthase activity in response to high glucose treatment of ARPE - 19 (Figure 4.6). Citrate synthase resides exclusively in the mitochondrial matrix and catalyzes the condensation of acetyl - CoA and oxaloacetate to citrate, in the first step of the tricarboxylic acid cycle. As such, it is widely used as a mitochondrial content marker [202] . The effect of glucose concentration on citrate synthase activity in ARPE - 19 cells is presented in Figure 4.6. No statistically significant effect of high glucose was observed after 24 h ( n = 5, p > 0.05). At 48 h, the 25 mM glucose treatment increased citrate synthase activity to 164.3% ± 5.8% vs. 5.5 mM glucose control, followed by the reduction in activity to 75% ± 1.6% of the control at 72 h ( p < 0.05, n = 5). To evaluate whether these glucose - induced changes in citrate synthase activity was mediated by ASM, parallel measurements were conducted on ARPE - 19 cells incubated in high - glucose conditions with daily, intermittent treatments with 15 µM desipramine. These treatments abolished the biphasic Figure 4.6. Citrate synthase activity in ARPE - 19 cells. White bars = control cells, black bars = 25 mM glucos e - treated cells, gray bars = 25 mM glucose - treated cells with daily 1 h treatment with 15 µM desipramine. * p < 0.05, n = 5. 86 hyperglycemia - induced response in citrate synthase activity, and the desipramine - treated group displayed a time profile closely corresponding to that of the control group ( p > 0.05, n = 5). 4. 4 Discussion Diabetes is a multifactorial pathological process resulting in micro - and macrovascular complications. Diabetic retinopathy is a common microvascular complication of diabetes, which results from hyperglycemia, dyslipidemia, and chronic inflammatory changes in the retina leading to blood retinal barrier br eakdown and disease progression [203] . Diabetic dyslipidemia results in both systemic and local changes to lipid metabolism and, in the retina, contributes to the pro - apoptotic changes seen in the inner and outer blood retinal barrier cellular components [ 16] . We previously demonstrated that diabetes leads to enhanced ASM expression predominantly in retinal endothelial - and retinal pigment epithelial cells [54] suggesting that sphingomyelin hydrolysis is the primary cause of cellular ceramide accumulation. Despite these findings, measurements of sphingolipid composition in diabetic rodent retinas revealed that ceramide levels are, in fact, decreased whereas glucosylceramides are increased [188] . Such results suggest hyperglycemia - induced diversion of cerami de toward the glycosylated forms in total retinal sphingolipid pools. The increased glycosylation in the diabetic retina was attributed to increases in uridine diphosphate glucose (UDP - glucose) production through the pentose phosphate pathway, rather than changes in enzymatic activity. As the pentose phosphate pathway occurs in the cytoplasm, we argue that an increase in glucosylceramide production due to higher UDP - glucose availability would be limited to the cytoplasm, rather than the mitochondria. In con trast to whole - retina sphingolipid measurements, we show in Figure 4.1A that diabetes - induced increases in relative ceramide levels can be detected in mitochondria after subcellular fractionation of whole retina. Similarly, mitochondria isolated from the r etinas of ASM - knockout animals displayed an inversion of the ceramide - to - sphingomyelin ratio (Figure 4.1B), demonstrating the direct connection 87 between ASM activity and mitochondrial ceramide accumulation. It is worth mentioning that the degree of these ch anges in the barrier cells was likely underestimated because the major component of the mitochondrial preparations from whole retina originate from photoreceptors. Diabetes - induced increase in ASM expression and activity is the highest in the cells that ma ke up the BRB, namely REC and RPE cells [54,55,185] . Smaller changes are observed in the Muller cells and microglia, and no changes are seen in the photoreceptors [54] . The changes in the whole - retina mitochondria preparations were thus diluted by the larg e population of the non - changing photoreceptor mitochondria and by the mitochondrial ceramide from endothelial, Muller, and microglia cells, which shows smaller changes. Here, we focused on the role of ASM - dependent sphingomyelin hydrolysis in barrier cell s. ASM - dependent mitochondrial ceramide accumulation is strongly supported by our present finding that a population of cellular ASM can be localized to mitochondrial membranes in RPE cells (Figure 4.5). These results are consistent with our previous report s that diabetes - induced ASM upregulation is a key player in blood retinal barrier breakdown and provide evidence for a proposed mechanism of metabolic dysfunction in retinal cells mediated by the accumulation of cellular ceramide. Although our results supp ort the role of mitochondrial ASM in the observed changes, we cannot presently rule out the contributions of alternative pathways, such as neutral - sphingomyelinase - and/or reverse - ceramidase - mediated mitochondrial ceramide generation, as described in other systems [63,196] . Our previous data show that neutral sphingomyelinase expression does not change in the diabetic retina [54] , and that inhibition of ceramide synthase has no effect on cytokine - induced pro - inflammatory changes in the retina and retinal ce lls [55] . In combination with the sensitivity of both the diabetes - induced changes in RCR and citrate synthase to desipramine reported here, we strongly argue that ASM plays a key role in RPE cell mitochondrial dysfunction. Alternatively, stress - induced pr oduction and transport of ceramide to mitochondria from distal sites has also been reported [204] . 88 In this work we focused on the RPE cells, a cellular component of the outer blood retinal barrier. As RPE cells were not a part of the whole - retina preparati ons, we examined mitochondria from the control and diabetic donors RPE cells separately from the whole - retina mitochondria. Human RPE cell culture could be used due to a well - known metabolic memory legacy effect. Metabolic memory was first described in dia betic patients as a prolonged effect of early glycemic control on the development of diabetic complications, even after glycemic control is established later in the course of disease progression [205] . The metabolic memory phenomenon is well accepted in th e field of diabetic complications [205 208] . The molecular mechanisms underpinning sustained metabolic memory are not fully understood. Recent work, however, has demonstrated that epigenetic modifications to mtDNA mismatch - repair machinery result in decrea sed transcript levels, decreased mitochondrial localization, and accumulation of mtDNA mutations [38]. As mtDNA is particularly vulnerable to ROS - induced DNA mutations, decreased functioning of repair machinery results in accumulation of damaged oxidative phosphorylation complexes and, ultimately, impairment of oxidative phosphorylation as a whole [39]. As these changes accumulate over time, they perpetuate a vicious cycle of oxidative stress and sustained inflammatory changes, leading to the progression of diabetic complications despite correction of diabetic hyperglycemia. These effects have been shown to occur in animal models as well as in cell culture models. Cells isolated from diabetic donor retinas or animal models retain their diabetic metabolic phe notype for several passages [205 210] . Human control and diabetic donor cells were previously shown to display metabolic memory characteristics right after the isolation and for up to eight passages [205] . Dysfunction of these cells is implicated in the de velopment of diabetic macular edema and they represent a site of significant diabetes - induced ASM upregulation [22,54] . Despite culturing control - and diabetic - derived RPE cells under identical, euglycemic, conditions, we detected increased ceramide and in flammatory gene expression in diabetic - derived RPE cells compared to controls (Figure 4.2). Furthermore, we showed that control - derived RPE cells retained their sensitivity to the diabetic milieu, as high - glucose treatment 89 resulted in enhanced inflammatory gene expression (Figure 4.2D). Accompanying these changes, diabetic - derived RPE cells displayed fragmented mitochondria and impaired mitochondrial - dependent metabolism (Figures 4.3 and 4.4). These results support the metabolic memory hypothesis, implying that diabetes induces permanent changes to cellular metabolism in the long term, despite achievement of a euglycemic state. Our results are consistent with previous reports detailing diabetes - induced mitochondrial fragmentation and impaired oxidative phosp horylation in retinal endothelial cells [211] , although mitochondrial fragmentation alone is insufficient to universally predict dysfunctional metabolism. The observation of diabetes - induced mitochondrial fragmentation was rationalized by the critical find ing of functional changes to oxidative phosphorylation (Figures 4.3, 4.4, and 4.6). Citrate synthase is a marker of mitochondrial content and its activity parallels electron transport chain capacity of the cell [202] . Steady - state mitochondrial content, ho wever, is controlled by the relative flux of mitochondrial biogenesis and mitophagy which are, in turn, related to mitochondrial fission and fusion dynamics. This quality - control mechanism is useful to clear bioenergetically dysfunctional mitochondria by f ission and subsequent mitophagy. It ensures a steady - state population of robust mitochondria capable of sustaining ATP synthesis rates over a wide range of metabolic demands. Indeed, diabetes - induced increases in mitophagy have been described in RPE cells with increased mitophagic flux attributed to ROS - dependent mitochondrial damage [58,212] . Our observations of diabetes - induced oxidative phosphorylation dysfunction (Figure 4.4), likely followed mitochondrial fragmentation (Figure 4.3), rationalize earlier reports of the increased mitophagic flux. Our data showed that diabetes - induced ASM upregulation led to an accumulation of ceramide in mitochondrial membranes that limits the maximal metabolic capacity of the respiratory chain. Combined with excessive ele ctron supply from glucose and adequate oxygenation in the hyperglycemic stage of the diabetes, such a restriction stimulates ROS production. This metabolic insult then leads to the production of dysfunctional mitochondria, which stimulates the 90 mitophagy pa thway and, under continuously elevated ASM levels, results in a steady population of fragmented and bioenergetically impaired mitochondria in RPE cells. Our current finding that desipramine treatment can rescue functional (RCR, Figure 4.4) and mitochondrial mass (Figure 4.6) changes in diabetic - derived RPE cells demonstrates that ASM - dependent ceramide metabolism plays a central role in diabetes - induced mi tochondrial damage. Desipramine belongs to a class of antidepressants known as tricyclic amines which are functional inhibitors of ASM activity [197,198] . Although there are reports that at a high dose, desipramine can also interact with mitochondrial prot eins directly, leading to impaired NADH oxidation, electron transport, and ATP synthase activity [213] daily) use in this study. Indeed, our results showed no changes of the basal rate (Figure 4.7) and substantial enhancement of the maximal oxidative phosphorylation function, which were not consistent with direct effects of desipramine on mitochondrial oxidative phosphorylation machinery. The observed changes rather sup port the effect of desipramine via inhibition of ASM activity, leading to the depletion of mitochondrial ceramide, reversing its inhibitory effect on the oxidative phosphorylation and increasing RCR, as described here. As a gross measure of mitochondrial fitness, the whole - cell RCR is sensitive to a range of metabolic processes including substrate delivery, maximal electron transport chain capacity, proton leakage, and outer mitochondrial membrane integrity [77,123] . A greater than two - fold decrease in the whole - cell RCR of diabetic - derived RPE cells suggests substantial impairment of oxidative phosphorylation with a concomitant decrease in mitochondrial ATP - synthesis capacity and increase in mitochondrial ROS generation. It is remarkable that the mitochond rial functional impairment was retained despite culturing the cells for several generations under standard, euglycemic conditions. Whether the diabetes - induced RCR changes arise from direct inhibition of substrate delivery, electron transport, or the phosp horylation system, such as direct ceramide inhibition of Complex III or ceramide - mediated formation of outer 91 mitochondrial membrane pores [71,73,186,187] , is the subject of ongoing research. The sensitivity of diabetic - derived RPE cells to desipramine stro ngly suggests ASM - mediated ceramide inhibition of oxidative phosphorylation. Depletion of mitochondrial ceramide upon ASM inhibition, therefore, would reverse these effects, reduce oxidative stress, and favor retention of a robust mitochondrial population, as we saw in this work (Figures 4.4 and 4.6). 92 Chapter 5: Conclusions and Future Directions 5. 1 Conclusions and Future Directions Respirometry is a powerful tool that is well suited as a screen for potential metabolic derangements and to perform highly detailed studies on specific segments of oxidative metabolism. Development of a flow - through microrespirometer, as described in Chapter 3, enables b ioenergetic characterizations of a variety of biological samples by significantly increasing sensitivity, permitting varied sample preparations , integrating into standard cell culture practices, and by executing novel experimental protocols. Despite these advantages, data presented in Chapter 3 indicate that cell suspension measurements suffer from large variances. As these variances were attributed to handling of dense cell suspensions and sedimentation of cells along the inlet and microchannel during load ing , implementation of flow focusing can be explored to remediate these effects. Additionally, multiplexing the oxygen sensing with reactive oxygen species quantitations and/or in - line glucose sensors can yield complementary information about diabetes - indu ced metabolic changes in whole cells or tissues . The data presented in Chapter 4 show that diabetic - derived retinal pigment epithelial cells display a significantly decreased respiratory control ratios compared to control - derived cells , suggesting that dia betes results in derangements in oxidative metabolism . T hese changes are additionally sensitive to the acid sphingomyelinase inhibitor desipramine, suggesting that acid sphingomyelinase - dependent mitochondrial ceramide accumulation impairs oxidative metabo lism. Ceramides are well known pro - apoptotic sphingolipids and have been reported to inhibit complex III activity and induce mitochondrial outer membrane permeabilization (see Chapter 1) . Application of the concepts described in Chapter 2 to the se results will aid in l ocalizing the functional consequences of acid - sphingomyelinase - dependent mitochondrial ceramide generation . Proceeding with this model system would therefore involve plasma membrane permeabilization of cell culture monolayers 93 with subsequent a ssessment for exogenous cytochrome c sensitivity and maximal ETC turnover . Results of these studies can resolve ceramide - mediated inhibition of ETC from induction of cytochrome c release. Decreases in maximal ETC turnover suggest flux limitations upstream of the pmf, including complex III as a candidate, whereas exogenous cytochrome c sensitivity suggests an increasing population of apoptotic cells in treatment conditions. Flux limitations in ETC turnover can then be further probed to determine complex - or substrate - specific respiratory rates to localize the metabolic dysfunction , as described in Chapter 2 . Complementary to the functional studies, assessment of diabetes - induced mitochondrial fragmentation (Chapter 4) in the presence of ASMase inhibition can be used to show downstream effects of ASMase - induced mitochondrial dysfunction. Ceramide accumulation can play a role in ROS generation, through ETC inhibition, or in mitochondrial fission due to its role in regulation of membrane biophysical properties . Reports of mitochondrial fragmentation and increased flux through mitophagy suggest that diabetes - induced mitochondrial dysfunction stimulates organelle removal pathways and ceramides may p lay a significant role in the process. Finally, l ocalization of ASMase to mitochondrial membranes by immunoblotting is also reported in Chapter 4. Similar localization patterns were seen using primary cells and ARPE - 19, a cell line, strongly suggesting that a portion of ASMase is localized to mitochondria. To extend these re sults, quantitative assays for ASMase activity can confirm presence of active ASMase in mitochondria and further to probe for treatment - induced changes to mitochondrial ASMase activity. Electron microscopy can help confirm localization to mitochondrial mem branes and reveal whether the detected ASMase localizes to intra - organelle contact points. 94 REFERENCES 95 REFERENCES 1. Diabetes 2014 Report Card ; 2014; 2. National Diabetes Statistics Report ; Atlanta, 2020; 3. Chen, R.; Ovbiagele, B.; Feng, W. Diabetes and Stroke: Epidemiology, Pathophysiology, Pharmaceuticals and Outcomes. Am. J. Med. Sci. 2016 , 351 , 380 386, doi:10.1016/j.physbeh.2017.03.040. 4. Fong, D.S.; Aiello, L.P.; Ferris, F.L.; Klein, R. Diabetic retin opathy. Diabetes Care 2004 , 27 , 2540 2553, doi:10.2337/diacare.27.10.2540. 5. Cheung, N.; Mitchell, P.; Wong, T.Y. Diabetic Retinopathy. Lancet 2010 , 376 , 124 136, doi:10.1016/S0140 - 6736(09)62124 - 3. 6. Bhavsar, A.R. Diabetic Retinopathy 2019. 7. Solomon, S.D.; Chew, E.; Duh, E.J.; Sobrin, L.; Sun, J.K.; VanderBeek, B.L.; Wykoff, C.C.; Gardner, T.W. Diabetic retinopathy: A position statement by the American Diabetes Association. Diabetes Care 2017 , 40 , 412 418, doi:10.2337/dc16 - 2641. 8. Wright, A. D.; Dodson, P.M. Medical management of diabetic retinopathy: fenofibrate and ACCORD Eye studies. Eye 2011 , 25 , 843 9, doi:10.1038/eye.2011.62. 9. Simó, R.; Hernández, C. Prevention and treatment of diabetic retinopathy: evidence from large, randomized tri als. The emerging role of fenofibrate. Rev. Recent Clin. Trials 2012 , 7 , 71 80. 10. Nabulsi, A.; Folsom, A.; White, A.; Patsch, W.; Heiss, G.; Wu, K.; Szklo, M. The Effect of Intensive Treatment of Diabetes on the Development and Progression of Long - Term Complications in Insulin - Dependent Diabetes Mellitus. N. Engl. J. Med. 1993 , 328 , 1069 1075. 11. Lyons, T.J.; Jenkins, A.J.; Zheng, D.; Lackland, D.T.; McGee, D.; Garvey, W.T.; Klein, R.L. Diabetic retinopathy and serum lipoprotein subclasses in the DCCT/ EDIC cohort. Invest. Ophthalmol. Vis. Sci. 2004 , 45 , 910 8. 12. Tse, D.; Williamson, E.; et al. Effect of fenofibrate on the need for laser treatment for diabetic retinopathy (FIELD study): a randomised controlled trial. Lancet 2007 , 370 , 1687 1697, doi:10.1016/S0140 - 6736(07)61607 - 9. 13. Group, A. to C.C.R. in D. (ACCORD) S. Effects of Intensive Glucose Lowering in Type 2 Diabetes. N. Engl. J. Med. 2008 , 3 58 , 225 237. 14. Baldeweg, S.E.; Yudkin, J.S. Implications of the United Kingdom prospective diabetes study. Diabetes Journals 2002 , 25 , doi:10.1016/S0095 - 4543(05)70132 - 9. 96 15. Nathan, D.M. The diabetes control and complications trial/epidemiology of diabetes interventions and complications study at 30 years: Overview. Diabetes Care 2014 , 37 , 9 16, doi:10.2337/dc13 - 2112. 16. Hammer, S.S.; Busik, J. V. The role of dyslipidemia in diabetic retinopathy. Vision Res. 2017 , 139 , 228 236, doi:10.1016/j.visres.2017.04.010. 17. Rübsam, A.; Parikh, S.; Fort, P.E. Role of inflammation in diabetic retinopathy. Int. J. Mol. Sci. 2018 , 19 , 1 31, doi:10.3390/ijms19040942. 18. Chew, E.Y.; Davi s, M.D.; Danis, R.P.; Lovato, J.F.; Perdue, L.H.; Greven, C.; Genuth, S.; Goff, D.C.; Leiter, L.A.; Ismail - beigi, F.; et al. The Effects of Medical Management on the Progression of Diabetic Retinopathy in Persons with Type 2 Diabetes: The ACCORD Eye Study. Ophthalmology 2014 , 121 , 2443 2451, doi:10.1016/j.ophtha.2014.07.019.The. 19. Preliminary Report on Effects of Photocoagulation Therapy. Am. J. Ophthalmol. 1976 , 81 , 383 396, doi:10.1016/0002 - 9394(76)90292 - 0. 20. Early Photocoagulation for Diabetic Reti nopathy: ETDRS Report Number 9. Ophthalmology 1991 , 98 , 766 785, doi:10.1016/S0161 - 6420(13)38011 - 7. 21. Deschler, E.K.; Sun, J.K.; Silva, P.S. Side - effects and complications of laser treatment in diabetic retinal disease. Semin. Ophthalmol. 2014 , 29 , 290 300, doi:10.3109/08820538.2014.959198. 22. Romero - Aroca, P.; Baget - Bernaldiz, M.; Pareja - Rios, A.; Lopez - Galvez, M.; Navarro - Gil, R.; Verges, R. Diabetic Macular Edema Pathophysiology: Vasogenic versus Inflammatory. J. Diabetes Res. 2016 , 2016 , doi:10.1155/2016/2156273. 23. Antonetti, D.A.; Barber, A.J.; Bronson, S.K.; Freeman, W.M.; Gardner, T.W.; Jefferson, L.S.; Kester, M.; Kimball, S.R.; Krady, J.K.; LaNoue, K.F.; et al. Diabetic retinopathy: seeing beyond glucose - induced microvascular disea se. Diabetes 2006 , 55 , 2401 11, doi:10.2337/db05 - 1635. 24. Funk, R. Blood Supply of the Retina. Ophthalmic Res. 1997 , 29 , 320 325, doi:10.1159/000268030. 25. Allen, C.L.; Malhi, N.K.; Whatmore, J.D.; Bates, D.O.; Arkill, K.P. Non - invasive measurement of retinal permeability in a diabetic rat model. Microcirculation 2020 , 1 , doi:10.1111/micc.12623. 26. Simó, R.; Hernández, C. Neurodegeneration is an early event in diabetic retinopathy: Therapeutic implications. Br. J. Ophthalmol. 2012 , 96 , 1285 1290, doi:10.1136/bjophthalmol - 2012 - 302005. 27. Thebeau, C.; Zhang, S.; Kolesnikov, A. V.; Kefalov, V.J.; Semenkovich, C.F.; Rajagopal, R. Light deprivation reduces the severity of experimental diabetic retinopathy. Neurobiol. Dis. 2020 , 137 , 104754, doi:10.1016/j.nbd.2020.104754. 28. Carrasco, E.; Hernandez, C.; Miralles, A.; Huguet, P.; Farres, J.; Simo, R.; Hernández, C.; Miralles, A.; Huguet, P.; Farrés, J.; et al. L ower Somatostatin Expression Is an Early Event in Diabetic Retinopathy and Is Associated With Retinal Neurodegeneration. Diabetes Care 2007 , 30 , 2902 2908, doi:10.2337/dc07 - 0332.Additional. 97 29. Barber, A.J.; Lieth, E.; Khin, S.A.; Antonetti, D.A.; Buchana n, A.G.; Gardner, T.W. Neural apoptosis in the retina during experimental and human diabetes: Early onset and effect of insulin. J. Clin. Invest. 1998 , 102 , 783 791, doi:10.1172/JCI2425. 30. Bogdanov, P.; Corraliza, L.; Villena, J.A.; Carvalho, A.R.; Garc ia - Arumí, J.; Ramos, D.; Ruberte, J.; Simó, R.; Hernández, C. The db/db mouse: A useful model for the study of diabetic retinal neurodegeneration. PLoS One 2014 , 9 , doi:10.1371/journal.pone.0097302. 31. Duh, E.J.; Sun, J.K.; Stitt, A.W. Diabetic retinopat hy: current understanding, mechanisms, and treatment strategies. JCI insight 2017 , 2 , 1 13, doi:10.1172/jci.insight.93751. 32. Busik, J. V.; Mohr, S.; Grant, M.B. Hyperglycemia - Induced reactive oxygen species toxicity to endothelial cells is dependent on paracrine mediators. Diabetes 2008 , 57 , 1952 1965, doi:10.2337/db07 - 1520. 33. Brownlee, M. Biochemistry and Molecular Cell Biol ogy of Diabetic Complications. Nature 2001 , 414 , 813 820, doi:10.1038/414813a. 34. LEE, A.Y.W.; CHUNG, S.S.M. Contributions of polyol pathway to oxidative stress in diabetic cataract. FASEB J. 1999 , 13 , 23 30, doi:10.1096/fasebj.13.1.23. 35. McNulty, M.; Mahmud, A.; Feely, J. Advanced Glycation End - Products and Arterial Stiffness in Hypertension. Am. J. Hypertens. 2007 , 20 , 242 247, doi:10.1016/j.amjhyper.2006.08.009. 36. Xu, J.; Chen, L.J.; Yu, J.; Wang, H.J.; Zhang, F.; Liu, Q.; Wu, J. Involvement of A dvanced Glycation End Products in the Pathogenesis of Diabetic Retinopathy. Cell. Physiol. Biochem. 2018 , 48 , 705 717, doi:10.1159/000491897. 37. Koya, D.; King, G.L. Perspectives in Diabetes Protein Kinase C Activation and the Development of Diabetic Com plications. Diabetes 1998 , 47 , 859 866. 38. Korshunov, S.S.; Skulachev, V.P.; Starkov, A.A. High protonic potential actuates a mechanism of production of reactive oxygen species in mitochondria. FEBS Lett. 1997 , 416 , 15 18, doi:10.1016/S0014 - 5793(97)01159 - 9. 39. Krauss, R.M. Lipids and lipoproteins in patients with type 2 diabetes. Diabetes Care 2004 , 27 , 1496 1504, doi:10.2337/diacare.27.6.1496. 40. Chew, E.Y.; Klein, M.L.; Ferris, F.L.; Remaley, N.A.; Murphy, R.P.; Chantry, K.; Hoogwerf, B.J.; Miller, D. Association of elevated serum lipid levels with retinal hard exudate in diabetic retinopathy. Early Treatment Diabetic Retinopathy Study (ETDRS) Report 22. Arch. Ophthalmol. 1996 , 114 , 1079 84. 41. Cetin, E.N.; Bulgu, Y.; Ozdemir, S.; Topsakal, S.; Akin, F.; Aybek, H.; Yildirim, C. Association of serum lipid levels with diabetic retinopathy. Int. J. Ophthalmol. 2013 , 6 , 346 349, doi:10.3980/j.issn.2222 - 3959.2013.03.17. 42. Tikhonenko, M.; Lydic, T.A .; Wang, Y.; Chen, W.; Opreanu, M.; Sochacki, A.; McSorley, K.M.; Renis, R.L.; Kern, T.; Jump, D.B.; et al. Remodeling of retinal Fatty acids in an animal model of 98 diabetes: a decrease in long - chain polyunsaturated fatty acids is associated with a decrease in fatty acid elongases Elovl2 and Elovl4. Diabetes 2010 , 59 , 219 27, doi:10.2337/db09 - 0728. 43. Huang, C.; Seregin, S.S.; et al. ELOVL4 - Mediated Producti on of Very Long - Chain Ceramides Stabilizes Tight Junctions and Prevents Diabetes - Induced Retinal Vascular Permeability. Diabetes 2018 , 67 , 769 781, doi:10.2337/db17 - 1034. 44. Busik, J. V; Reid, G.E.; Lydic, T.A. Global Analysis of Retina Lipids by Complem entary Precursor Ion and Neutral Loss Mode Tandem Mass Spectrometry. In Lipidomics. Methods in Molecular Biology ; Armstrong, D., Ed.; Humana Press: Totowa, NJ, 2009; pp. 33 70 ISBN 978 - 1 - 60761 - 322 - 0. 45. Hannun, Y.A.; Obeid, L.M. Sphingolipids and their m etabolism in physiology and disease. Nat. Rev. Mol. Cell Biol. 2018 , 19 , 175 191, doi:10.1038/nrm.2017.107. 46. Novgorodov, S. a; Szulc, Z.M.; Luberto, C.; Jones, J. a; Bielawski, J.; Bielawska, A.; Hannun, Y. a; Obeid, L.M. Positively charged ceramide is a potent inducer of mitochondrial permeabilization. J. Biol. Chem. 2005 , 280 , 16096 105, doi:10.1074/jbc.M411707200. 47. Hannun, Y.A.; Obeid, L.M. Principles of bioactive lipid signalling: Lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 2008 , 9 , 13 9 150, doi:10.1038/nrm2329. 48. Newton, J.; Lima, S.; Maceyka, M.; Spiegel, S. Revisiting the sphingolipid rheostat: Evolving concepts in cancer therapy. Exp. Cell Res. 2015 , 333 , 195 200, doi:10.1016/j.yexcr.2015.02.025. 49. Taniguchi, M.; Kitatani, K.; Kondo, T.; Hashimoto - Nishimura, M.; Asano, S.; Hayashi, A.; Mitsutake, S.; Igarashi, Y.; Umehara, H.; Takeya, H.; et al. Regulation of autophagy and its - phosph ate in the mammalian target of rapamycin pathway. J. Biol. Chem. 2012 , 287 , 39898 39910, doi:10.1074/jbc.M112.416552. 50. Jenkins, R.W.; Canals, D.; Idkowiak - Baldys, J.; Simbari, F.; Roddy, P.; Perry, D.M.; Kitatani, K.; Luberto, C.; Hannun, Y.A. Regulated secretion of acid sphingomyelinase: Implications for selectivity of ceramide formation. J. Biol. Chem. 2010 , 285 , 35706 3 5718, doi:10.1074/jbc.M110.125609. 51. Schuchman, E.H.; Desnick, R.J. Types A and B Niemann - Pick disease. Mol. Genet. Metab. 2017 , 120 , 27 33, doi:10.1016/j.ymgme.2016.12.008. 52. Yu, Z.F.; Nikolova - Karakashian, M.; Zhou, D.; Cheng, G.; Schuchman, E.H.; Mattson, M.P. Pivotal role for acidic sphingomyelinase in cerebral ischemia - induced ceramide and cytokine production, and neuronal apoptosis. J. Mol. Neurosci. 2000 , 15 , 85 97, doi:1 0.1385/JMN:15:2:85. 53. Santana, P.; Peña, L.A.; Haimovitz - Friedman, A.; Martin, S.; Green, D.; McLoughlin, M.; Cordon - Cardo, C.; Schuchman, E.H.; Fuks, Z.; Kolesnick, R. Acid sphingomyelinase - deficient human lymphoblasts and mice are defective in radiati on - induced apoptosis. Cell 1996 , 86 , 189 199, doi:10.1016/S0092 - 8674(00)80091 - 4. 99 54. Opreanu, M.; Tikhonenko, M.; Bozack, S.; Lydic, T.A.; Reid, G.E.; McSorley, K.M.; Sochacki, A.; Perez, G.I.; Esselman, W.J.; Kern, T.; et al. The unconventional role of a cid sphingomyelinase in regulation of retinal microangiopathy in diabetic human and animal models. Diabetes 2011 , 60 , 2370 8, doi:10.2337/db10 - 0550. 55. Opreanu, M.; Lydic, T.A.; Reid, G.E.; McSorley, K.M.; Esselman, W.J.; Busik, J. V Inhibition of cytoki ne signaling in human retinal endothelial cells through downregulation of sphingomyelinases by docosahexaenoic acid. Invest. Ophthalmol. Vis. Sci. 2010 , 51 , 3253 63, doi:10.1167/iovs.09 - 4731. 56. Van Blitterswijk, W.J.; Van Der Luit, A.H.; Veldman, R.J.; Verheij, M.; Borst, J. Ceramide: Second messenger or modulator of membrane structure and dynamics? Biochem. J. 2003 , 369 , 199 211, doi:10.1042/BJ20021528. 57. y, N.; Huang, C.; Blanchard, G.J.; et al. Role of Acid Sphingomyelinase in Shifting the Balance between Proinflammatory and Reparative Bone Marrow Cells in Diabetic Retinopathy. Stem Cells 2016 , 34 , 972 983, doi:10.1002/stem.2259. 58. Devi, T.S.; Yumnamch a, T.; Yao, F.; Somayajulu, M.; Kowluru, R.A.; Singh, L.P. TXNIP mediates high glucose - induced mitophagic flux and lysosome enlargement in human retinal pigment epithelial cells. Biol. Open 2019 , 8 , 1 13, doi:10.1242/bio.038521. 59. Du, Y.; Miller, C.M.; Kern, T.S. Hyperglycemia increases mitochondrial superoxide in retina and retinal cells. Free Radic. Biol. Med. 2003 , 35 , 1491 1499, doi:10.1016/j.freeradbiomed.2003.08.018. 60. Kumar, B.; Kowluru, A.; Kowluru, R.A. Lipotoxicity augm ents glucotoxicity - induced mitochondrial damage in the development of diabetic retinopathy. Invest. Ophthalmol. Vis. Sci. 2015 , 56 , 2985 92, doi:10.1167/iovs.15 - 16466. 61. Trudeau, K.; Molina, A.J.A.; Guo, W.; Roy, S. High Glucose Disrupts Mitochondrial M orphology in Retinal Endothelial Cells: Implications for Diabetic Retinopathy. Am. J. Pathol. 2010 , 177 , 447 455, doi:10.2353/ajpath.2010.091029. 62. Roy, S.; Trudeau, K.; Roy, S.; Tien, T.; Barrette, K.F. Mitochondrial dysfunction and endoplasmic reticul um stress in diabetic retinopathy: mechanistic insights into high glucose - induced retinal cell death. Curr. Clin. Pharmacol. 2013 , 8 , 278 84. 63. Kogot - Levin, A.; Saada, A. Ceramide and the mitochondrial respiratory chain. Biochimie 2014 , 100 , 88 94, doi: 10.1016/j.biochi.2013.07.027. 64. Dai, Q.; Liu, J.; Chen, J.; Durrant, D.; McIntyre, T.M.; Lee, R.M. Mitochondrial ceramide increases in UV - irradiated HeLa cells and is mainly derived from hydrolysis of sphingomyelin. Oncogene 2004 , 23 , 3650 3658, doi:10. 1038/sj.onc.1207430. 65. Yu, J.; Novgorodov, S.A.; Chudakova, D.; Zhu, H.; Bielawska, A.; Bielawski, J.; Obeid, L.M.; Kindy, M.S.; Gudz, T.I. JNK3 signaling pathway activates ceramide synthase leading to mitochondrial dysfunction. J. Biol. Chem. 2007 , 282 , 25940 9, doi:10.1074/jbc.M701812200. 100 66. Managò, A.; Becker, K.A.; Carpinteiro, A.; Wilker, B.; Soddemann, M.; Seitz, A.P.; Edwards, M.J.; Grassmé, H.; Szabò, I.; Gulbins, E. Pseudomonas aeruginosa Pyocyanin Induces Neutrophil Death via Mitochondrial Re active Oxygen Species and Mitochondrial Acid Sphingomyelinase. Antioxid. Redox Signal. 2015 , 22 , 1097 1110, doi:10.1089/ars.2014.5979. 67. BIRBES, H.; EL BAWAB, S.; HANNUN, Y.A.; OBEID, L.M. Selective hydrolysis of a mitochondrial pool of sphingomyelin in duces apoptosis. FASEB J. 2001 , 15 , 2669 2679, doi:10.1096/fj.01 - 0539com. 68. Chang, K. - T.; Anishkin, A.; Patwardhan, G.A.; Beverly, L.J.; Siskind, L.J.; Colombini, M. Ceramide channels: destabilization by Bcl - xL and role in apoptosis. Biochim. Biophys. Acta - Biomembr. 2015 , 1848 , 2374 2384, doi:10.1016/j.bbamem.2015.07.013. 69. C olombini, M. Membrane channels formed by ceramide. Handb. Exp. Pharmacol. 2013 , 109 26, doi:10.1007/978 - 3 - 7091 - 1368 - 4_6. 70. Colombini, M. Ceramide channels and mitochondrial outer membrane permeability. J. Bioenerg. Biomembr. 2017 , 49 , 57 64, doi:10.1007 /s10863 - 016 - 9646 - z. 71. Siskind, L.J.; Kolesnick, R.N.; Colombini, M. Ceramide channels increase the permeability of the mitochondrial outer membrane to small proteins. J. Biol. Chem. 2002 , 277 , 26796 803, doi:10.1074/jbc.M200754200. 72. Perera, M.N.; Ga nesan, V.; Siskind, L.J.; Szulc, Z.M.; Bielawski, J.; Bielawska, A.; Bittman, R.; Colombini, M. Ceramide channels: influence of molecular structure on channel formation in membranes. Biochim. Biophys. Acta 2012 , 1818 , 1291 1301, doi:10.1016/j.bbamem.2012.0 2.010. 73. Gudz, T.I.; Tserng, K. - Y.; Hoppel, C.L. Direct Inhibition of Mitochondrial Respiratory Chain Complex III by Cell - permeable Ceramide. J. Biol. Chem. 1997 , 272 , 24154 24158, doi:10.1074/jbc.272.39.24154. 74. Zimorski, V.; Ku, C.; Martin, W.F.; G ould, S.B. Endosymbiotic theory for organelle origins. Curr. Opin. Microbiol. 2014 , 22 , 38 48, doi:10.1016/j.mib.2014.09.008. 75. Nobel Prize in Chemistry for Biological Energy Transfer Available online: https://www.nobelprize.org/prizes/chemistry/1978/pr ess - release/ (accessed on Mar 5, 2020). 76. Morelli, A.M.; Ravera, S.; Calzia, D.; Panfoli, I. An update of the chemiosmotic theory as suggested by possible proton currents inside the coupling membrane. Open Biol. 2019 , 9 , doi:10.1098/rsob.180221. 77. Brand, M.D.; Nicholls, D.G. Assessing mitochondrial dysfunction in cells. Biochem. J. 2011 , 435 , 297 312, doi:10.1042/BJ20110162. 78. Ortiz - Prado, E.; Dunn, J.F.; Vasconez, J.; Castillo, D.; Viscor, G. Partial pressure of oxygen in the human body: a gener al review. Am. J. Blood Res. 2019 , 9 , 1 14. 79. Gnaiger, E.; Lassnig, B.; Kuznetsov, a; Rieger, G.; Margreiter, R. Mitochondrial oxygen affinity, 101 respiratory flux control and excess capacity of cytochrome c oxidase. J. Exp. Biol. 1998 , 201 , 1129 1139. 80. Gnaiger, E. Polarographic Oxygen Sensors, the Oxygraph, and High - Resolution Respirometry to Assess Mitochondrial Function. In Drug - Induced Mitochondrial Dysfunction ; Dykens, J.A., Will, Y., Eds.; John Wiley & Sons, Inc, 2008; pp. 327 351. 81. Pesta, D.; Gnaiger, E. High - Resolution Respirometry: OXPHO S Protocols for Human Cells and Permeabilized Fibers from Small Biopsies of Human Muscle. In Methods in Molecular Biology ; Humana Press, 2012; pp. 25 58. 82. Salabei, J.K.; Gibb, A.A.; Hill, B.G. Comprehensive measurement of respiratory activity in permea bilized cells using extracellular flux analysis. Nat. Protoc. 2014 , 9 , 421 438, doi:10.1038/nprot.2014.018. 83. Clark, L.C.; Wolf, R.; Granger, D.; Taylor, Z. Continuous Recording of Blood Oxygen Tensions by Polarography. J. Appl. Physiol. 1953 , 6 , 189 19 3. 84. Otto, A.M. Warburg effect(s) a biographical sketch of Otto Warburg and his impacts on tumor metabolism. Cancer Metab. 2016 , 4 , 1 8, doi:10.1186/s40170 - 016 - 0145 - 9. 85. Mookerjee, S.A.; Goncalves, R.L.S.; Gerencser, A.A.; Nicholls, D.G.; Brand, M.D. The contributions of respiration and glycolysis to extracellular acid production. Biochim. Biophys. Acta - Bioenerg. 2015 , 1847 , 171 181, doi:10.1016/j.bbabio.2014.10.005. 86. Lighton, J.R.B. Measuring Metabolic Rates: A Manual for Scientists ; 1st ed.; O xford University Press: New York, New York, 2008; Vol. 1; ISBN 9780195310610. 87. Gerencser, A.A.; Neilson, A.; Choi, S.W.; Edman, U.; Yadava, N.; Oh, R.J.; Ferrick, D.A.; Nicholls, D.G.; Brand, M.D. Quantitative microplate - based respirometry with correct ion for oxygen diffusion. Anal. Chem. 2009 , 81 , 6868 78, doi:10.1021/ac900881z. 88. Kondrashina, A. V.; Papkovsky, D.B.; Dmitriev, R.I. Measurement of cell respiration and oxygenation in standard multichannel biochips using phosphorescent O2 - sensitive probes. Analyst 2013 , 138 , 4915, doi:10.1039/c3an00658a. 89. Severinghaus, J.W. First e lectrodes for blood PO2 and PCO2 determination. J. Appl. Physiol. 2004 , 97 , 1599 1600, doi:10.1152/classicessays.00021.2004.Editorial. 90. Bhalla, N.; Jolly, P.; Formisano, N.; Estrela, P. Introduction to biosensors. Essays Biochem. 2016 , 60 , 1 8, doi:10. 1042/EBC20150001. 91. Severinghaus, J.W.; Freeman, B.A. Electrodes for Blood PO, and pC0, Determination. J. Appl. Physiol. 1958 , 13 , 515 520. 92. Chance, B.; Williams, G.R. Respiratory Enzymes in Oxidative Phosphorylation. J. Biol. 1955 , 1 , 409 428. 93. Hinkle, P.C. P/O ratios of mitochondrial oxidative phosphorylation. Biochim. Biophys. Acta - 102 Bioenerg. 2005 , 1706 , 1 11, doi:10.1016/j.bbabio.2004.09.004. 94. Oomen, P.E.; Skolimowski, M.; Verpoorte, S. Implementing Oxygen Control in Chip - Based Cell and Tissue Culture Systems. Lab Chip 2016 , 16 , 3394 3414, doi:10.1039/C6LC00772D. 95. Papkovsky, D.B.; Dmitriev, R.I. Biological detection by optical oxygen sensing. Chem. Soc. Rev. 2013 , 42 , 8700 8732, doi:10.1039/c3cs60131e. 96. Ochs, C.J.; Kasuya, J.; Pav esi, A.; Kamm, R.D. Oxygen levels in thermoplastic microfluidic devices during cell culture. Lab Chip 2014 , 14 , 459 62, doi:10.1039/c3lc51160j. 97. Khan, D.H.; Roberts, S.A.; Cressman, J.R.; Agrawal, N. Rapid Generation and Detection of Biomimetic Oxygen Concentration Gradients in Vitro. Sci. Rep. 2017 , 7 , 1 11, doi:10.1038/s41598 - 017 - 13886 - z. 98. Lam, R.H.W.; Kim, M.C.; Thorsen, T. Culturing aerobic and anaerobic bacteria and mammalian cells with a microfluidic differential oxygenator. Anal. Chem. 2009 , 81 , 5918 5924, doi:10.1021/ac9006864. 99. Pham, T.D.; Wallace, D.C.; Burke, P.J. Microchambers with solid - state phosphorescent sensor for measuring single mitochondrial respiration. Sensors 2016 , 16 , doi:10.3390/s16071065. 100. Zand, K.; Pham, T.; Jr, A.D. Nanofluidic Platform for Single Mitochondria Analysis Using Fluorescence Microscopy. 2013 , 85 , 6018 6025. 101. Kelbauskas, L.; Ashili, S. P.; Lee, K.B.; Zhu, H.; Tian, Y.; Meldrum, D.R. Simultaneous multiparameter cellular energy metabolism profiling of small populations of cells. Sci. Rep. 2018 , 8 , 1 12, doi:10.1038/s41598 - 018 - 22599 - w. 102. Bénit, P.; Chrétien, D.; Porceddu, M.; Yanicostas , C.; Rak, M.; Rustin, P. An Effective, Versatile, and Inexpensive Device for Oxygen Uptake Measurement. J. Clin. Med. 2017 , 6 , 58, doi:10.3390/jcm6060058. 103. Molter, T.W.; Holl, M.R.; Dragavon, J.M.; McQuaide, S.C.; Anderson, J.B.; Young, A.C.; Burgess , L.W.; Lidstrom, M.E.; Meldrum, D.R. A new approach for measuring single - cell oxygen consumption rates. IEEE Trans. Autom. Sci. Eng. 2008 , 5 , 32 40, doi:10.1109/TASE.2007.909441. 104. Molter, T.W.; Mcquaide, S.C.; Suchorolski, M.T.; Strovas, T.J.; Lloyd, W.; Meldrum, D.R.; Lidstrom, M.E.; Hall, B.; Northlake, N.E.; Rm, P. A microwell array device capable of measuring single - cell oxygen consumption rates. Sensors Actuators B Chem 2009 , 135 , 678 686, doi:10.1016/j.snb.2008.10.036.A. 105. Oppegard, S.C.; Bl ake, A.J.; Williams, J.C.; Eddington, D.T. Precise control over the oxygen conditions within the Boyden chamber using a microfabricated insert. Lab Chip 2010 , 10 , 2366, doi:10.1039/c004856a. 106. Rogers, G.W.; Brand, M.D.; Petrosyan, S.; Ashok, D.; Elorza, A.A.; Ferrick, D.A.; Murphy, A.N. High throughput microplate respiratory measurements using minimal quantities of isolated mitochondria. PLoS One 2011 , 6 , e21746, doi:10.1371/journal.pone.0021746 . 103 107. Divakaruni, A.S.; Paradyse, A.; Ferrick, D.A.; Murphy, A.N.; Jastroch, M. Analysis and interpretation of microplate - based oxygen consumption and pH data ; 1st ed.; Elsevier Inc., 2014; Vol. 547; ISBN 1557 - 7988 (Electronic)r0076 - 6879 (Linking). 108. Divakaruni, A.S.; Wiley, S.E.; Rogers, G.W.; Andreyev, A.Y.; Petrosyan, S.; Loviscach, M.; Wall, E.A.; Yadava, N.; Heuck, A.P.; Ferrick, D.A.; et al. Thiazolidinediones are acute, specific inhibitors of the mitochondrial pyruvate carrier. Proc. Natl. Acad . Sci. 2013 , 110 , 5422 7, doi:10.1073/pnas.1303360110. 109. Haller, T.; Ortner, M.; Gnaiger, E. A Respirometer for Investigating Oxidative Cell Metabolism: Toward Optimization of Respiratory Studies. Anal. Biochem. 1994 , 218 , 338 342. 110. Gnaiger, E.; S teinlechner - Maran, R.; Méndez, G.; Eberl, T.; Margreiter, R. Control of mitochondrial and cellular respiration by oxygen. J. Bioenerg. Biomembr. 1995 , 27 , 583 596, doi:10.1007/BF02111656. 111. Steinlechner - Maran, R.; Eberl, T.; Kunc, M.; Margreiter, R.; G naiger, E. Oxygen dependence of respiration in coupled and uncoupled endothelial cells. Am. J. Physiol. - Cell Physiol. 1996 , 271 , doi:10.1152/ajpcell.1996.271.6.c2053. 112. Iliescu, C.; Taylor, H.; Avram, M.; Miao, J.; Franssila, S. A practical guide for the fabrication of microfluidic devices using glass and silicon. Biomicrofluidics 2012 , 6 , 16505 1650516, doi:10.1063/1.3689939. 113. Waheed, S.; Cabot, J.M.; Macdonald, N.P.; Lewis, T.; Guijt, R.M.; Paull, B.; Breadmore, M.C. 3D printed microfluidic dev ices: Enablers and barriers. Lab Chip 2016 , 16 , 1993 2013, doi:10.1039/c6lc00284f. 114. Gebhardt, A. Understanding Additive Manufacturing ; 1st ed.; Hanser Publications: Cincinnati, 2011; ISBN 9781569905074. 115. Gross, B.C.; Erkal, J.L.; Lockwood, S.Y.; Chen, C.; Spence, D.M. Evaluation of 3D printing and its potential impact on biotechnology and the chemical sciences. Anal. Chem. 2014 , 86 , 3240 3253, doi:10.1021/ac403397r. 116. Rimington, R.P.; Capel, A.J.; Play er, D.J.; Bibb, R.J.; Christie, S.D.R.; Lewis, M.P. Feasibility and Biocompatibility of 3D - Printed Photopolymerized and Laser Sintered Polymers for Neuronal, Myogenic, and Hepatic Cell Types. Macromol. Biosci. 2018 , 18 , 1 12, doi:10.1002/mabi.201800113. 11 7. Gross, B.C.; Anderson, K.B.; Meisel, J.E.; McNitt, M.I.; Spence, D.M. Polymer Coatings in 3D - Printed Fluidic Device Channels for Improved Cellular Adherence Prior to Electrical Lysis. Anal. Chem. 2015 , 87 , 6335 6341, doi:10.1021/acs.analchem.5b01202. 1 18. Nock, V.; Blaikie, R.J.; David, T. Patterning, integration and characterisation of polymer optical oxygen sensors for microfluidic devices. Lab Chip 2008 , 8 , 1300 1307, doi:10.1039/b801879k. 119. Carraway, E.R.; Demas, J.N.; DeGraff, B.A.; Bacon, J.R . Photophysics and Photochemistry of Oxygen Sensors Based on Luminescent Transition - Metal Complexes. Anal. Chem. 1991 , 63 , 337 104 342, doi:10.1021/ac00004a007. 120. Fluorescence Quenching Studies Available online: https://www.chem.uzh.ch/de/study/download/ye ar2/che211.html. 121. Forstner, H.; Gnaiger, E. Calculation of Equilibrium Oxygen Concentration. In Polarographic Oxygen Sensors: Aquatic and Physiological Applications ; Gnaiger, E., Forstner, H., Eds.; Springer - Verlag Berline Heidelberg: Heidelberg, 1983 ; pp. 321 333 ISBN 978 - 3 - 642 - 81865 - 3. 122. Kneas, K.A.; Xu, W.; Demas, J.N.; Degraff, B.A. Oxygen sensors based on luminescence quenching: Interactions of tris(4,7 - diphenyl - 1,10 - phenanthroline)ruthenium(II) chloride and pyrene with polymer supports. Appl. Spectrosc. 1997 , 51 , 1346 1351, doi:10.1366/0003702971942024. 123. Murphy, M.P. How understanding the control of energy metabolism can help investigation of mitochondrial dysfunction, regulation and pharmacology. Biochim. Biophys. Acta - Bioenerg. 2001 , 1504 , 1 11, doi:10.1016/S0005 - 2728(00)00234 - 6. 124. Rolfe, D.F.S.; Brown, G.C. Cellular energy utilization and molecular origin of standard metabolic rate in mammals. Physiol. Rev. 1997 , 77 , 731 758, doi:10.1152/physrev.1997.77.3.731. 125. Clerc, P.; Car ey, G.B.; Mehrabian, Z.; Wei, M.; Hwang, H.; Girnun, G.D.; Chen, H.; Martin, S.S.; Polster, B.M. Rapid detection of an ABT - 737 - sensitive primed for death state in cells using microplate - based respirometry. PLoS One 2012 , 7 , 15 18, doi:10.1371/journal.pone.0042487. 126. Moreadith, R.W.; Lehninger, A.L. The pathways of glutamate and glutamine oxidation by tumor cell mitochondria. Role of mitochondrial NAD(P)+ - dependent malic enzyme. J. Biol. Chem. 1984 , 259 , 6215 6221. 127. Hutter, E.; Renner, K.; Pfister, G.; Stöckl, P.; Jansen - Dürr, P.; Gnaiger, E. Senescence - associated changes in respiration and oxidative phosphorylation in primary human fibroblasts. Biochem. J. 2004 , 380 , 919 928, doi:10.1042/BJ20040095. 128. Keuper, M.; Jastroc h, M.; Yi, C.X.; Fischer - Posovszky, P.; Wabitsch, M.; Tschöp, M.H.; Hofmann, S.M. Spare mitochondrial respiratory capacity permits human adipocytes to maintain ATP homeostasis under hypoglycemic conditions. FASEB J. 2014 , 28 , 761 770, doi:10.1096/fj.13 - 238 725. 129. Kuznetsov, A. V.; Strobl, D.; Ruttmann, E.; Königsrainer, A.; Margreiter, R.; Gnaiger, E. Evaluation of mitochondrial respiratory function in small biopsies of liver. Anal. Biochem. 2002 , 305 , 186 194, doi:10.1006/abio.2002.5658. 130. Gnaiger, E. Mitochondrial Pathways and Respiratory Control An Introduction to OXPHOS Analysis ; 2014; ISBN 9783950239966. 131. Brand, M.D.; Chien, L.F.; Diolez, P. Experimental discrimination between proton leak and redox slip during mitochondrial electron transpor t. Biochem. J. 1994 , 297 , 27 29, doi:10.1042/bj2970027. 132. Harper, M.; Brand, D. The Quantitative Contributions of Mitochondrial Proton Leak and ATP 105 Turnover Reactions to the Changed Respiration Rates of Hepatocytes from Rats of Different Thyroid Status . J. Biol. Chem. 1993 , 268 , 14850 14860. 133. Divakaruni, A.S.; Brand, M.D. The regulation and physiology of mitochondrial proton leak. Physiology 2011 , 26 , 192 205, doi:10.1152/physiol.00046.2010. 134. Ruas, J.S.; Siqueira - Santos, E.S.; Amigo, I.; Rodrigues - Silva, E.; Kowaltowski, A.J.; Castilho, R.F. Underestimation of the maximal capacity of the mitochondrial electron transport system in oligomycin - treated cells. PLoS One 2016 , 11 , 1 20, doi:10.1371/journal.pone.0150967. 135. Jaber, S.M.; Yadava, N.; Polster, B.M. Mapping Mitochondrial Respiratory Chain Deficiencies by Respirometry: Beyond the Mito Stress Test. Exp. Neurol. 2020 , doi:https://doi.org/10.1016/j.expneurol.2020.113282. 136. Fell, D.A. Understanding the control of metabolism ; Fell, D.A., Snell, K., Eds.; Ashgate Publishing: Surrey, 1997; 137. Eigentler, A.; Draxl, A.; Wiethüchter, A. Laboratory protocol: citrate synthase a mitochondrial marker enzyme. Mitochondrial Physiol. N etw. 2015 , 04 , 1 11. 138. CHANCE, B.; WILLIAMS, G.R. Respiratory enzymes in oxidative phosphorylation. IV. The respiratory chain. J. Biol. Chem. 1955 , 217 , 429 438. 139. Picard, M.; Taivassalo, T.; Ritchie, D.; Wright, K.J.; Thomas, M.M.; Romestaing, C.; Hepple, R.T. Mitochondrial structure and function are disrupted by standard isolation methods. PLoS One 2011 , 6 , e18317, doi:10.1371/journal.pone.0018317. 140. Benador, I.Y.; Veliova, M.; Mahdaviani, K.; Petcherski, A.; Wikstrom, J.D.; Assali, E.A.; Acín - Pérez, R.; Shum, M.; Oliveira, M.F.; Cinti, S.; et al. Mitochondria Bound to Lipid Droplets Have Unique Bioenergetics, Composition, and Dynamics that Support Lipid Droplet Expansion. Cell Metab. 2018 , 27 , 869 - 885.e6, doi:10.1016/j.cmet.2018.03.003. 141. Kuznetsov, A. V.; Veksler, V.; Gellerich, F.N.; Saks, V.; Margreiter, R.; Kunz, W.S. Analysis of mitochondrial function in situ in permeabilized muscle fibers, tissues and cells. Nat. Protoc. 2008 , 3 , 965 976, doi:10.1038/nprot.2008.61. 142. Salabei, J.K. ; Gibb, A.A.; Hill, B.G. Comprehensive measurement of respiratory activity in permeabilized cells using extracellular flux analysis Joshua. Nat. Protoc. 2014 , 9 , 421 438, doi:10.1016/j.pestbp.2011.02.012.Investigations. 143. Saks, V.A.; Belikova, Y.O.; Ku znetsov, A. V. In vivo regulation of mitochondrial respiration in cardiomyocytes: specific restrictions for intracellular diffusion of ADP. Biochim. Biophys. Acta 1991 , 1074 , 302 311, doi:10.1016/0304 - 4165(91)90168 - G. 144. Mathers, K.E.; Staples, J.F. Sap onin - permeabilization is not a viable alternative to isolated mitochondria for assessing oxidative metabolism in hibernation. Biol. Open 2015 , 4 , 858 864, doi:10.1242/bio.011544. 145. Picard, M.; Ritchie, D.; Wright, K.J.; Romestaing, C.; Thomas, M.M.; Rowan, S.L.; Taivassalo, T.; 106 Hepple, R.T. Mitochondrial functional impairment with aging is exaggerated in isolated mitochondria compared to permeabilized myofibers. Aging Cell 2010 , 9 , 1 032 1046, doi:10.1111/j.1474 - 9726.2010.00628.x. 146. Munier - Lehmann, H.; Vidalain, P.O.; Tangy, F.; Janin, Y.L. On dihydroorotate dehydrogenases and their inhibitors and uses. J. Med. Chem. 2013 , 56 , 3148 3167, doi:10.1021/jm301848w. 147. Gnaiger, E. Mit Mitochondrial Physiol. Netw. 2011 , 11 , 1 4. 148. Gnaiger, E. Capacity of oxidative phosphorylation in human skeletal muscle. New perspectives of mitochondrial physiology. Int. J. Bi ochem. Cell Biol. 2009 , 41 , 1837 1845, doi:10.1016/j.biocel.2009.03.013. 149. Puchowicz, M.A.; Varnes, M.E.; Cohen, B.H.; Friedman, N.R.; Kerr, D.S.; Hoppel, C.L. Oxidative phosphorylation analysis: Assessing the integrated functional activity of human sk eletal muscle mitochondria - Case studies. Mitochondrion 2004 , 4 , 377 385, doi:10.1016/j.mito.2004.07.004. 150. Kimelberg, H.K.; Nicholls, P. Kinetic Studies on the Interaction c and Cytochrome of TMPD with c used p - phenylene - diamine routinely in the manometric assay of cytochrome c oxidase activity in heart muscle preparations chrome oxidase . He. Arch. Biochem. Biophys. 1969 , 133 , 327 335. 151. Gnaiger, E.; Boushel, R.; Søndergaard, H.; Munch - Andersen, T.; Damsgaard, R.; Hagen, C.; Díez - Sánchez, C.; Ara, I.; Wright - Paradis, C.; Schrauwen, P.; et al. Mitochondrial coupling and capacity of oxidative phosphorylation in skeletal muscle of Inuit and Caucasians in the arctic winter. Scand. J. Med. Sci. Sport. 201 5 , 25 , 126 134, doi:10.1111/sms.12612. 152. Kasper, J.D.; Meyer, R.A.; Beard, D.A.; Wiseman, R.W. Effects of altered pyruvate dehydrogenase activity on contracting skeletal muscle bioenergetics. Am. J. Physiol. - Regul. Integr. Comp. Physiol. 2019 , 316 , R 76 R86, doi:10.1152/ajpregu.00321.2018. 153. Wanders, R.J.A.; Westerhoff, H. V. Sigmoidal Relation between Mitochondrial Respiration and log ([ATP]/[ADP])OUtunder Conditions of Extramitochondrial ATP Utilization. Implications for the Control and Thermodynamics of Oxidative Phosphorylation. Biochemist ry 1988 , 27 , 7832 7840, doi:10.1021/bi00420a037. 154. Jekabsons, M.B.; Nicholls, D.G. In Situ respiration and bioenergetic status of mitochondria in primary cerebellar granule neuronal cultures exposed continuously to glutamate. J. Biol. Chem. 2004 , 279 , 32989 33000, doi:10.1074/jbc.M401540200. 155. Zirath, H.; Rothbauer, M.; Spitz, S.; Bachmann, B.; Jordan, C.; Müller, B.; Ehgartner, J.; Priglinger, E.; Mühleder, S.; Redl, H.; et al. Every breath you take: Non - invasive real - time oxygen biosensing in two - and three - dimensional microfluidic cell models. Front. Physiol. 2018 , 9 , 1 12, doi:10.3389/fphys.2018.00815. 156. Kathuria, A.; Brouwers, N.; Buntinx, M.; Harding, T.; Auras, R. Effect of MIL - 53 (Al) MOF particles on the chain mobility and crystallizatio n of poly(L - lactic acid). J. Appl. Polym. Sci. 2017 , 53 , 45690, doi:10.1002/app.45690. 107 157. Auras, R.A.; Singh, S.P.; Singh, J.J. Evaluation of oriented poly(lactide) polymers vs. existing PET and oriented PS for fresh food service containers. Packag. Tec hnol. Sci. 2005 , 18 , 207 216, doi:10.1002/pts.692. 158. Frezza, C.; Cipolat, S.; Scorrano, L. Organelle isolation: functional mitochondria from mouse liver, muscle and cultured fibroblasts. Nat. Protoc. 2007 , 2 , 287 95, doi:10.1038/nprot.2006.478. 159. G reen, E.J.; Carritt, D.E. Oxygen Solubility in Sea Water: Thermodynamic Influence of Sea Salt. Science (80 - . ). 1967 , 157 , 191 193. 160. Gillanders, R.N.; Tedford, M.C.; Crilly, P.J.; Bailey, R.T. Thin film dissolved oxygen sensor based on platinum octaethylporphyrin encapsulated in an elastic fluorinated polymer. Anal. Chim. Acta 2004 , 502 , 1 6, doi:10.1016/j.aca.2003.09.053. 161. Thomas, P.C.; Raghavan, S.R.; Forry, S.P. Regulating Oxygen Levels in a Microfluidic Device. Anal. Chem. 2011 , 83 , 8821 8824, doi:10.1021/ac202300g. 162. Krab, K.; Kempe, H.; Wikström, M. Explaining the enigmatic KM for oxygen in cytochrome c oxidase: A kinetic model. Biochim. Biophys. Acta - Bioenerg. 2011 , 1807 , 348 358, doi:10.1016/J.BBABIO.2010.12.015. 163. Vellonen, K. - S.; Malinen, M.; Mannermaa, E.; Subrizi, A.; Toropainen, E.; Lou, Y. - R.; Kidron, H.; Yliperttula, M.; Urtti, A. A critical assessment of in vitro tissue models for ADME and drug delivery. J. Control. release 2014 , 190C , 94 114, doi:10.1016/j.jconrel.2014.06.044. 164. Horan, M.P.; Pichau d, N.; Ballard, J.W.O. Review: Quantifying mitochondrial dysfunction in complex diseases of aging. Journals Gerontol. - Ser. A Biol. Sci. Med. Sci. 2012 , 67 A , 1022 1035, doi:10.1093/gerona/glr263. 165. Arato - Oshima, T.; Matsui, H.; Wakizaka, A.; Homareda , H. Mechanism responsible for oligomycin - induced occlusion of Na+ within Na/K - ATPase. J. Biol. Chem. 1996 , 271 , 25604 25610, doi:10.1074/jbc.271.41.25604. 166. Glaser, E.; Norling, B.; Kopecky, J.; Ernster, L. Comparison of the Effects of Oligomycin and Dicyclohexylcarbodiimide on Mitochondrial ATPase and Related Reactions. Eur. J. Biochem. 1982 , 121 , 525 531, doi:10.1111/j.1432 - 1033.1982.tb05818.x. 167. Padman, B.S.; Bach, M.; Lucarelli, G.; Prescott, M.; Ramm, G. The protonophore CCCP interferes with l ysosomal degradation of autophagic cargo in yeast and mammalian cells. Autophagy 2013 , 9 , 1862 1875, doi:10.4161/auto.26557. 168. Heytler, P.G. Uncoupling of Oxidative Phosphorylation by Carbonyl Cyanide Phenylhydrazones. I. Some Characteristics of m - Cl - C CP Action on Mitochondria and Chloroplasts. Biochemistry 1963 , 2 , 357 361, doi:10.1021/bi00902a031. 169. Yun, H.; Kim, K.; Lee, W.G. Effect of a dual inlet channel on cell loading in microfluidics. Biomicrofluidics 2014 , 8 , 1 9, doi:10.1063/1.4901929. 170 . Kolnik, M.; Tsimring, L.S.; Hasty, J. Vacuum - assisted cell loading enables shear - free mammalian 108 microfluidic culture. Lab Chip 2012 , 12 , 4732 4737, doi:10.1039/c2lc40569e.Vacuum - assisted. 171. Grist, S.M.; Chrostowski, L.; Cheung, K.C. Optical oxygen sensors for applications in microfluidic cell culture. Sensors 2010 , 10 , 9286 9316, doi:10.3390/s101009286. 172. Abaci, H.E.; Devendra, R.; Smith, Q.; Gerecht, S.; Drazer, G. Design and development of microbioreactors for long - term cell culture in controlled oxygen microenvironments. Biomed. Microdevices 2012 , 14 , 145 52, doi:10.1007/s10544 - 011 - 9592 - 9. 173. Lemasters, J.J. The ATP - to - Oxygen Stoichiometries of Oxidative Phosphorylation by Rat Liver Mitochondria. J. Biol. Chem. 1984 , 259 , 13123 13130. 174. Chau, L.; Doran, M.; Cooper - White, J. A novel multishear microdevice for studying cell mechanics. Lab Chip 2009 , 9 , 1897 1902, doi:10.1039/b823180j. 175. Shao, J.; Wu, L.; Wu, J.; Zheng, Y.; Zhao, H.; Jin, Q.; Zhao, J. Integrated microfluidic chip for endothelial cells culture and analysis exposed to a pulsatile and oscillatory shear stress. Lab Chip 2009 , 9 , 3118 3125, doi:10.1039/b909312e. 176. Oyre, S.; Pedersen, E.M.; Ringgaard, S.; Boesiger, P.; Paaske, W.P. In vivo wall shear stress measured by magnetic resonance velocity mapping in the normal human abdominal aorta. Eur. J. Vasc. Endovasc. Surg. 1997 , 13 , 263 271, doi:10.1016/S1078 - 5884(97)80097 - 4. 177. Fröhlich, E.; Bonstingl, G.; Höfler, A .; Meindl, C.; Leitinger, G.; Pieber, T.R.; Roblegg, E. Comparison of two in vitro systems to assess cellular effects of nanoparticles - containing aerosols. Toxicol. Vitr. 2013 , 27 , 409 417, doi:10.1016/j.tiv.2012.08.008. 178. Chen, F.; Luo, Y.; Tsoutsos, N.G.; Maniatakos, M.; Shahin, K.; Gupta, N. Embedding Tracking Codes in Additive Manufactured Parts for Product Authentication. Adv. Eng. Mater. 2018 , 1800495 , 1 8, doi:10.1002/adem.201800495. 179. Robins, M.; Solomon, J.B.; Samei, E. Can a 3D task transf er function accurately represent the signal transfer properties of low - contrast lesions in non - linear CT systems? Med. Imaging 2018 Phys. Med. Imaging 2018 , 10573 , 148, doi:10.1117/12.2294588. 180. Simó, R.; Villarroel, M.; Corraliza, L.; Hernández, C.; Garcia - Ramírez, M. The retinal pigment epithelium: something more than a constituent of the blood - retinal barrier -- implications for the pathogenesis of diabetic retinopathy. J. Biomed. Biotechnol. 200 9 , 2010 , 15, doi:10.1155/2010/190724. 181. R. Sparrrow, J.; Hicks, D.; P. Hamel, C. The Retinal Pigment Epithelium in Health and Disease. Curr. Mol. Med. 2010 , 10 , 802 823, doi:10.2174/156652410793937813. 182. Berkowitz, B.A.; Olds, H.K.; Richards, C.; J oy, J.; Rosales, T.; Podolsky, R.H.; Childers, K.L.; Brad Hubbard, W.; Sullivan, P.G.; Gao, S.; et al. Novel imaging biomarkers for mapping the impact of mild mitochondrial uncoupling in the outer retina in vivo. PLoS One 2020 , 15 , 1 16, doi:10.1371/journa l.pone.0226840. 183. Saint - 109 RPE - derived soluble VEGF in the maintenance of the choriocapillaris. Proc. Natl. Acad. Sci. U. S. A. 2009 , 106 , 18751 18756, doi:10.107 3/pnas.0905010106. 184. Ion, G.; Fajka - Boja, R.; Kovács, F.; Szebeni, G.; Gombos, I.; Czibula, Á.; Matkó, J.; Monostori, É. Acid sphingomyelinase mediated release of ceramide is essential to trigger the mitochondrial pathway of apoptosis by galectin - 1. Ce ll. Signal. 2006 , 18 , 1887 1896, doi:10.1016/j.cellsig.2006.02.007. 185. Tikhonenko, M.; Lydic, T.A.; Opreanu, M.; Li Calzi, S.; Bozack, S.; McSorley, K.M.; Sochacki, A.L.; Faber, M.S.; Hazra, S.; Duclos, S.; et al. N - 3 polyunsaturated Fatty acids prevent diabetic retinopathy by inhibition of retinal vascular damage and enhanced endothelial progenitor cell reparative function. PLoS One 2013 , 8 , e55177, doi:10.1371/journal.pone.0055177. 186. Novgorodov, S.A.; Gudz, T.I. Ceramide and mitochondria in ischemi c brain injury. Int. J. Biochem. Mol. Biol. 2011 , 2 , 347 61. 187. France - Lanord, V.; Brugg, B.; Michel, P.P.; Agid, Y.; Ruberg, M. Mitochondrial free radical signal in ceramide - J. Neurochem. 1997 , 69 , 1612 1621, doi:10.1046/j.1471 - 4159.1 997.69041612.x. 188. Fox, T.E.; Han, X.; Kelly, S.; Merrill, A.H.; Martin, R.E.; Anderson, R.E.; Gardner, T.W.; Kester, M. Diabetes alters sphingolipid metabolism in the retina: A potential mechanism of cell death in diabetic retinopathy. Diabetes 2006 , 55 , 3573 3580, doi:10.2337/db06 - 0539. 189. Jaffe, G.J.; Earnest, K.; Fulcher, S.; Lui, M.; Houston, L.L. Antitransferrin Receptor Immunotoxin Inhibits Proliferating Human Retinal Pigment Epithelial Cells. Arch. Ophthalmol. 1990 , 108 , 1163 1168, doi:10.100 1/archopht.1990.01070100119046. 190. Clayton, D.A.; Shadel, G.S. Purification of mitochondria by sucrose step density gradient centrifugation. Cold Spring Harb. Protoc. 2014 , 2014 , 1115 1117, doi:10.1101/pdb.prot080028. 191. Clayton, D.A.; Shadel, G.S. I solation of mitochondria from cells and tissues. Cold Spring Harb. Protoc. 2014 , 2014 , 1040 1041, doi:10.1101/pdb.top074542. 192. Lydic, T.A.; Busik, J. V.; Reid, G.E. A monophasic extraction strategy for the simultaneous lipidome analysis of polar and no npolar retina lipids. J. Lipid Res. 2014 , 55 , 1797 1809, doi:10.1194/jlr.D050302. 193. Byeon, S.K.; Lee, J.Y.; Lee, J.S.; Moon, M.H. Lipidomic profiling of plasma and urine from patients with Gaucher disease during enzyme replacement therapy by nanoflow l iquid chromatography - tandem mass spectrometry. J. Chromatogr. A 2015 , 1381 , 132 139, doi:10.1016/j.chroma.2015.01.004. 194. Haimi, P.; Uphoff, A.; Hermansson, M.; Somerharju, P. Software tools for analysis of mass spectrometric lipidome data. Anal. Chem. 2006 , 78 , 8324 8331, doi:10.1021/ac061390w. 195. Levitsky, Y.; Pegouske, D.J.; Hammer, S.S.; Frantz, N.L.; Fisher, K.P.; Muchnik, A.B.; Saripalli, A.R.; Kirschner, P.; Bazil, J.N.; Busik, J. V.; et al. Micro - respirometry of whole cells and isolated mitoch ondria. RSC Adv. 2019 , 9 , 33257 33267, doi:10.1039/c9ra05289e. 110 196. Novgorodov, S.A.; Wu, B.X.; Gudz, T.I.; Bielawski, J.; Ovchinnikova, T. V; Hannun, Y.A.; Obeid, L.M. Novel pathway of ceramide production in mitochondria: thioesterase and neutral ceramid ase produce ceramide from sphingosine and acyl - CoA. J. Biol. Chem. 2011 , 286 , 25352 62, doi:10.1074/jbc.M110.214866. 197. Erdreich - Epstein, A.; Tran, L.B.; Bowman, N.N.; Wang, H.; Cabot, M.C.; Durden, D.L.; Vlckova, J.; Reynolds, C.P.; Stins, M.F.; Groshe n, S.; et al. Ceramide signaling in fenretinide - induced endothelial cell apoptosis. J. Biol. Chem. 2002 , 277 , 49531 7, doi:10.1074/jbc.M209962200. 198. Perry, D.M.; Newcomb, B.; Adada, M.; Wu, B.X.; Roddy, P.; Kitatani, K.; Siskind, L.; Obeid, L.M.; Hannun, Y.A. Defining a Role for Acid Sphingomyelinase in the p38/Interleukin - 6 Pathway. J. Biol. Chem. 2014 , 289 , 22401 22412, doi:10.1074/jbc.M114.589648. 1 99. Novgorodov, S.A.; Riley, C.L.; Yu, J.; Keffler, J.A.; Clarke, C.J.; Van Laer, A.O.; Baicu, C.F.; Zile, M.R.; Gudz, T.I. Lactosylceramide contributes to mitochondrial dysfunction in diabetes. J. Lipid Res. 2016 , 57 , 546 562, doi:10.1194/jlr.M060061. 20 0. Kappler, L.; Li, J.; Häring, H.U.; Weigert, C.; Lehmann, R.; Xu, G.; Hoene, M. Purity matters: A workflow for the valid high - resolution lipid profiling of mitochondria from cell culture samples. Sci. Rep. 2016 , 6 , 1 10, doi:10.1038/srep21107. 201. Fra nko, A.; Baris, O.R.; Bergschneider, E.; Von Toerne, C.; Hauck, S.M.; Aichler, M.; Walch, A.K.; Wurst, W.; Wiesner, R.J.; Johnston, I.C.D.; et al. Efficient isolation of pure and functional mitochondria from mouse tissues using automated tissue disruption and enrichment with anti - TOM22 magnetic beads. PLoS One 2013 , 8 . 202. Larsen, S.; Nielsen, J.; Hansen, C.N.; Nielsen, L.B.; Wibrand, F.; Stride, N.; Schroder, H.D.; Boushel, R.; Helge, J.W.; Dela, F.; et al. Biomarkers of mitochondrial content in skeletal muscle of healthy young human subjects. J. Physiol. 2012 , 590 , 3349 60, doi:10.1113/jphysiol.2012.230185. 203. Kern, T.S.; Antonetti, D.A.; Smith, L.E.H. Pathophysiology of Diabetic Retinopathy: Contribution and Limitations of Laboratory Research. Ophtha lmic Res. 2019 , 62 , 196 202, doi:10.1159/000500026. 204. Babiychuk, E.B.; Atanassoff, A.P.; Monastyrskaya, K.; Brandenberger, C.; Studer, D.; Allemann, C.; Draeger, A. The targeting of plasmalemmal ceramide to mitochondria during apoptosis. PLoS One 2011 , 6 , doi:10.1371/journal.pone.0023706. 205. Kowluru, R.A. Diabetic retinopathy, metabolic memory and epigenetic modifications. Vision Res. 2017 , 139 , 30 38, doi:10.1016/j.visres.2017.02.011. 206. Alivand, M.R.; Soheili, Z.S.; Pornour, M.; Solali, S.; Sabo uni, F. Novel Epigenetic Controlling of Hypoxia Pathway Related to Overexpression and Promoter Hypomethylation of TET1 and TET2 in RPE Cells. J. Cell. Biochem. 2017 , 118 , 3193 3204, doi:10.1002/jcb.25965. 207. Desjardins, D.; Liu, Y.; Crosson, C.E.; Ablonczy, Z. Histone Deacetylase Inhibition Restores Retinal Pigment Epithelium Function in Hyperglycemia. PLoS One 2016 , 11 , 1 16, doi:10.1371/journal.pone.0162596. 111 208. Dolinko, A.H.; Chwa, M.; Atilano, S.R.; Kenn ey, M.C. African and Asian Mitochondrial DNA Haplogroups Confer Resistance Against Diabetic Stresses on Retinal Pigment Epithelial Cybrid Cells In Vitro. Mol. Neurobiol. 2020 , 57 , 1636 1655, doi:10.1007/s12035 - 019 - 01834 - z. 209. Peng, Q.H.; Tong, P.; Gu, L .M.; Li, W.J. Astragalus polysaccharide attenuates metabolic memory - triggered ER stress and apoptosis via regulation of miR - 204/SIRT1 axis in retinal pigment epithelial cells. Biosci. Rep. 2020 , 40 , 1 15, doi:10.1042/BSR20192121. 210. Roy, S.; Sala, R.; C agliero, E.; Lorenzi, M. Overexpression of fibronectin induced by diabetes or high glucose: Phenomenon with a memory. Proc. Natl. Acad. Sci. U. S. A. 1990 , 87 , 404 408, doi:10.1073/pnas.87.1.404. 211. Duraisamy, A.J.; Mohammad, G.; Kowluru, R.A. Mitochond rial fusion and maintenance of mitochondrial homeostasis in diabetic retinopathy. Biochim. Biophys. Acta - Mol. Basis Dis. 2019 , 1865 , 1617 1626, doi:10.1016/j.bbadis.2019.03.013. 212. Qi, X.; Mitter, S.K.; Yan, Y.; Busik, J. V; Grant, M.B.; Boulton, M.E. Diurnal Rhythmicity of Autophagy Is Impaired in the Diabetic Retina. Cells 2020 , 9 , 1 17, doi:10.3390/cells9040905. 213. Weinbach, E.C.; Costa, J.L.; Nelson, B.D.; Claggett, C.E.; Hundal, T.; Bradley, D.; Morris, S.J. Effects of tricyclic antidepressant drugs on energy - linked reactions in mitochondria. Biochem. Pharmacol. 1986 , 35 , 1445 1451, doi:10.1016/0006 - 2952(86)90108 - 5.