LIPIDOMIC INSIGHTS INTO NEURODEGENERATIVE DISEASES AND SYSTEMIC LUPUS ERYTHEMATOSUS USING MASS SPECTROMETRY By Seyedeh Elham Pourmand A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemistry – Doctor of Philosophy 2024 ABSTRACT This dissertation explores the intricate roles of polyunsaturated fatty acid (PUFA) metabolites, particularly oxylipins, in the pathology of neurodegenerative diseases and systemic lupus erythematosus (SLE), emphasizing their significance in disease diagnostics and therapeutics. We examined oxylipins' biosynthesis and physiological impacts, focusing on metabolic pathways, primarily cytochrome P450 and epoxide hydrolase (CYP-EH) metabolism. This work underscores the critical role of oxylipins as potential biomarkers, positioning lipidomics as a transformative frontier in biomedical research. Using the aging model organism Caenorhabditis elegans (C. elegans), we developed an analytical methodology to investigate the interplay between aging, neurodegenerative disease, and CYP-EH metabolism of PUFA. The method employing solid phase extraction, high-performance liquid chromatography coupled with tandem mass spectrometry (SPE-HPLC-MS/MS) was developed and validated to quantify nanomolar levels of potent oxylipins accurately and precisely in C. elegans across different ages and under various treatments. This analytical methodology will serve as a tool to significantly enhance our understanding of the molecular relationship between PUFA CYP-EH metabolism and aging-related diseases. Later, we investigate the impact of SLE pathogenesis on PUFA metabolism with a focus on the CYP-EH metabolic pathway. Through serum analysis from SLE and healthy Western populations, with the help of targeted lipidomics and comprehensive machine learning analysis, we found EH metabolites, such as 14,15-DiHETrE, to be significantly downregulated in subjects with SLE across different sexes, races, and age groups. Classification and regression machine learning analysis further revealed that some CYP-EH-derived oxylipins are important predictors of SLE, indicating a link with disease status and severity and identifying potential biomarkers. Studies show oxidative stress plays a critical role in both neurodegenerative disease and SLE. Lastly, we aimed to further understand the mechanisms behind oxidative stress and its connection with lipid metabolism. We investigated the impact of chemically induced oxidative stress on lipid and oxylipin profiles using C. elegans as an animal model. Methodologies include targeted LC- MS/MS and untargeted high-resolution mass spectrometry (HR-MS) analysis to monitor changes in oxylipin and lipid concentrations under oxidative stress conditions. The upregulation of dihydroxy PUFA metabolites alongside unchanged epoxy metabolites suggests enzyme regulation under oxidative stress. The findings of this pilot study reveal substantial alterations in oxylipin profiles and lipid concentrations, indicating a complex interplay between lipid metabolism and oxidative stress responses. This dissertation underscores the essential role of PUFAs and their CYP-EH metabolites in understanding disease mechanisms. It demonstrates how lipidomics, coupled with advanced analytical techniques such as targeted and untargeted mass spectrometry, can effectively identify biomarkers and therapeutic targets, acquire critical insights into cellular conditions and disease processes and thereby advance personalized medicine. Copyright by SEYEDEH ELHAM POURMAND 2024 NULLIUS IN VERBA v ACKNOWLEDGMENTS I am profoundly grateful to my loving parents, Kazem and Farokhlagha, who have always encouraged me to follow my heart and dreams, and I am equally thankful to my siblings, Mohammad and Mohana, for their unwavering love and companionship. The distance over the past six years has been challenging for all of us, but I hope my achievements reflect the sacrifices made and make you all happy and proud. I extend my heartfelt thanks to my advisor, Prof. Sing Lee, whose patience and support have greatly contributed to my development in science. I immensely appreciate my second reader, Prof. Jones, whom I learned a lot from, and his support and positive attitude were always heartwarming and encouraging. My gratitude extends to Prof. Bernard and Prof. Swain, who have supported me as members of my Ph.D. committee. I also want to acknowledge my current and former lab mates—Maris, Fan, Devon, Olivia, Morteza, Derek, Jennifer, Tommy, Sara, Sachini, Angel, Ali, and Megha. Their collaboration has enriched my laboratory experience, making it both enjoyable and rewarding. Special thanks go to all my considerate and supportive friends in the USA and Iran. I am forever blessed and grateful to have such wonderful individuals in my life, and I look forward to more laughter and adventure. I have to express my gratitude to my kitten, Penelope Cruz-Pourmand, who played a crucial role in maintaining my mental health and brought immense joy and love into my daily life. I dedicate my Ph.D. to my parents who have inspired me the most. Additionally, this is for my grandmothers, who were denied the opportunity to pursue their education as kids because of their gender, and for all the brave and exceptional women who fought and sacrificed during the Iranian freedom movement in 2022, specially Hanane Kia from my hometown Noshahr. Their courage has motivated me every day, being their voice is my identity till the last day of my life. vi TABLE OF CONTENTS LIST OF ABBREVIATIONS ................................................................................................... viii CHAPTER 1. INTRODUCTION TO THE IMPORTANCE OF POLYUNSATURATED FATTY ACIDS AS BIOMARKERS AND ANALYTICAL TOOLS TO DETECT AND QUANTIFY THEM ..............................................................................................................................................1 BIBLIOGRAPHY .....................................................................................................................71 CHAPTER 2. QUANTITATIVE PROFILING METHOD FOR OXYLIPINS IN CAENORHABDITIS ELEGANS BY LIQUID CHROMATOGRAPHY COUPLED WITH TANDEM MASS SPECTROMETRY ..........................................................................................86 BIBLIOGRAPHY ...................................................................................................................113 CHAPTER 3. THE EFFECTS OF SYSTEMIC LUPUS ERYTHEMATOSUS (SLE) ON UNSATURATED LIPIDS AND THEIR OXIDATIVE METABOLISM IN THE WESTERN POPULATION ............................................................................................................................118 BIBLIOGRAPHY ...................................................................................................................170 CHAPTER 4. INVESTIGATING THE EFFECT OF TBHP-INDUCED OXIDATIVE STRESS ON OXYLIPIN AND GLOBAL LIPID PROFILE USING TRIPLE QUADRUPOLE AND HR- MS ..............................................................................................................................................176 BIBLIOGRAPHY ...................................................................................................................212 APPENDIX 1: SUPPORTING INFORMATION FOR CHAPTER 2 .......................................217 APPENDIX 2: SUPPORTING INFORMATION FOR CHAPTER 3 .......................................235 vii LIST OF ABBREVIATIONS 17,18-EEQ 17,18-Epoxyeicosatetraenoic acid 14,15-EEQ 14,15-Epoxyeicosatetraenoic acid 11,12-EEQ 11,12-Epoxyeicosatetraenoic acid 8,9-EEQ 5,6-EEQ 8,9-Epoxyeicosatetraenoic acid 5,6-Epoxyeicosatetraenoic acid 17,18-DiHETE 17,18-Dihydroxyeicosatetraenoic acid 14,15-DiHETE 14,15-Dihydroxyeicosatetraenoic acid 11,12-DiHETE 11,12-Dihydroxyeicosatetraenoic acid 8,9-DiHETE 8,9-Dihydroxyeicosatetraenoic acid 5,6-DiHETE 5,6-Dihydroxyeicosatetraenoic acid 12,13-EpOME 12,13-Epoxyoctadecamonoenoic acid 9,10-EpOME 9,10-Epoxyoctadecamonoenoic acid 12,13-DiHOME 12,13-Dihydroxyoctadecamonoenoic acid 9,10-DiHOME 9,10-Dihydroxyoctadecamonoenoic acid 14,15-EET 14,15-Epoxyeicosatrienoic acid 11,12-EET 11,12-Epoxyeicosatrienoic acid 8,9-EET 5,6-EET 8,9-Epoxyeicosatrienoic acid 5,6-Epoxyeicosatrienoic acid 14,15-DiHET 14,15-Dihydroxyeicosatrienoic acid 11,12-DiHET 11,12-Dihydroxyeicosatrienoic acid 8,9-DiHETrE 8,9-Dihydroxyeicosatrienoic acid 5,6-DiHET 5,6-Dihydroxyeicosatrienoic acid viii 14,15-EED 14,15-Epoxyeicosadienoic acid 11,12-EED 11,12-Epoxyeicosadienoic acid 8,9-EED 8,9-Epoxyeicosadienoic acid 14,15-DHED 14,15-Dihydroxyeicosadienoic acid 11,12-DHED 11,12-Dihydroxyeicosadienoic acid 8,9-DHED 8,9-Dihydroxyeicosadienoic acid 15,16-EpODE 15,16-Epoxyoctadecadienoic acid 12,13-EpODE 12,13-Epoxyoctadecadienoic acid 9,10-EpODE 9,10-Epoxyoctadecadienoic acid 15,16-DiHODE 15,16-Dihydroxyoctadecadienoic acid 12,13-DiHODE 12,13-Dihydroxyoctadecadienoic acid 9,10-DiHODE 9,10-Dihydroxyoctadecadienoic acid 13-HODE 13-Hydroxyoctadecadienoic acid 9-HODE 20-HETE 19-HETE 15-HETE 12-HETE 9-HETE 8-HETE 5-HETE 9-Hydroxyoctadecadienoic acid 20-Hydroxyeicosatetraenoic acid 19-Hydroxyeicosatetraenoic acid 15-Hydroxyeicosatetraenoic acid 12-Hydroxyeicosatetraenoic acid 9-Hydroxyeicosatetraenoic acid 8-Hydroxyeicosatetraenoic acid 5-Hydroxyeicosatetraenoic acid 15(S)-HETE 15(S)-Hydroxyeicosatetraenoic acid 20-HEPE 20-Hydroxyeicosapentaenoic acid ix 18-HEPE 15-HEPE 12-HEPE 8-HEPE 5-HEPE LA DGLA AA α-LA EPA DGA CYP EH 18-Hydroxyeicosapentaenoic acid 15-Hydroxyeicosapentaenoic acid 12-Hydroxyeicosapentaenoic acid 8-Hydroxyeicosapentaenoic acid 5-Hydroxyeicosapentaenoic acid Linoleic Acid Dihomo-gamma-linolenic Acid Arachidonic Acid Alpha-Linolenic Acid Eicosapentaenoic Acid Docosahexaenoic Acid Cytochrome P450 Enzyme Epoxide Hydrolase Enzyme C. elegans Caenorhabditis elegans x CHAPTER 1. INTRODUCTION TO THE IMPORTANCE OF POLYUNSATURATED FATTY ACIDS AS BIOMARKERS AND ANALYTICAL TOOLS TO DETECT AND QUANTIFY THEM 1 ABSTRACT This chapter underscores the pivotal role of polyunsaturated fatty acid (PUFA) metabolites, particularly oxylipins, in the pathology of various diseases and highlights their significance in medical research and diagnostics. A detailed historical analysis of dietary shifts in omega-3 and omega-6 fatty acid ratios exposes potential connections to prevalent health issues. It outlines the metabolic transformations accompanying changes in human diet and lifestyle. The biosynthesis and physiological impacts of oxylipins are explored, focusing on metabolic pathways such as cyclooxygenase (COX) and lipoxygenase (LOX), as well as the roles of cytochrome P450-derived oxylipins and epoxide hydrolase. Moreover, the chapter delves into the identification and evolution of bioactive chemicals and biomarkers, emphasizing the critical role of oxylipins as valuable biomarkers with potential clinical applications. This narrative sets the stage for the emerging field of lipidomics, which is positioned as a transformative frontier in biomedical research. By integrating advanced analytical techniques like mass spectrometry with sophisticated separation technologies, the chapter not only reviews the current capabilities in lipid profiling but also outlines the perspective shifts from theoretical frameworks to empirical studies aimed at unearthing novel biomarkers and therapeutic strategies. Ultimately, this exploration of lipidomics is depicted as a journey to enhance our understanding of biological systems and to harness this knowledge for the advancement of personalized medicine. 2 1. Diversity roles of lipids in biological systems and structural complexity Exploring the nature of lipids is challenging due to their extraordinary diversity in size, structure, and chemical composition. As indispensable elements of biological systems, lipids play important roles in a host of different physiological processes. Cells harbor an extensive array of distinct lipids—amounting to tens of thousands each fulfilling specific functions and interacting with myriad proteins to manage their metabolism and transport. As fundamental components of cell architecture, lipids are essential to the construction of various membrane structures, such as the plasma membrane, nuclear envelope, endoplasmic reticulum (ER), and Golgi apparatus. Lipids are vital for the operation of transport vesicles, including endosomes and lysosomes1,2. The variations in lipid composition across different organelles, cell types, and tissues suggest unique lipid elements for various cellular functions. The LIPID MAPS Structure Database (LMSD) provides details on more than 30,000 lipid structures and their properties, while the LIPID MAPS Proteome Database (LMPD) includes data on over 1,200 lipid-related proteins for both humans and mice3. Complicating matters further, the literature employs many names for these structures, including lipid bodies, lipid droplets, lipid particles, lipid-protein particles, and oil bodies. There appears to be no distinct characteristic that sets apart these entities with different names. Despite their evident significance and crucial roles, lipids have not undergone as thorough investigations as proteins4. Altogether, one straightforward way to define lipids is based on their solubility, lipids usually have poor solubility in polar solvents like water but appreciable solubility when it comes to non-polar solvents such as chloroform, and alcohols. Solubility refers to the maximum quantity of a substance that can dissolve in a given amount of another substance, forming a saturated solution5. Interactions of lipids, including sterol esters and glycerolipids, with other lipids are non-covalent. The hydrophobic regions of proteins through their hydrocarbon 3 chains, which is evident in structures such as adipose tissue fat and albumin-fatty acid complexes. Similarly, while polar lipids such as glycerophospholipids and sphingolipids can form hydrogen bonds and electrostatic interactions with proteins due to their polar head groups, it is also the hydrophobic tails of these lipids that play a crucial role by interacting with the hydrophobic parts of proteins. These interactions are integral to the organization and function of cellular membranes, mitochondria, endoplasmic reticulum, and serum lipoprotein complexes, where the hydrophobic tails align and associate with the proteins' hydrophobic domains, facilitating membrane integrity and protein functionality. Hydrogen bonding and electrostatic are predominant interactions at the lipid/water interface, while hydrophobic interactions govern the ordering of membranes in the lipid bilayer core 6,7. The critical micelle concentrations (CMC) can vary significantly among lipids. For example, phospholipids like phosphatidylcholine typically have CMC values in the low micromolar range. Glycolipids and sphingolipids can have CMC values in the nanomolar to micromolar range, depending on their structure8. In biological membranes, polar lipids form lipid bilayer or hexagonal-II structures, contributing to the amphipathic balance, with the bilayer being the typical arrangement in living cells. The hexagonal-II structure involves lipid polar group- formed water tubes packed into a two-dimensional hexagonal array. The amphipathic balance strongly leans toward hydrocarbon affinity, and the polar group is typically less hydrated than in bilayer-forming lipids. Moreover, numerous lipids possess an 'amphiphilic' nature, featuring molecular components that interact with both water and a hydrophobic moiety within the same molecule. This amphiphilic property holds significant relevance in biological structures, contributing to the versatile roles of lipids in various contexts. Lipids form cell membranes in biological systems where phospholipids arrange themselves to create a bilayer due to their amphiphilic nature. This lipid bilayer serves as 4 a barrier that separates the internal cellular environment from the external surroundings, enabling essential cellular processes and interactions9. The hydrophobic traits of lipids are commonly derived from fatty acids esterified with glycerol or forming amide linkages with sphingosine, while others originate from the steroid ring system. This dual affinity for both hydrophilic and hydrophobic environments allows lipids to play essential roles in the formation and stability of biological membranes and contributes to their functionality in diverse biological and culinary contexts6,10. 2. Different classes of lipids Historically, lipid biology has been predominantly studied by chemists, biochemists, and biophysicists, leading to a nomenclature grounded in chemical scaffolds and connections. In contrast to proteins, which are primarily categorized based on their biological functions (such as kinase or acyl transferase), lipids are typically classified based on their chemical structures. This has resulted in potential confusion, as lipids with similar chemical compositions may bear different names (e.g., diacylglycerol and ceramide). In contrast, others with similar names may exhibit distinct structures (C14 and C26 ceramides) and, presumably, diverse functions4. Another example is eicosatetraenoic acid (20:4), which can have different isomers, including omega-6 (20:4) arachidonic acid (AA) or omega-3 (20:4) polyunsaturated fatty acid juniperonic acid (JuA). They are both 20-carbon polyunsaturated fatty acids (PUFA), but they differ in the positioning of their double bonds. This seemingly minor structural difference has significant functional consequences. AA is a precursor for the synthesis of eicosanoids, a class of potent signaling molecules involved in inflammation, blood clotting, and other physiological processes. In contrast, the other one is not a good substrate for eicosanoid synthesis11. By adopting a more structural perspective on lipids, as biologists do with proteins, researchers may be able to better elucidate the 5 complex roles these molecules play in various physiological and pathological processes, leading to the development of more targeted and effective interventions for lipid-related diseases and disorders4. With this structural understanding of lipids in mind, the following sections will explore the diverse classes of these biomolecules and their unique functions. 2.1. Triglycerides Triglycerides are essential lipid compounds found in the human body, composed of a glycerol molecule esterified to three fatty acid chains of varying length and composition (Figure 1) 12. They are the main constituents of vegetable fat and body fat, serving as a major energy source for the body, providing 9 kilocalories per gram of free fatty acids. Triglycerides can be derived from dietary sources, such as oils, meats, and dairy products, or synthesized in the body from excess carbohydrates, alcohol, and fats. Triglycerides serve as the primary energy storage molecules within the body. They are synthesized in the liver and intestines from fatty acids and then transported through the bloodstream by chylomicrons QM (intestinal origin) or VLDL (hepatic origin) in plasma. The catabolism of these triglycerides is governed by the lipoprotein lipase enzyme complex. Moreover, their clearance from the plasma is facilitated by liver receptors, specifically the LDL receptor (LDLR) and the LDL receptor-related protein 1 (LRP-1)12. These lipid molecules are classified based on the saturation of the fatty acid chains, which can be saturated or unsaturated, and their structural comparison between the chains is heterogeneous in nature. For example, in saturated triglycerides, all three fatty acid chains lack double bonds. These triglycerides are typically found in animal fats such as butter and lard, and some tropical oils like coconut oil. Due to their saturated nature, they are solid or semi-solid at room temperature and have a higher melting point. Meanwhile, in unsaturated triglycerides, one or more of the fatty acid chains contain double bonds. These can be further classified based on the number of double bonds 6 present. Monounsaturated triglycerides contain at least one fatty acid chain with a single double bond. An example is oleic acid, predominantly found in olive oil and avocados. Polyunsaturated triglycerides have fatty acid chains with two or more double bonds. Common examples include linoleic acid (LA) which is abundant in sunflower oil and flaxseed oil, respectively. Double bonds not only affect the physical state of these lipids at room temperature making unsaturated fats generally liquid but also impact their nutritional properties and roles in health13. Triglycerides play a crucial role in energy storage, insulation, and protection of vital organs, and they are transported in the bloodstream within lipoprotein particles14. Elevated levels of triglycerides have been associated with an increased risk of cardiovascular disease, making monitoring and management an essential aspect of maintaining health15. Figure 1.1. Example of triglyceride structure 2.2. Glycolipids Glycolipids are a class of lipids with a carbohydrate moiety attached to a glycosidic bond (Figure 1.2). They are ubiquitous in various organisms, including bacteria, plants, and animals, and are essential components of cellular membranes, where they play a crucial role in maintaining membrane stability and facilitating cell-cell communication. Glycolipids play a role as regulators of signal transduction and receptor anchors for proteins. They are primarily located on the outer leaflet of cellular membranes. For instance, glycolipid monosialodihexosylganglioside (GM3), plays a role in influencing the function of receptor tyrosine kinases (RTKs), including the 7 OOOOOO epidermal growth factor receptor (EGFR). By interfering with the process of EGFR autophosphorylation, GM3 can contr. ol the downstream signaling pathways that govern cellular proliferation and differentiation16. Glycolipids are classified based on their structure, with two main classes being glycosphingolipids, which are composed of a sphingolipid or a glycerol group with one or two fatty acids, and glycoglycerolipids, which are comprised of a glycerol group with one or two fatty acids. Glycolipids are synthesized in the Golgi-apparatus, and the dysfunction of glycolipid metabolism has been linked to several diseases, such as the development of type 2 diabetes mellitus. Individuals with type 2 diabetes mellitus are at greater risk for heart damage and vascular diseases, highlighting the importance of these molecules in maintaining cellular hemostasis 17. Glycolipids, especially glycosphingolipids (GSLs) like gangliosides, are primarily degraded through the lysosomal pathway involving specific hydrolytic enzymes. Cofactor proteins "activator proteins” facilitate the degradation of micelle-forming GSLs by water-soluble lysosomal enzymes18. Figure 1.2. Example of glycolipid structure 2.3. Steroids Steroids are intricate lipophilic molecules that govern various physiological functions throughout an organism's lifespan (Figure 1.3). Synthesized from cholesterol in specialized endocrine cells within the adrenal gland, ovary, and testis, steroid hormones like cortisol, aldosterone, estradiol, and testosterone are released into circulation as needed. 8 OOOOOOHHOHOOHO Figure 1.3. Example of sterol lipid structure Once these hormones enter the cells, they move freely to activate nuclear receptors inside the cells. These receptors function as multi-domain, ligand-dependent transcriptional regulators within the cell nucleus. The activated nuclear receptors then modulate the expression of hundreds to thousands of specific target genes19. Steroids, modified structural forms of cholesterol are critical for maintaining the cell membrane structure and as signaling molecules., Several of steroids serve as endogenous endocrine hormones. Subsequent modifications of the steroid structure and resultant functions occur in various tissues and organs, including the liver, skin epidermis, brain, and prostate19. The major sites of steroid inactivation and catabolism are the liver and, to a lesser extent, the kidney. Mechanisms of steroid catabolism include reduction, oxidation, hydroxylation, and conjugation with sulfuric or glucuronic acid to create water-soluble steroid sulfates and glucuronides excreted in the urine20. 2.4. Phospholipids Phospholipids, constituting a distinct subclass of lipids, boast a unique chemical structure featuring a hydrophilic phosphate head group and hydrophobic lipid tails. This unique structure bestowing them with distinctive chemical properties that are pivotal for cellular membranes. Structurally, phospholipids comprise a glycerol or sphingosine backbone, two fatty acid chains, and a phosphate-containing head group. This structural duality enables the formation of bilayers, with 9 HOOH hydrophobic tails oriented away from water and hydrophilic heads interacting with aqueous environments (Figure 1.4). Figure 1.4. Example of phospholipid structure The diversity of phospholipids is attributed to variations in the head group and fatty acid composition. Primarily, phospholipids serve as structural components of cell membranes, constituting the lipid bilayer that acts as a selectively permeable barrier, overseeing the regulation of ions and molecules entering and exiting cells, like Phosphatidylcholine (PC). PC is the most abundant phospholipid in cell membranes, making up around 50% of the total phospholipids, and is essential for the structure and function of cell membranes21. Furthermore, their involvement in cellular signaling is notable, as the hydrophilic heads can interact with water-soluble signaling molecules; for example, phosphatidylinositol 4,5-bisphosphate (PIP2) is a phospholipid found in the inner leaflet of the plasma membrane. When cell surface receptors are activated, PIP2 can be broken down by the enzyme phospholipase C (PLC) to produce the second messengers inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 causes the release of calcium from intracellular stores, while DAG activates protein kinase C, initiating a series of signaling events. 21. Specific phospholipids also act as precursors to signaling molecules, including second messengers engaged in intracellular communication. As previously mentioned, AA released from phospholipids produces eicosanoids, which are potent signaling molecules. To comprehend cellular physiology fully, it is imperative to grasp the dynamic interplay between different phospholipid types and their arrangement within membranes. Moreover, disruptions in 10 OOOOOPHOOOH phospholipid metabolism have been linked to various diseases, underscoring the significance of ongoing research in this field. This exploration aims to delve deeper into the intricacies of phospholipid biology, elucidating their diverse roles and significance in the intricate machinery of living cells22. The catabolism of phospholipids in lysosomes involves the sequential removal of the polar head group and fatty acyl chains by phospholipases and lipases. Deficiencies in the enzymes involved in phospholipid catabolism can lead to lysosomal storage disorders like Batten disease, where phospholipids and their metabolites accumulate in lysosomes23. 2.5. Sphingolipids Sphingolipids are a distinct class of lipids with a shared sphingoid base backbone, which is N- acylated with various fatty acids to form a range of structural classes, including sphingoid bases and their derivatives, ceramides, and complex sphingolipids (Figure 1.5). The foundational sphingoid bases, primarily sphingosine, and sphinganine in mammals, are amino alcohols that hold hydroxyl groups as hydrophilic head groups and are critical for synthesizing ceramides and complex sphingolipids. Beyond their structural roles, sphingolipids are potent bioactive lipid mediators in many cellular processes. They influence cell migration24, proliferation25, and the regulation of apoptosis26. For example, sphingosine-1-phosphate (S1P) is a sphingolipid-derived signaling molecule integral to pathways that govern cell survival and growth. In membrane architecture, sphingomyelins, a type of sphingolipid, are paramount in ensuring membrane stability and insulation, especially in nerve cells, where they are essential to the integrity of myelin sheaths. Cerebrosides, another sphingolipid variant, are vital for the proper functioning of the nervous system by maintaining neuronal cell membrane integrity. Additionally, gangliosides, complex sphingolipids predominantly located in the nervous system, play critical roles in cell-to- cell recognition and signaling, which are fundamental for neuronal function27,28. 11 Figure 1.5. Example of sphingolipid structure 2.6. Ceramides Ceramides are a crucial class of sphingolipids with significant roles in various biological processes. They are essential intermediates in the biosynthesis and metabolism of all sphingolipids consisting of a long-chain or sphingoid base linked to a fatty acid via an amide bond (Figure 1.6). These lipids are vital components of cellular membranes, with unique compositions and functions, particularly in the skin 29. Ceramides exhibit structural diversity originally from fatty acid chain length, degree of saturation, and the presence of hydroxyl groups or double bonds in the sphingoid base. Ceramides are inherently hydrophobic due to their hydrocarbon chains, contributing to their insolubility in aqueous solutions. C18 ceramide is the most abundant form in the nervous system, impacting membrane rigidity and organization. The tight packing ability of ceramides in plasma membranes enhances molecular rigidity and influences membrane curvature, affecting cellular processes like vesiculation and fusion, emphasizing the importance of ceramides in cellular signaling, skin barrier function, and their involvement in various pathophysiological conditions when sphingolipid homeostasis is disrupted29. Figure 1.6. Example of ceramide structure 12 OHNH2OHNHOHOHO 2.7. Fatty Acids Fatty acids, the simplest lipid type, consist of a lengthy hydrocarbon chain with a carboxyl group (-COOH) at one end and can be saturated, monounsaturated, or polyunsaturated, with one or more double bonds between carbon atoms. PUFAs, crucial dietary components for humans and mammals, fall into the categories of omega-9, omega-6, and omega-3, depending on the position of the terminal double bond relative to the methyl end of the fatty acids. For instance, omega-6 linolenic acid (LA,18:2) has double bonds at carbons 9 and 12, while omega-3 alpha-linolenic acid (ALA, 18:3) has double bonds at carbons 9, 12, and 15. ALA is an essential fatty acid that cannot be synthesized by the human body and must be obtained through the diet. It is found in plant sources such as flaxseed, chia seeds, and walnuts. These essential fatty acids can undergo elongation and desaturation processes, yielding a spectrum of 20- and 22-carbon omega-6 like AA and omega-3 fatty acids like eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA)30. EPA and DHA are long-chain omega-3 fatty acids that can be synthesized from ALA, but the extent of this conversion is modest and controversial. For example, one study reported a 15% conversion31, whereas another study found 0.2%32; both reported that the conversion to DHA was much less than that to EPA. A notable disparity was observed between males and females regarding their reaction to a diet enriched with ALA. Females exhibited a considerably higher elevation in the EPA levels within plasma phospholipids, averaging an increase of +2.0% of total fatty acids, following six months of consuming an ALA-abundant diet, in contrast to males who showed a mean increase of just +0.7% 33. The efficiency of this conversion process plays a crucial role in maintaining adequate levels of EPA and DHA in the body, especially for those who do not consume sufficient amounts of these fatty acids in their diet. Moreover, this conversion rate is notably higher in infants and children but decreases as individuals age. Since humans cannot 13 convert omega-6 to omega-3 fatty acids, the dietary intake determines the crucial ratio between the two34. Despite their structural similarity, omega-6 and omega-3 fatty acids exhibit divergent biological functions, which will be explained later35. Palmitic acid (16:0) is one of the most abundant saturated fatty acids in the human body. It can make up around 20-30% of total fatty acids. Oleic acid (18:1, n-9), is a monounsaturated fatty acid and is commonly found in various tissues. It can constitute about 10-20% of total fatty acids. LA is an essential PUFA comprising 10-20% of total fatty acids. ALA is another essential PUFA typically found in lower amounts than LA, constituting around 0.5-1% of total fatty acids. Stearic acid (18:0), another saturated fatty acid, can make up approximately 10-15% of total fatty acids36–38 (Figure 1.7) depicts some of the most prevalent classes of fatty acids. Figure 1.7. Representative structures of fatty acids. Different subclasses of fatty acyls include straight chain, methyl branched, unsaturated, hydroperoxy, hydroxy, oxo, epoxy, methoxy, cyano, thia, carbocyclic, heterocyclic, mycolic, and dicarboxylic fatty acids 14 3. Function of lipids Comprehending the functions of lipids in biological processes poses numerous challenges. It is essential to 1) identify the involvement of specific lipid families and species in the given process; 2) visualize these lipids within pertinent cellular compartments or structures, ideally in live cells to facilitate turnover assessment; and 3) perturb lipid levels to conduct phenotypic analyses for functional insights4. (Figure 1.8) represents some of the most important functions of lipids in cellular biology. 3.1. Storage and provision of energy Lipid droplets (LDs) are widespread organelles found across various organisms, including animals, plants, fungi, and bacteria10. LDs are designed to store energy in the form of neutral lipids, like triglycerides and sterol esters. Neutral lipids are in a hydrophobic core of LDs, which stores energy and can act as lipid signaling molecules. A monolayer of polar lipids, mostly phospholipids, envelops this core, providing stability and isolating it from the cell's aqueous environment39. The synthesis of neutral lipids occurs in the endoplasmic reticulum (ER) membrane through enzymes like diglyceride acyltransferase 1 (DGAT1); other than the synthesis of neural lipids, several additional processes are involved in the biogenesis of LDs, including the accumulation of neutral lipids within the ER membrane, the coalescence of these lipids into larger droplets, and eventually leading to the generation of LDs that emerge as organelles either partially or entirely distinct from the ER40. LDs can enlarge by synthesizing neutral lipids locally or in specific tissues such as adipocytes through fusion. The neutral lipids stored within LDs can be broken down when the body requires energy. There are two main pathways for that purpose: lipolysis or lipophagy. Lipolysis involves lipid droplet-associated enzymes like adipose triglyceride lipase (ATGL), whereas lipophagy utilizes lysosomal acid lipase (LAL) to act on LDs that are transported to 15 autolysosomes through autophagy. The fatty acids released from the breakdown of neutral lipids are then used for energy production through mitochondrial or peroxisomal beta-oxidation within the tricarboxylic acid (TCA) cycle. This cycle is a key metabolic pathway that occurs in the mitochondrial matrix and consists of a series of enzyme-catalyzed reactions that oxidize acetyl- CoA, derived from sources such as pyruvate, fatty acids, and amino acids, to generate energy stored in the form of ATP41,42. LDs serve a crucial function in maintaining energy homeostasis within the cell. This homeostatic role is vital for the regular functioning of both individual cells and the entire organism, and disruptions in this balance are associated with various human pathologies, including obesity, diabetes, cardiovascular disease, and fatty liver disease43,44. 3.2. Membrane lipid layer formation Cellular membranes, essential for dynamic cellular processes, are amphipathic lipids and protein assemblies. Their fluid nature allows for rotational, translational, and trans-bilayer lipid movements45. The composition of lipids in the plasma membrane (PM) significantly shapes its properties. Asymmetry between membrane leaflets is evident, with the PM cytoplasmic leaflet enriched in phosphatidylethanolamine (PE) and phosphatidylserine (PS), while the outer leaflet is rich in sphingolipids46. Membrane fluidity, even in the presence of saturated lipids, is modulated through integrated sterols disrupting acyl chain packing. Being inflexible, sterols can increase rigidity when incorporated into flexible unsaturated lipid bilayers47. The distribution of cholesterol within the membranes of different organelles within a cell is not uniform. This non-uniform distribution, or asymmetry, of cholesterol, affects how each organelle's membrane behaves regarding its structure, fluidity, and function48. Membrane cholesterol's impact on G protein- coupled receptors (GPCRs) has been established, while its effects on other membrane-bound components have not. This is exemplified by the discovery of cholesterol within the structure of 16 the beta-2 adrenergic receptor and the identification of a common cholesterol binding pattern in a significant subset of GPCRs49. In vivo studies using cholesterol analogs have shown both direct and indirect interactions between cholesterol and GPCRs; cholesterol has been found to bind directly to the oxytocin receptor, whereas it appears to affect cholecystokinin receptor activity indirectly by altering the properties of the membrane49. These findings demonstrate the delicate balance between lipid composition and physical attributes in cellular membranes underscores their pivotal role in orchestrating fundamental cellular functions. 3.3. Membrane lipid composition and signaling The lipid constituents of cellular membranes are pivotal in maintaining cellular homeostasis. They dictate membrane characteristics such as fluidity, flexibility, and phase partitioning, which are essential for proper cell signaling and protein interactions. Macrodomain formation within the membrane, a product of lipid phase separation, results in the segregation of signaling components, which profoundly influences their movement and interactions within the membrane, consequently shaping signal transduction processes45. The specific arrangement of phospholipids within the plasma membrane is essential for recruiting and activating proteins that are crucial for signaling processes. Variations in phospholipid types can lead to different protein binding affinities and activities, as illustrated by the sensitivity of the small GTPase K-Ras activity affected by various lipid environments. Through molecular dynamics simulations, it has been revealed that the association of K-Ras with specific plasma membrane phospholipids influences its spatial orientation and access to the catalytic domain, affecting its signaling capabilities. These simulations provide insight into the molecular interactions with phospholipids that dictate K-Ras activity50. K-Ras activity is also modulated by its nucleotide state, which influences its lipid membrane affinity. This nucleotide-dependent affinity results in varied signaling outcomes, adding 17 another layer of regulation to K-Ras's role in cellular signaling51. Lipids act as dynamic platforms that anchor and regulate proteins, facilitating the complex web of interactions necessary for downstream signaling. Their diverse structures provide a basis for varied protein-lipid and protein- protein interactions within signaling pathways. Specifically, lipids can modulate the affinity of proteins to their endogenous ligands, such as altering GPCR conformations to affect ligand binding. For instance, protein interactions with other protein partners by concentrating specific proteins in lipid rafts. The functional activities of signaling proteins, including influencing the catalytic activity of membrane-associated enzymes or the activation state of receptors52. Moreover, their diverse structures provide a basis for varied protein affinities and activities within signaling pathways. For instance, SH2 domains, essential for phosphotyrosine signaling, demonstrate specificity for lipid and protein interactions. SH2 domains specifically recognize and bind to phosphorylated tyrosine residues within specific peptide sequences on target proteins. This interaction is highly specific, with different SH2 domains recognizing distinct phosphotyrosine- containing motifs. Some SH2 domains can also interact with specific phospholipids, particularly PIPs. These interactions are often electrostatic, involving positively charged residues on the SH2 domain and the negatively charged phosphate groups of the lipids. This dual specificity allows SH2 domains to integrate information from the lipid environment and protein phosphorylation states, fine-tuning signaling responses53. A recent study by Kundu et al. has shown that lipids like ganglioside GM2 can activate receptors such as beta1 integrins. Such activation can initiate a signaling cascade leading to cell migration and tumor progression, demonstrating a novel, direct signaling role for ganglioside GM254. Similarly, in plants, lipids modulate signaling pathways in response to environmental stresses, including temperature extremes, drought, and nutrient depletion, further illustrating the universal signaling importance of lipids across life forms55. 18 In summary, the diverse interplay between lipid composition and cellular signaling highlights the critical role of lipids in cellular processes. These studies not only enhance our understanding of cellular homeostasis but also offer promising avenues for targeted therapeutic strategies56. Figure 1.8. represents the biological functions of lipids, from signal transduction to energy metabolism. Lipids provide a dynamic and flexible barrier from the extracellular including environment. Lipid/membrane mediates membrane exocytosis and endocytosis, influencing the formation and fusion of vesicles. Lipids are involved in signaling by acting as first and second messengers. Lipids are a key input into the citric acid cycle. This cycle is vital to produce ATP, utilizing the energy from fatty acid metabolism to power the cell, made by BioRender trafficking processes, 4. PUFAs in diet over time 4.1. How different diets contain different omega-3/omega-6 fatty acid ratio Epidemiological studies indicate that individuals susceptible to coronary heart disease (CHD) benefit from incorporating omega-3 fatty acids from both plant and marine sources into their diets57. While the optimal dosage remains uncertain, daily intakes of EPA+DHA ranging from 0.5 19 to 1.8 grams, obtained either from fish or supplements, significantly decrease heart disease and overall mortality rates57. These findings align with the 2000 American Heart Association (AHA) Dietary Guidelines recommending the inclusion of a minimum of two servings of fish, especially fatty fish, per week. Regarding ALA, a total intake of 1.5 to 3 grams per day is suggested, yet conclusive evidence from prospective, randomized clinical trials is needed57. The ratio of omega- 3 to omega-6 fatty acids in the diet can vary depending on the types of foods that are consumed. Different types of diet contain different ratios of these two types of fatty acids. For example, the typical Western diet, containing high levels of processed foods, is often made with vegetable oils such as corn, soybean, and sunflower oil, which are high in omega-6 fatty acids. Research has implicated saturated fats and cholesterol as primary contributors to the increasing prevalence of heart disease. Consequently, this has led to recommendations to replace saturated fat and cholesterol with unsaturated fats, which are presumed to be a healthier alternative58. Vegetable oils, known for their reduced saturated fat content, were validated in clinical trials for their ability to lower serum cholesterol levels. However, an unintended consequence of consuming more vegetable oil is the potential elevation of the omega-6 to omega-3 fatty acids ratio. An increase in the omega-6 to omega-3 fatty acid ratio has been linked to a prevalence of chronic inflammation and the onset of various chronic diseases over the past decades. In some countries such as Japan, fish and seafood are a traditional part of the diet, and these foods are rich in omega-3 fatty acids such as EPA and DHA. However, many plant-based foods such as flaxseed, chia seeds, and walnuts are good sources of ALA, which can be converted into EPA and DHA in the body, although this conversion is limited. Ethnic differences in omega6/omega3 percentage in thrombocyte phospholipids in Europe and the United States, Japan, and Greenland Eskimos are 50, 12, and 1. In contrast, the percentage of all deaths from cardiovascular disease in those three 20 ethnicities is 45, 12, and 7% (Table 1.1)59. A traditional Mediterranean diet is rich in whole, unprocessed foods such as fruits, vegetables, whole grains, and legumes, as well as fatty fish, nuts, and olive oil. This diet tends to have a lower ratio of omega-6 to omega-3 fatty acids (7.10)60, as the focus is on consuming whole foods rather than processed foods high in omega-6 fatty acids61. A vegetarian or vegan diet may also have a lower ratio of omega-6 to omega-3 fatty acids, as it typically includes a higher intake of plant-based foods that are rich in ALA, such as flaxseed, chia seeds, and walnuts. The key to achieving a healthy ratio of omega-6 to omega-3 fatty acids in the diet is to focus on consuming whole, unprocessed foods and reduce the consumption of processed foods that are high in omega-6 fatty acids. Including sources of omega-3 fatty acids, such as fatty fish or plant-based sources, can also help to balance the ratio. Table 1.1. Fatty acid concentrations in thrombocyte phospholipids and % of death from cardiovascular disease among different ethnic54 Arachidonic acid (%) Eicosapentaenoic acid (%) Ratio of ω6/ω3 (%) Mortality from cardiovascular disease (%) Japan 26 0.5 50 45 Eskimos 21 1.6 12 12 Europe and United States 8.3 8.0 1 7 4.2. How much has omega-6/omega-3 fatty acid changed over the last centuries, when, and why? Various sources of evidence suggest that the human evolutionary diet featured an omega-6 to omega-3 essential fatty acid (EFA) ratio of around 1, while contemporary Western diets exhibit a ratio ranging from 15/1 to 16.7/1. This imbalance, characterized by excessive omega-6 levels in modern diets, contradicts the dietary patterns under which human genetics were established. High levels of omega-6 PUFA) and an imbalanced omega-6/omega-3 ratio in contemporary Western diets are believed to contribute to the onset of various diseases, including cardiovascular disease, cancer, and inflammatory and autoimmune disorders. In contrast, increased levels of omega-3 21 PUFA, which result in a lower omega-6/omega-3 ratio, have been shown to have protective effects against these diseases. For example, in secondary prevention of cardiovascular disease, an omega- omega-6/ omega-3 ratio of 4:1 was associated with a substantial 70% reduction in total mortality. Similarly, for patients with colorectal cancer, an omega-6/omega-3 ratio of 2.5:1 was found to decrease rectal cell proliferation59. In rheumatoid arthritis patients, an omega-6/omega-3 ratio of 2-3/1 suppressed inflammation, and an omega-6/omega-3 ratio of 5/1 had a positive impact on asthma patients, whereas an omega-6/omega-3 ratio of 10/1 had adverse effects. These studies emphasize the disease-specific nature of the optimal ratio, aligning with the understanding that chronic diseases are influenced by multiple genes and factors. Consequently, the therapeutic dose of omega-3 fatty acids may vary based on the severity of the disease resulting from genetic predisposition. A lower omega-6/omega-3 ratio is deemed more favorable in mitigating the risk of prevalent chronic diseases in Western societies and developing countries, which are now extending to the global population59. The ratio of omega-6 to omega-3 fatty acids in the diet is important because these two types of fatty acids have opposing effects on the body's inflammatory response. Omega-6 fatty acids are generally considered pro-inflammatory, while omega-3 fatty acids have anti-inflammatory properties. While inflammation is a natural response that helps the body fight infection, chronic inflammation can contribute to the development of many diseases. Early human diets, characterized by lean meat, fish, green leafy vegetables, fruits, nuts, berries, and honey, shaped the genetic nutritional requirements of modern humans. Cereal grains, including wheat, maize, and rice, now contribute significantly to the world's grain production, constituting 75% of the total product62,63. This shift in dietary patterns has major nutritional implications, given that cereal grains are rich in carbohydrates and omega-6 fatty acids but deficient in omega-3 fatty acids and 22 antioxidants, especially compared to green leafy vegetables. The nutritional composition of contemporary diets, characterized by high grain consumption, has been linked to increased risks of insulin resistance, hyperinsulinemia, coronary heart disease, hypertension, diabetes, and obesity. Notably, the dietary transition from diverse, calorie-dilute foods consumed by Paleolithic humans to the calorie-concentrated nature of modern diets has occurred over a relatively short period—less than 500 generations64. The impact of this rapid dietary change extends beyond caloric density, influencing the balance between omega-6 and omega-3 fatty acids, essential components for health. The Western diet, with an omega-6 to omega-3 ratio of 15–20/1 compared to the ancestral 1/1 ratio, is considered deficient in omega-3 fatty acids64,65. There is a complex interplay between diet, evolution, and health, emphasizing that the genetic changes in response to dietary influences occurred over millions of years, with omega-3 fatty acids naturally present in the foods consumed during the long evolutionary history of the genus Homo. However, the unprecedented dietary changes of the past 100–150 years, marked by a significant increase in the omega-6/omega-3 ratio, pose novel challenges to human health, challenging our evolutionary precedent66,67. 5. Different metabolic pathways of PUFA Oxylipins are a broad and diverse group of bioactive lipid metabolites derived from the oxidation of PUFAs, including LA (C18:2n-6), AA (C20:4n-6, AA), EPA (C20:5n-3), and DHA (C22:6n- 3). The synthesis of oxylipins starts from the hydrolysis of the plasma membrane PUFA by phospholipases, specifically, phospholipase A (PLA) enzymes are specialized in the hydrolysis of phospholipids at specific positions on the glycerol backbone: PLA1 enzymes target the sn-1 position to release fatty acids, while PLA2 enzymes act at the sn-2 position, also liberating fatty acids. Additionally, patatin-like PLA enzymes exhibit non-specific lipid acyl hydrolase (LAH) 23 activity, granting them the ability to hydrolyze phospholipids at both the sn-1 and sn-2 positions. Subsequently, the released PUFAs are further metabolized by three metabolic pathways such as cyclooxygenase (COX), lipoxygenase (LOX), or cytochrome P450 (CYP) pathways. Alternatively, there's the possibility of their nonenzymatic conversion into oxylipins through reacting with reactive oxygen species (ROS)68,69. Figure 9 illustrates enzymatic and non- enzymatic pathways involved in the metabolism of PUFA. 1) Cyclooxygenase (COX) pathway: This pathway involves the conversion of AA to prostaglandins, which are involved in inflammation and other physiological processes. COX enzymes convert AA to prostaglandin H2 (PGH2), which is then converted to various prostaglandins such as PGE2, PGD2, and PGF2α by specific prostaglandin synthases, COX enzyme, is necessary for the formation of thromboxane A2. 24 Figure 1.9. Schematic representation of oxylipin metabolic pathways 2) Lipoxygenase (LOX) pathway: This pathway involves the conversion of PUFAs to leukotrienes, which are involved in inflammation, immunity, and allergic reactions. LOX enzymes convert PUFAs to hydroperoxy eicosatetraenoic acids (HPETEs), which are then converted to leukotrienes 25 such as LTB4 and LTC4 by specific leukotriene synthases. The LOX family of enzymes introduces molecular oxygen into arachidonic acid at specific carbon positions, leading to the formation of various hydroperoxyETEs and then forming HETE isomers depending on the LOX enzyme involved. Common LOX enzymes include 5-LOX, 12-LOX, and 15-LOX, which generate 5- HETE, 12-HETE, and 15-HETE, respectively. 3) Cytochrome P450 (CYP) pathway: This pathway involves the conversion of PUFAs to epoxyeicosatrienoic acids (EpETrEs) and hydroxyeicosatetraenoic acids (HETEs), which are involved in vascular function and inflammation. CYP enzymes convert PUFAs to EpETrEs and HETEs, which have various biological activities, a more detailed explanation of this enzyme is provided in the following section. 6. Bioactive compounds and biomarkers 6.1. Definition of biomarkers and the historical aspect of how these chemicals were discovered “A characteristic that is objectively measured and quantified as an indicator of normal biological processes, pathogenic processes or pharmacological response to a therapeutic intervention”. Biological markers, or biomarkers, are measurable indicators of biological processes or conditions in the body. They can be used to diagnose diseases, monitor disease progression, and evaluate the effectiveness of treatments. Biomarkers can be proteins, DNA, RNA, metabolites, or other molecules that can be detected in bodily fluids or tissues. They serve as indicators for normal and pathogenic processes, as well as responses to therapeutic interventions. Biomarkers are crucial tools in understanding disease prediction, causation, diagnosis, progression, regression, or treatment outcomes 70. Biomarkers have been extensively used in various fields, such as infections, immunological and genetic disorders, and cancer research. The use of biomarkers in research has 26 increased due to the need for direct measurements in disease causation, unaffected by recall bias, and capable of providing insights into the absorption and metabolism of substances. For example, neuroscientists use biomarkers to diagnose and treat nervous system disorders and explore their mechanisms. These biomarkers are collected from various biological sources such as blood, brain, cerebrospinal fluid, skin, and urine to study both healthy and diseased states of the nervous system. Advances in molecular biology and laboratory technologies have facilitated the measurement of complex biomarkers. Molecular biomarkers provide a robust tool for clinical researchers to understand and address a wide range of neurological diseases, with applications spanning analytic epidemiology, clinical trials, disease prevention, diagnosis, and management70. Biomarkers play a multifaceted role in medicine, offering diverse applications that significantly impact disease screening, characterization, and treatment. These versatile indicators can be effectively employed to: 1) Screen for diseases: biomarkers play a vital role in early disease detection, serving as essential tools for efficient screening programs. Their utilization enables the identification of potential health concerns at their incipient stages, allowing for timely intervention. This proactive approach enhances the effectiveness of healthcare strategies and improves patient outcomes. 2) Characterize diseases: the characterization of diseases, such as the identification of trinucleotide repeats, is facilitated by biomarkers. This enables a better understanding of the underlying nature of various medical conditions71. 3) Rule out, diagnose, stage, and monitor diseases: biomarkers contribute to the entire spectrum of disease management, from ruling out potential conditions to accurate diagnosis, staging, and ongoing monitoring of disease progression. 4) Inform prognosis: they play a crucial role in predicting the likely course and outcome of diseases, providing valuable insights into the potential trajectory of a patient's health. 5) Individualize therapeutic interventions: 27 biomarkers enable a personalized approach to therapeutic interventions by monitoring responses to treatments. This individualized strategy ensures that medical interventions are tailored to the unique characteristics of each patient. 6) Predict outcomes in response to therapies: biomarkers offer predictive capabilities, allowing healthcare professionals to anticipate the likely outcomes of specific therapeutic interventions. This foresight aids in making informed decisions about the course of treatment. 7) Predict adverse drug reactions: biomarkers play a vital role in predicting and mitigating adverse reactions to drugs, contributing to safer and more effective medication regimens. 8) Predict and guide treatment of drug toxicity: in instances of drug toxicity, biomarkers, such as the measurement of serum concentrations following medication overdose, provide critical information to guide the treatment process. 9) Identify cell types: biomarkers extend beyond systemic considerations to include cellular-level insights. They are instrumental in identifying specific cell types, aiding in histological analyses, and contributing to a deeper understanding of tissue characteristics. In essence, biomarkers serve as indispensable tools across the entire spectrum of healthcare, from early detection and accurate diagnosis to personalized treatment strategies and ongoing monitoring, enhancing the quality of patient care72. The concept of biomarkers dates to ancient times when physicians used certain physical signs or symptoms to diagnose illness or disease. For example, the presence of fever was considered a biomarker of infection. In the modern era, the discovery of biomarkers was driven by advances in biochemistry and molecular biology. One of the earliest examples of biomarker discovery was the identification of the enzyme creatine kinase, which is released into the blood following muscle damage. This discovery was made in the 1950s and paved the way for the use of creatine kinase as a biomarker of heart attack and other conditions that cause muscle damage73. Another significant development in biomarker discovery was the identification of prostate-specific antigen (PSA) in 28 the 1980s. PSA is a protein produced by the prostate gland and is elevated in the blood of men with prostate cancer. The discovery of PSA revolutionized the diagnosis and management of prostate cancer, as it provided a simple blood test that could detect the disease at an early stage74. More recently, advances in genomics and proteomics have led to the discovery of numerous biomarkers associated with various diseases and conditions. For example, genetic mutations and changes in gene expression patterns can serve as cancer biomarkers, while the presence of specific proteins in the blood or urine can indicate the presence of kidney disease, liver disease, or other conditions74. 6.2. Oxylipins are useful biomarkers of diseases Oxylipins are oxygenated bioactive lipid metabolites derived from PUFAs and play important roles in regulating inflammation, immune responses, and other physiological processes in the body. Because of their diverse functions and their levels change with different disease conditions, oxylipins could be useful biomarkers of diseases. Some reasons why oxylipins are useful biomarkers of diseases are: 1) Specificity: oxylipins are highly specific when it comes to function and can be used to identify the type and severity of the disease. For example, certain oxylipins are associated with specific types of cancer, such as breast cancer75 or colon cancer76, while others are associated with cardiovascular disease77, diabetes78, or neurological disorders (explained in detail in the following section). 2) Sensitivity: The biochemical sensitivity of oxylipins as biomarkers refers to their ability to detect and respond quickly to early biochemical changes within a biological system. For example, during the early phase of the inflammatory process, specific enzymes like COX-2 are induced and increase the production of proinflammatory oxylipins, such as prostaglandins, from PUFAs. These proinflammatory oxylipins act locally, recruit immune cells, increase vascular permeability, and mediate other inflammatory responses. Because this is the 29 early step of inflammation, these proinflammatory oxylipins can be detected before others. The sensitivity of oxylipins is such that they provide an early warning signal of specific pathological conditions. Their levels may increase significantly in response to even slight changes in the cellular state, providing a biomarker that is responsive to the early stages of inflammation. 3) Non- invasive: Oxylipins can be measured in various biological fluids such as blood, urine, and saliva, which can be obtained in a relatively less invasive manner. This makes them a convenient and safe alternative to traditional diagnostic tests that may require invasive procedures like biopsy for cancer79. 4) Predictive value: The levels of certain oxylipins can be used to predict disease progression and response to treatment; for example, in the context of breast cancer research, oxylipins have been investigated as potential biomarkers for predicting disease progression and treatment response. A recent study on breast cancer patients with adjuvant anastrozole therapy, a common treatment for hormone receptor-positive breast cancer, showed significant alterations in specific oxylipins, such as 12-HETE and 13-HODE, in response to anastrozole treatment80. To further solidify the case for oxylipins as valuable biomarkers in disease, it's important to consider the practical implications of their use in a clinical setting. The integration of oxylipin profiling into clinical practice could revolutionize patient care by enabling more precise and individualized therapeutic strategies. For instance, monitoring oxylipin levels over time gives clinicians a dynamic view of a patient's response to treatment, allowing for real-time adjustments to optimize therapeutic outcomes. Moreover, the potential for oxylipins to act as early disease biomarkers opens the possibility for preventative interventions. By identifying at-risk individuals through changes in their oxylipin profiles before the onset of overt disease, clinicians could intervene earlier, potentially halting or reversing disease progression. 30 The increasing prevalence of personalized medicine also underscores the value of oxylipins. Given their role in reflecting an individual's unique biochemical milieu, oxylipin profiles could tailor treatments according to a patient’s specific biological context, thereby enhancing the efficacy of therapeutic interventions and reducing the likelihood of adverse effects. Oxylipins are a promising group of biomarkers that can provide valuable insights into the underlying mechanisms of various diseases. Their specificity, sensitivity, non-invasiveness, and predictive value make them a valuable tool for disease diagnosis and prognosis, and they hold the key to unlocking a new horizon in personalized treatment79. 6.3. Effect of PUFA Metabolite on different diseases 6.3.1. Cardiovascular diseases Oxylipins such as prostaglandins, leukotrienes, and thromboxanes can modulate platelet aggregation, vascular tone, inflammation, and the pathogenesis of cardiovascular diseases such as atherosclerosis, myocardial infarction, and stroke. The initiation of atheromatous lesions is well- established, primarily involving the oxidation of low-density lipoproteins within the vessel wall's intima. One study focused on plasma samples from three cohorts: 61 acute coronary syndrome (ACS) patients (group A), 49 acute ischemic stroke (AIS) patients (group D), and 82 controls (group K). Utilizing untargeted lipidomics, the research aimed to identify potential lipid pattern relationships across these patient categories. Hydroxy fatty acid fatty acyls (FAHFAs) emerged as the most distinctive lipids in group K compared to groups A and D; higher levels in group K suggest an increased anti-inflammatory activity, which is lower in group A and even lower in group D 81. In the heart, EpETrEs provide cardioprotective effects during ischemia/reperfusion injury29. EpETrEs are categorized as endothelium-derived hyperpolarizing factors (EDHFs); they are produced in endothelial cells and cause hyperpolarization in vascular smooth muscle cells by 31 activating large-conductance calcium-activated potassium (BK(Ca)) channels77. One study aimed to compare the dilating potency of docosahexaenoic acid and its five CYP epoxy metabolites, epoxydocosapentaenoates (EDPs), in porcine coronary arterioles precontracted with endothelin. The five EDP regioisomers induce vessel dilation with EC50 values ranging from 0.5 to 24 pM, while the EDP hydrolyzed product 13,14-dihydroxy docosapentaenoic acid (13,14-DHDP) has an EC50 value of 30 +/- 22 nM 82. Moreover, the hydroxyeicosatetraenoic acids (HETEs), derivatives of AA, are produced via LOX metabolism and have been implicated in various physiological and pathological processes related to cardiovascular health. These LOX metabolites have played roles in chemotaxis, the modulation of vascular tone, and the stimulation of vascular endothelial growth factor production. Mid-chain HETEs such as 5-, 8-, 9-, 11-, and 12-HETE are critical in the migration and activation of leukocytes. In contrast, 15-HETE tends to exert opposing effects and is a precursor to lipoxins (LXs), which are important for inflammation resolution. Additionally, there is evidence that the production of mid-chain HETEs is heightened in cases of essential hypertension, suggesting a potential role in its development. These findings highlight the distinct and sometimes contrasting effects of EpETrEs and HETEs within vascular biology77. In conclusion, the body of research surrounding oxylipins has elucidated their multifaceted roles in cardiovascular health and disease. As bioactive lipid metabolites, oxylipins such as prostaglandins, leukotrienes, and thromboxanes are pivotal in modulating physiological responses like platelet aggregation, vascular tone, and inflammation, which are vital components in the pathogenesis of cardiovascular conditions, including atherosclerosis, myocardial infarction, and stroke. The insights gained from lipidomic studies, such as the differential levels of anti- inflammatory FAHFAs in various patient cohorts, have highlighted the potential of oxylipins as biomarkers for disease states and therapeutic response. 32 6.3.2. Neurodegenerative diseases Recent studies have shed light on the potential therapeutic effects of omega-3 fatty acids in cognitive disorders such as Alzheimer's disease (AD). The impact of a 6-month administration of DHA-rich omega-3 FA supplementation on plasma FA profiles in patients with mild to moderate AD is investigated in the OmegAD study, which involved 174 AD patients randomized to daily intake of 2.3 g omega-3 FA or placebo for the first 6 months, followed by all patients receiving the omega-3 FA preparation for the next 6 months. The analysis of baseline and changes in plasma levels of major omega-3 FAs in 165 patients during omega-3 FA supplementation revealed a significant association between the preservation of cognitive functioning (assessed by ADAS-cog scores) and increasing plasma omega-3 FA levels over time, this relationship held regardless of gender. The findings from this study suggest a dose-response relationship between plasma omega- 3 FA levels and preserving cognition, emphasizing the potential benefits of graded doses in future omega-3 FA trials for individuals with mild AD83. In another study, researchers utilized transgenic Fat-1 mice that carry an omega-3 FA desaturase. This genetic modification enables these mice to synthesize omega-3 PUFA from omega-6 PUFA endogenously. The expression of the fat-1 gene in these mice led to a notable increase in the omega-3/omega-6 PUFA ratio in the brain, with a significant elevation of DHA levels observed (+11%, p < 0.001), including a specific increase in DHA by 5% (p < 0.001) in 20-month-old mice. Interestingly, the expression of fat-1 also resulted in a reduction of soluble Aβ₄₂ levels (-41%, p < 0.01) in the brains of 20-month-old mice. However, it did not affect the levels of insoluble forms of Aβ₄₀ and Aβ₄₂ in the 3 × Tg-AD mouse model. These findings suggest a potential link between endogenous omega-3/omega-6 FA ratio and Alzheimer's disease pathology84,85. Omega-3 fatty acids can be metabolized into a range of oxylipins that have anti-inflammatory, pro-resolving, and neuroprotective properties. The 33 beneficial effects observed from the above-mentioned studies could be triggered by their downstream oxylipins. Utilizing preserved samples from the Emory Goizueta Aging and Alzheimer’s Disease Research Center, a randomized, double-blind, placebo-controlled trial conducted in Sweden, undertook a comparative analysis of oxylipin levels in individuals with AD and cognitively healthy subjects, using cerebrospinal fluid (CSF) from 150 AD patients and 139 controls, as well as plasma from 148 AD patients and 133 controls. The CSF of AD patients was characterized by increased levels of CYP enzyme-derived EpOMEs, specifically 9,10- and 12,13- EpOME from LA, along with their metabolites, 9,10- and 12,13-DiHOMEs. Additionally, a decrease in the CYP-sEH enzyme-derived 17,18-DiHETE from EPA was observed. In the plasma, a rise in EPA-derived 17,18-DiHETE was noted in the AD group, with a simultaneous decrease in the EPA metabolites 5-, 12-HEPE, and DHA-derived 4-, 14-HDoHE86. The other study involved 174 patients with mild to moderate AD; participants received either an omega-3 supplement (1.7g DHA and 0.6g EPA) or a placebo daily for six months. The primary outcome was cognitive function measured by the Mini-Mental State Examination (MMSE) score. There was no significant difference in cognitive function (MMSE scores) between the omega-3 and placebo groups after six months of supplementation. They concluded that omega-3 fatty acid supplementation did not delay cognitive decline in this study population of mild to moderate AD patients over six months. Some limitations noted were the relatively short 6-month duration and the fact that the study may have been underpowered to detect small cognitive changes. So, this trial represents one of the examples in which omega-3 supplementation did not show significant benefits in cognitive outcomes in patients with mild to moderate Alzheimer's disease, a neurodegenerative condition, over the 6-month study period87. 34 While some studies suggest the potential benefits of omega-3 PUFAs for cognitive function and reducing the risk of AD, clinical trial results have been inconsistent. The other six months-long trial found no significant cognitive improvements with omega-3 supplementation. This highlights the need for more research to determine the efficacy of omega-3 PUFAs in neurodegenerative diseases like Alzheimer's, considering factors such as dosage, formulation, timing, and potential genetic influences87,86. 6.3.3. Cancer PUFA metabolites can have both pro- and anti-cancer effects depending on the type of metabolite and the stage of cancer. For example, prostaglandins and leukotrienes promote cancer cell growth and survival, while others, such as resolvins and protectins, inhibit cancer cell proliferation and induce apoptosis88. Numerous studies have established that consuming omega-3 PUFAs reduces the susceptibility to colorectal cancer; however, the underlying mechanisms remain unclear. One study applied targeted metabolomics to investigate how omega-3 PUFAs’ supplementation has anti-colorectal cancer effects through eicosanoid signaling. Three well-defined isocaloric diets are used, containing 10 wt% total fat to investigate the effects of omega-3 PUFAs on colorectal cancer. In the control diet the ratio of omega-6-to-omega-3 PUFA is ≈ 69.3:1, the DHA diet the ratio of omega-6-to-omega-3 PUFA is ≈ 1.26:1, and in the DHA-high diet, the ratio of omega-6-to-omega- 3 PUFA is ≈ 0.56:1. Their findings demonstrate that a diet rich in omega-3 PUFAs suppressed the colorectal tumors growth of MC38, a mouse colon adenocarcinoma cell line, and induced changes in the fatty acid and eicosanoid metabolite profiles in C57BL/6 mice. Notably, the dietary intake of omega-3 PUFAs led to a significant increase in the levels of Epoxydocosapentaenoic acids (EDPs), which are metabolites of omega-3 PUFAs generated by CYP P450 enzymes, both in the plasma and tumor tissue of mice under treatment. Furthermore, systemic administration of EDPs 35 (at a dose of 0.5 mg/kg/day) inhibits MC38 tumor growth in mice. The pro-oncogenic gene expression, including c-myc, Axin2, and C-jun, in tumor tissues was reduced. These collective results support the notion that EDPs may have contributed to the anti-colorectal cancer effects associated with omega-3 PUFAs89. In another study utilizing LC-MS/MS-based targeted lipidomics, investigation reveals that the CYP monooxygenase pathway is a primary route for the in vivo metabolism of LA. Importantly, CYP monooxygenase, the CYP2C subfamily, is responsible for the colon cancer-promoting effects of LA, as demonstrated by the failure of an LA- rich diet to exacerbate colon cancer in mice with genetic KO of CYP monooxygenase. Moreover, this study also revealed that the pro-cancer effects of LA are mediated by CYP monooxygenase by converting LA to epoxy octadecenoic acids (EpOMEs), which significantly promote colon tumorigenesis via gut microbiota-dependent mechanisms. This study underscores the critical role of CYP monooxygenase in the health implications of LA, establishing a unique mechanistic connection between dietary intake of fatty acids and cancer risk. These findings can inform the development of more effective dietary guidelines for optimal LA consumption and identify subpopulations particularly susceptible to the adverse effects of LA90. Table 2 shows a summary of the effect of PUFA metabolites on different types of cancer. Table 1.2. Some examples of the effect of PUFA metabolites on different cancer types. They are categorized based on the enzyme88 Enzyme PUFA Metabolite Effect COX LOX CYP AA EPA AA DHA EPA AA EPA DHA PGE2 PGE3 Induces primary tumor growth and metastasis. Inhibits cancer cells proliferation and invasion in lung cancer cell 5-HETE 4-HDHA 15-HEPE Induces angiogenesis and tumor progression Anti-angiogenesis effect Inhibits the formation of 5-HETE and PGE2 as well as cell proliferation EpETrE 17,18-EpETE 19,20-EpDPE Induces colon cancer progression. Inhibits endothelial cell proliferation and VGFE-induced angiogenesis. Inhibits angiogenesis and suppresses 70% of tumor metastasis. 36 6.3.4. Inflammatory diseases Inflammation is a natural response to infection and injury in the host organism. However, when inflammation becomes excessive or inappropriate, it contributes to various acute and chronic human diseases. This inflammatory process involves the production of inflammatory cytokines and eicosanoids derived from AA, including prostaglandins, thromboxanes, and leukotrienes. Long-chain omega-3 PUFA supplementation decreases the production of inflammatory eicosanoids, cytokines, reactive oxygen species, and the expression of adhesion molecules. The anti-inflammatory effects of long-chain omega−3 PUFAs could occur through multiple mechanisms, such as replacing AA to minimize the production of proinflammatory AA-derived eicosanoids and changing the inflammatory gene expression via the modulation of transcription factor activation. Long-chain omega-3 PUFAs are metabolized to generate their corresponding anti-inflammatory/pro-resolving mediators known as resolvins. Given their potent anti- inflammatory properties, these fatty acids may have therapeutic applications in various acute and chronic inflammatory conditions. Clinical evidence supports their efficacy in specific contexts, such as rheumatoid arthritis, but remains limited in others, including inflammatory bowel diseases and asthma91. Furthermore, larger, well-designed trials are necessary to thoroughly evaluate the therapeutic utility of long-chain omega-3 PUFAs in inflammatory diseases. Animal experiments have demonstrated a robust correlation between AA levels in inflammatory cells and the cells' capacity to generate downstream eicosanoids, including PGE2. Consequently, elevating the dietary intake of AA in rats or augmenting its presence in the human diet can lead to increased AA within inflammatory cells. Dietary supplementation with fish oil decreases T lymphocyte proliferation in healthy older humans. T lymphocytes, pivotal components of the immune system, play a central role in the 37 development of autoimmune diseases. In a healthy immune response, T cells help protect the body against pathogens; however, when immune tolerance is compromised, the same cells may erroneously attack self-tissues. This loss of self-tolerance, often due to dysfunctional regulatory T cells (Tregs), can lead to the activation and proliferation of autoreactive T cells, which recognize and respond to the body’s proteins as if they were foreign invaders. The resultant misguided immune response manifests as chronic inflammation and tissue damage, characteristic of autoimmune conditions92,93. Moreover, studies involving supplementation with (3.1–8.4 g) of EPA and DHA per day reduce reactive oxygen species (specifically superoxide or hydrogen peroxide) produced by stimulated human neutrophils by 30% to 55%. In cell culture experiments, it has been demonstrated that EPA and DHA can impede the synthesis of IL-1β and TNF-α by monocytes, as well as the synthesis of IL-6 and IL-8 by venous endothelial cells 94. Altogether, PUFA metabolites have diverse effects on different diseases, and their roles are complex and multifactorial. Further research is needed to fully understand the mechanisms of action and clinical significance of PUFA metabolites in various diseases. 7. Oxylipin receptors As discussed in previous sections, lipid metabolites activate diverse nuclear receptors with a wide range of biological effects. A nuclear receptor with extensive roles in governing gene expression is the peroxisome proliferator-activated receptor (PPAR). These receptors control various biological pathways, developmental processes, conversion to neoplastic states, inflammation, and wound healing. Moreover, PPARs are pivotal in regulating energy metabolism, including lipid and carbohydrate metabolism. Essentially, they function as critical regulators of the metabolic processes linked to the oxidation and storage of dietary lipids. This is facilitated by their capacity 38 to act as sensors for fatty acids and their intermediary metabolites95. Table 1.3 summarizes the effect of the different PPAR subtypes on fatty acid metabolic regulation96,97. Table 1.3. The activity of PPAR subtypes, their gene targets, and metabolic regulation effect, with a focus on fatty acids. Abbreviation: LTB4: Leukotriene B4, HETE: hydroxyeicosatetraenoic acid, PGJ: prostaglandin J97 PPAR Subtype PPARα Metabolic Regulation It functions as a lipid sensor and regulates energy combustion. Additionally, it is important in the development of fatty liver disease It is involved in fatty acid oxidation, improves lipid profiles and reduces adiposity Ligand PUFA, LTB4, 8- HETE PUFA PUFA, 15-HETE, PGJ2 It is regulating the genes participating in the release, transport, and storage of fatty acids PPARβ/δ PPARγ GPCRs are the other essential receptors for PUFAs and their metabolites. Oxylipins influence adipocyte functions such as lipogenesis, lipolysis, and secretion and phenotypic conversions like browning or whitening processes. The modulation of adipocyte functions primarily occurs through the activation of membrane GPCRs by oxylipins, given the sensitivity of adipocyte functions to cAMP- and Ca++-mediated pathways. Gs-coupled receptors increase cAMP levels; Gi-coupled receptors decrease cAMP levels; and Gq-coupled receptors increase intracellular Ca++ levels. Different GPCRs and their oxylipin ligand and functions are mentioned in Table 4. This table outlines the key GPCRs that interact with specific oxylipins and the functions they mediate. The identified GPCRs, such as Free Fatty Acid Receptor 1 (FFA1), Free Fatty Acid Receptor 4 (FFA4), Leukotriene B4 Receptor 1 (BLT1), Cysteinyl Leukotriene Receptor 1 (CysLT1), Prostaglandin E (EP) receptor subtypes 1 through 4 EP1-4, DP2 are the two receptor subtypes for PGD2, play crucial roles in regulating processes like inflammation, metabolism, vascular function, and platelet aggregation in response to different oxylipin signaling molecules. These GPCRs are critical players in mediating the diverse effects of oxylipins on cellular signaling and physiological responses98,99. 39 Table 1.4. GPCRs are activated by oxylipins and their functions. Abbreviations: FFA1 (GPR40): Free Fatty Acid Receptor 1, also known as G Protein-Coupled Receptor 40, FFA4 (GPR120): Free Fatty Acid Receptor 4, also known as G Protein-Coupled Receptor 120, BLT1: Leukotriene B4 Receptor 1, CysLT1: Cysteinyl Leukotriene Receptor 1, EP1-4: Prostaglandin E2 Receptors 1-4, DP1-2: Prostaglandin D2 Receptors 1-2 GPCR Oxylipin Ligands Functions FFA1 (GPR40) 15-HETE, 12-HETE, 5-HETE FFA4 (GPR120) Resolvin E1, Protectin D1, Maresin 1 BLT1 CysLT1 EP1-4 DP1-2 LTB4 Cysteinyl leukotrienes (LTC4, LTD4, LTE4) Prostaglandin E2 Prostaglandin D2 Regulates insulin secretion, glucose and lipid metabolism Mediates anti-inflammatory and insulin- sensitizing effects Involved in leukocyte chemotaxis and activation Regulates bronchoconstriction and vascular permeability Diverse functions including inflammation, vasodilation, and nociception Involved in inflammation, vasodilation, and platelet aggregation Early findings identified the DGLA-derived PGE1 as an anti-lipolytic agent across various species. Similarly, PGE2 demonstrated anti-lipolytic properties, likely through the EP3 receptor, known to inhibit cAMP-mediated pathways induced by lipolytic agents. Notably, the activation of Gαi- coupled receptors by PGE1 and PGE2, particularly through EP3 receptors, causes a decrease in cAMP levels, which is a key second messenger that promotes lipolysis, so reducing its levels inhibits fat breakdown. While Gαq-coupled receptor-activating PGF2α lacks this lipolysis inhibitory effect. The control of lipolysis by oxylipins is underscored by in vivo evidence showing that inhibiting COX activity with indomethacin enhances lipolysis under fasting conditions. Oxylipins play a pivotal role in interacting between inflammation and lipolysis in adipocytes. They control and are controlled by inflammatory cytokines, with omega-6-PUFA-derived metabolites generally proinflammatory and omega-3-PUFA-derived metabolites exhibiting pro-resolving or anti-inflammatory properties through their function as PPARγ ligands98. 40 8. CYP 450-EH pathway 8.1. Introduction to cytochrome P450 The first CYP enzyme was discovered by Klingenberg and Garfinkel as an unknown pigment that in its reduced form binds to carbon monoxide, producing an absorption peak at wavelength 450 nm100,101. CYPs are involved in many reactions, such as heteroatoms oxidation and dealkylation, couplings of an aryl ring, formation of the ring, rearrangement of oxygenated molecules, and carbon-carbon bond cleavage and desaturation. The monooxygenase activity of CYPs, including the epoxidation and hydroxylation, has been discussed extensively elsewhere102,103. CYP proteins, heme-thiolate proteins, play a primary role in the synthesis and metabolism of endogenous biological molecules, including steroid hormones, cholesterol, fatty acids, and various xenobiotics like different drugs, primarily through oxidation processes104. CYPs share a distinctive residue sequence of FXXGXbXXCXG, where Xb is a basic residue, and cysteine is axially positioned to the heme. Additionally, they exhibit a peak at 450 nm upon carbon monoxide binding to the Fe (II) of the heme105,106. The Human Genome Project has identified 57 genes that code for various CYP enzymes, categorized into 18 families and 43 sub-families based on amino acid sequence similarities. Each CYP enzyme is labeled with a number indicating the family, a letter for the subfamily, and a second number unique to the specific enzyme (e.g., CYP2J2). In prokaryotes, CYP enzymes are soluble, whereas in mammals, CYPs are mostly membrane-bound proteins located in the endoplasmic reticulum (ER) or mitochondria. The catalytic domains of membrane- associated CYPs are immersed in the membrane. The active site is linked to both the cytosolic environment and the membrane through access channels, allowing substrates from either compartment to interact with the enzymes107,108. 50 CYP enzymes are primarily involved in the metabolism of xenobiotics; they are in the ER. The remaining CYP enzymes are situated in the 41 mitochondrial membrane and are involved in the metabolism and biosynthesis of endogenous molecules109,110. Although primarily expressed in the liver, these enzymes are found in various other tissues, including the skin, brain, lung, intestinal mucosa, and kidney. CYPs metabolize PUFAs to produce epoxy PUFAs (Ep-PUFAs) or hydroxy-PUFAs (hydroxylase activity). The specific CYP enzyme and the type of PUFA substrate define the final product. Moreover, the structure, regio/stereoselectivity, and catalytic properties of each CYP enzyme are influenced by the enzyme’s active site’s structure106,111,112. In the context of AA metabolism, Ep-PUFAs are primarily produced by CYP2B, 2C, and 2J sub-families, while omega and omega-1 hydroxylated AA are predominantly generated by 1A, 4A, and 4F subfamilies113. To illustrate, CYP2C and 2J subfamilies can convert AA into four regioisomers of EpETrEs, namely 5,6-EpETrE, 8,9-EpETrE, 11,12-EpETrE, and 14,15-EpETrE, depending on the specific double bond involved in oxygen Figure 1.10. Crystal structures of P450 105AS1, image from the RCSB PDB (RCSB.org)105 insertion. Each EpETrE product may exist as the R, S- or the S, R-stereoisomer114,115. Studies have proposed that CYP isoforms can be efficiently involved in LA metabolism. For example, CYP2C9 42 converts LA to both regioisomers 9,10- and 12,13-EpOMEs in the human liver. Regarding EPA and DHA, research demonstrates that CYP isoforms in humans, rats, and mice can metabolize these compounds116. Specifically, human isoforms such as CYP2C8, 9, 18, and 19, along with CYP2J2, can epoxidize both EPA and DHA117,118. Notably, the catalytic activities of CYP2C isoforms for EPA and DHA closely resemble those for AA. In contrast, CYP2J2 exhibits significantly higher rates, displaying 9- and 2-times greater efficiency in metabolizing EPA and DHA, respectively, than AA117,119. The regioselectivity of cytochrome P450 enzymes in metabolizing PUFA is a key factor in producing specific bioactive lipid epoxides. For instance, CYP1A1 and CYP2J2 are particularly adept at generating the 19,20-EpEDP regioisomer, while CYP2D6 stands out for its ability to produce not only 19,20-EpEDP but also a significant amount of 17,18-EpETE. Additionally, CYP2E1 exhibits a distinct preference for catalyzing the formation of the 14,15-EpETrE isomer. These patterns of regioselectivity highlight the specialized roles that different CYP enzymes play in lipid metabolism and their potential impact on physiological processes106. 8.2. Introduction to epoxide hydrolase The primary role of EH is the breakdown of xenobiotic epoxides and Ep-PUFAs. There are four different EH isoforms. The first identified mammalian EH was the microsomal epoxide hydrolase (mEH) encoded by EPHX1. This enzyme, which is bound to membranes, is anchored to the surface of the ER or plasma membrane via its N-terminal region120. Another type of mammalian EH identified is the soluble epoxide hydrolase (sEH), encoded by the EPHX2 gene. This bi-functional homodimeric enzyme is in both the cytosol and peroxisomes. The C-terminal region is responsible for its epoxide hydrolase activity, while the N-terminal region exhibits phosphatase activity. sEH enzymes are extensively distributed across various tissues in the human body, with their expression 43 and activity levels being influenced by factors such as sex, tissue type, and age121. The recently identified mammalian epoxide hydrolases EH3 and EH4 are more highly expressed in the brain than in other tissues122. All EHs can convert Ep-PUFAs into corresponding 1,2-diols. At the same time, EH3 and EH4 are known to possess an aspartate nucleophile crucial for their hydrolytic activity, which distinguishes them from other α/β-hydrolase domain-containing (ABHD) proteins. This distinction is important as not all EHs share this specific structural feature, highlighting the diversity in catalytic mechanisms among different EHs123. The enzymatic activity of sEH can be elucidated through three distinct steps. It was proposed that initially, the epoxide enters a hydrophobic tunnel with an L-shaped structure, with the nucleophilic aspartate at the active site. Epoxide substrate is stabilized through hydrogen bonds between two tyrosine and the epoxide group of the substrate. Following this, the aspartate residue acts as an activated nucleophile, attacking the electrophilic carbon of the epoxide, resulting in the formation of an acylated intermediate. The final step involves hydrolysis of an ester bond of the bound acylated substrate catalyzed by a basic histidine, forming the 1,2-diol. This process ultimately reinstates the conserved aspartate present in all sEHs123. These four EH enzymes, with their distinct subcellular localizations, substrate preferences, and tissue-specific expression patterns, play important and complementary roles in the metabolism of epoxide-containing compounds and regulate various physiological processes in mammals. 9. Lipid profiling and the “Omics” era 9.1. History of lipid profiling The history of lipid profiling can be traced back to the mid-20th century, with a significant turning point marked by the advancement of chemical analysis of lipids. In the earlier perception, lipids were considered relatively unimportant compared to biomolecules like proteins and nucleic acids. 44 The development of analytical techniques such as chromatography and mass spectrometry revolutionized this dogma, enabling the identification and quantification of specific lipid molecules in biological samples. By the 1970s and 1980s, lipid profiling emerged as an independent field of study by investigating the roles of lipids in physiological processes, including energy metabolism, membrane structure and function, and cell signaling. Concurrently, researchers delved into the connections between lipids and diseases like heart disease, diabetes, and cancer. The "omics" era, which began between the 1990s and 2000s, represented a significant leap forward in lipid profiling124. This era, encompassing genomics, proteomics, metabolomics, and lipidomics, facilitated the comprehensive study of biological molecules on a larger scale. A pivotal development during this era was the introduction of mass spectrometry-based lipidomics, enabling the identification and quantification of hundreds or thousands of lipid species in a single analysis. This approach revolutionized the study of lipid profiles in various biological samples, including blood, urine, and tissues, providing crucial insights into the roles of lipids in health and disease. The term "omics" saw rapid expansion following the completion of the human genome in the early 2000s. Genomics, the initial "omic," paved the way for subsequent disciplines such as transcriptomics, proteomics, metabolomics, and lipidomics124,125. The latter, a subset of metabolomics, focuses explicitly on mapping and quantifying all lipids within cells, tissues, or organisms, providing insights into lipid alterations associated with health and disease126,127. Lipidomics, which emerged in early 2000, heavily relies on advancements in analytical technologies, notably gas chromatography (GC) and liquid chromatography (LC)128. Mass spectrometry with various ionization technologies, like electrospray ionization (ESI), matrix- assisted laser desorption/ionization (MALDI), and atmospheric pressure chemical ionization (APCI) have vastly improved sensitivity128,129. The field of lipid research advances in parallel with 45 the progress of analytical techniques, allowing for the determination of lipid profile alterations associated with diseases. These advances enable the detection of changes in lipid metabolism or pathway modulation, even in complex biological systems, providing new insights into discovering and characterizing disease molecular biomarkers. As analytical techniques and computational tools continue to evolve, we can expect even more significant progress in lipid research in the foreseeable future129. 9.2. Challenges in the identification and quantification of lipids Our understanding of lipids is less extensive than that of other biomolecules because of two specific challenges: 1) the intricate nature of lipids and their collective functions, and 2) an absence of analytical techniques comparable to those employed in protein analysis, hindering the visualization and manipulation of lipid levels both globally and locally. In this section, we are focusing on the second reason for the identification and quantification of lipids by analytical techniques; more specific challenges in this regard are: 1) A diverse range of lipids: Lipids constitute a diverse group of molecules, as discussed in section (2-1), including triglyceride, phospholipids, sterols, and glycolipids. Each lipid class has a distinct chemical structure and physicochemical properties that can influence their separation and detection, including different chain lengths, unsaturation levels, and isomers. This diversity can make distinguishing between different lipid species in a single analytical analysis challenging. 2) Complex sample matrix: Lipids are often found in complex biological matrices, such as plasma, tissues, and cells. These samples can contain many other molecules, such as proteins, carbohydrates, and nucleic acids, that can interfere with the analysis of lipids. 3) Sample preparation: Lipids are often tightly bound to cellular structures or proteins, making them difficult 46 to extract. Several extraction methods have been developed, but each has advantages and limitations6. 4) Quantification: Accurate quantification of lipids is challenging due to their dynamic range and varying concentrations in different sample types that could be several degrees of magnitude. Standardization of sample preparation and analysis is critical to achieving accurate and reproducible results. 5) Analytical techniques: The identification and quantification of lipids require specialized analytical techniques, such as MS and chromatography6. The identification and quantification of lipids can be a challenging task, and it requires careful consideration of the sample matrix, sample preparation, and analytical techniques to achieve accurate and reliable results6. 9.3. Sample preparation and lipid extraction To unravel the biochemical and biophysical intricacies of membrane lipids and elucidate their cellular functions, the preparation of highly pure and morphologically distinct membranes and subcellular fractions is crucial. Effective separation demands homogenization steps that break down cells into subcellular organelles while preserving the integrity of vesicular organelles, encapsulating hydrolases and enzymes capable of lipid composition alteration. If hydrolases and enzymes are released from these vesicular organelles, they could alter the lipid composition of the sample. Therefore, careful handling is necessary to prevent the breakdown of these organelles and ensure that the lipid profile being studied reflects the actual biological state rather than an artifact of the preparation process. This is crucial for accurately analyzing the lipids and understanding their cellular functions6. Tailoring the homogenization step to specific tissue characteristics is paramount and is influenced by diverse operational criteria. Pretreatment with collagenase or cellulase is essential for robust extracellular matrices or cell walls. The choice of homogenization buffers also plays a crucial role in mimicking in vivo pH and solute composition. Various 47 homogenization methods have been used, such as mechanical disruption (Potter-Elvehjem, rotating blade devices like Waring blender and Ultraturax, glass beads) or other shear forces (pressure/cavitation, freeze-thaw). Sonication and enzymatic digestion are also viable methods for cell homogenization to isolate subcellular organelles. Sonication utilizes ultrasonic waves to disrupt cell membranes, while enzymatic digestion employs specific enzymes to break down the cellular matrix gently; it uses enzymes like collagenase, trypsin, or other proteases to break down the extracellular matrix and connective tissues without the need for harsh mechanical forces. However, these methods may lead to an artefactual reshuffling of membrane subfractions or organelle rupturing and fusion of naturally separated membrane domains. For instance, homogenization using a loosely fitted Potter homogenizer results in clump formation, representing associations of different organelle membranes. The homogenization process must be carefully selected and optimized for each tissue type and research objective to achieve precise and meaningful insights into lipid biology. The goal is to maintain the natural state of the lipids as closely as possible, ensuring that the resulting data accurately reflect biological reality and provide a solid foundation for subsequent analyses and interpretations. Effective lipid extraction is crucial for studying membrane lipids' biochemical and biophysical properties. The choice of extraction method depends on the chemical nature of the lipids and their matrix within the cell. For instance, hydrophobic lipids may be extracted using non-polar solvents like ethyl ether or chloroform, while membrane-associated lipids may require polar solvents such as ethanol. Covalently bound lipids necessitate cleavage from complexes through acid-alkaline or enzymatic hydrolysis130. Preventing oxidation is crucial, requiring the use of freshly distilled, peroxide-free solvents. Highly unsaturated lipids may demand de-aeration by nitrogen bubbling. Extraction should be performed at or below room temperature to mitigate lipid peroxidation and 48 hydrolysis. Silanized glassware is essential to prevent lipid adhesion. Since lipids tend to bind to glass due to their amphipathic nature, when working with small quantities of these molecules, the loss of sample due to adhesion to glass can be significant, potentially leading to inaccurate measurements and loss of material, which could be critical in analytical or preparative procedures. Comprehensive lipid extraction protocols, emphasizing the use of glass tubes and creating one- phase systems with polar and non-polar solvents, are outlined on the Cyberlipid website [www.cyberlipid.org]. The Folch method, involving a two-immiscible-phase system, facilitates the separation of lipids from water-soluble constituents. Protocols derived from the Folch procedure, including potential modifications like centrifugation or the addition of saline buffers, are employed for efficient separation. Automation with solid-phase extraction protocols has been developed to enhance efficiency, reduce solvent volume, and minimize lipid oxidation during the simultaneous treatment of multiple samples131,132. Ensuring the integrity of lipid samples during storage is crucial, given the susceptibility of lipids to chemical alterations. The inherent protective mechanisms in living organisms are compromised during lipid isolation, necessitating the addition of antioxidants, such as butylated hydroxytoluene (BHT) and tocopherol, to the lipid extracts. Care must be taken with antioxidant concentrations, with levels below 0.00001 (w/w) recommended to avoid pro-oxidant effects. Contamination with heavy metals, particularly copper and iron, poses an additional risk, and using single-use vessels with polytetrafluoroethylene (PTFE) or Teflon fittings helps mitigate such concerns in lipidomics research6. Exposure to UV light and high storage temperatures can lead to harmful effects, such as the migration of double bonds, diene conjugation, and significant peroxidation, especially in concentrated solutions. A few steps help to mitigate these problems, like, keeping lipid extract at low temperatures (≤−20°C), in solvent-diluted solutions within 49 opaque containers, purging with an inert gas like nitrogen or argon, and sealing with PTFE caps and tape. Following these steps, polyunsaturated glycerophospholipids can remain stable for up to six months, while sphingolipids stay stable for longer periods due to their higher saturation levels. Sterols, steroids, and bile acids exhibit inherent resistance to chemical alterations, while dehydrocholesterol and ergosterol, with double bonds in the B-ring, require prompt analysis, handled under low illumination and an inert atmosphere. Evaporation of the solvent should be performed before analysis using nitrogen at controlled temperatures or a SpeedVac133,134. Figure 11-abriefly explains the sample preparation steps. Figure 1.11. A schematic diagram of the lipidomics approach. a) sample preparation steps, b) separation and detection techniques79 50 10. Different analytical techniques to detect and quantify lipid metabolites 10.1. Immunoassay Immunoassay is a widely used analytical technique that uses antibodies to selectively detect and measure specific molecules in a variety of samples. Immunoassay has been used to measure PUFA metabolites and oxylipins. To develop an immunoassay for PUFA metabolites and oxylipins, antibodies that specifically recognize the target molecules are first generated. These antibodies are then used to capture and detect the target molecules in biological samples. The immunoassay can be performed in a variety of formats, including enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA), and fluorescent immunoassay (FIA). RIA is a sensitive technique that uses radioactive isotopes as labels to detect antigens or antibodies. It involves competition between a radioactive-labeled antigen and an unlabeled antigen for binding to a limited amount of specific antibody. Radioactivity is measured using a scintillation counter, and the amount of radioactivity detected is inversely proportional to the concentration of the antigen or antibody in the sample. FIA uses fluorescent labels to detect antigens or antibodies. When the labeled antigen-antibody complex forms, it emits fluorescence that can be measured. Fluorescence is quantified using a fluorometer, with higher fluorescence indicating higher concentrations of the target antigen or antibody.135 In ELISA, the process begins with the immobilization of the antigen of interest on the surface of a microplate well, providing a stable platform for specific antibody binding (Figure 6, step 1). Following antigen coating, a primary antibody that is specific to the antigen is introduced into the well. This primary antibody binds to the antigen, forming an antigen-antibody complex (Figure 6, step 2). To enable detection, a secondary antibody, which recognizes the primary antibody and is conjugated to an enzyme such as horseradish peroxidase (HRP), is then added. This secondary antibody binds to the primary antibody, creating a sandwich of the antigen and two 51 antibodies (Figure 6, step 3). The addition of a substrate that reacts with HRP, typically TMB (tetramethylbenzidine), leads to a colorimetric change. The enzyme catalyzes the conversion of the substrate into a colored product, indicating the presence of the antigen-antibody complex (Figure 6, step 4). Finally, an acid is added to stop the reaction, fix the color change, and allow for stable measurement. The resulting color intensity, now a stable product, is proportional to the original antigen concentration in the sample, allowing for quantification upon measurement with a spectrophotometer (Figure 1.12, step 5). This colorimetric endpoint is a direct Figure 1.12. Schematic representation of the ELISA process, made by BioRender reflection of the antigen concentration, which can be quantified by comparing it against a standard curve derived from known antigen concentrations. Immunoassays have several advantages for measuring PUFA metabolites and oxylipins. They are specific and can distinguish between different types of metabolites and oxylipins. They are also sensitive, with detection limits in the nanomolar range. Immunoassays can be used to measure these molecules in a wide range of biological samples, including blood, urine, and tissues. However, there are also some limitations to using immunoassays. The antibodies used in the assay may cross-react with other molecules in the sample, leading to false positives or false negatives. The antibodies may also degrade over time, leading to decreased sensitivity and accuracy. Immunoassays are also limited in their ability to detect low levels of the target molecules and may not be able to detect all metabolites and 52 oxylipins. For an extended period, immunoassays such as enzyme immunoassay (EIA) and RIA were the predominant quantitative methods used for oxylipins due to their remarkable sensitivity. RIA has been specifically designed for measuring concentrations of IsoPs (8-iso-PGF2α) in both human plasma and urine136, 15-keto-dihydro-PGF2α in human plasma137, Prostaglandin E2 (PGE2) and LTB4 in human prostate tissues138, PGE2 in human plasma139,140, and an array of compounds including PGE2, PGF2α, PGI2, 6-oxo-PGF2α, TXA2, and TXB2141. Additionally, RIA has been employed to quantify PGF2α, PGI2, TXA2, 13,14-dihydro-15-keto PGF2α (M- PGF2α), 6-keto PGF1α, and TXB2 142 in human follicular fluid. ELISA, akin to EIA and RIA, necessitates specific antibodies. However, due to the inherent structural similarities among oxylipins, obtaining antibodies with the necessary specificity is improbable143. While RIA and EIA exhibit sensitivity levels adequate for quantifying minuscule quantities of oxylipins, they encounter certain limitations when dealing with tissue and plasma samples. These limitations contribute to a reduction in the sensitivity of immunoassays. For example, plasma proteins can bind to eicosanoids, leading to considerable immunological cross-reactivity among commercially available eicosanoid antibodies. This cross-reactivity can introduce challenges in accurately detecting and quantifying specific eicosanoids in biological samples using ELISA, and this limitation potentially affects the reliability and specificity of the assay results. For instance, the PGE2 antibody can manifest notable cross-reactivity with PGE3 and 8-iso-PGF2a144. To address these challenges and to segregate eicosanoids from plasma proteins, liquid-liquid extraction (LLE) or solid-phase extraction (SPE) methods are employed. Furthermore, the chromatographic isolation of eicosanoids becomes imperative to prevent instances of immunological cross- reactivity145. 53 10.2. Mass spectrometry-based techniques 10.2.1. HPLC In recent decades, HPLC has emerged as the predominant method for LC analysis in separating oxylipins, largely replacing thin-layer chromatography. HPLC separates components in a mixture by passing a pressurized liquid solvent (mobile phase) through a column filled with solid particles (stationary phase). Each component interacts differently with the stationary phase based on its physical and chemical properties, leading to varying migration rates and separation. The compounds are then detected and quantified based on their unique characteristics and retention time. Nonetheless, the traditional employment of HPLC with UV detection presents challenges. This is attributed to the low UV absorption range of wavelengths (235–280 nm) of conjugated dienes and keto groups, components commonly found in oxylipin molecules. This low UV absorption range makes it challenging to detect and quantify oxylipins; it can compromise the sensitivity and reliability of the assay results compared to other analytical techniques. Due to the absence of suitable chromophores in most oxylipins, sensitivity and selectivity remain inadequate when analyzing these molecules within intricate cellular or tissue extracts144,146. Consequently, plasma samples containing an abundance of UV-absorbing compounds prove unsuitable for discerning eicosanoids, which exhibit lower concentrations in plasma. Moreover, UV absorption is ineffective for prostaglandins (PGs), as they cannot absorb UV light at analytically useful wavelengths147. Furthermore, HPLC coupled with electrochemical detection enables the quantification of minute quantities (picogram or nanogram amounts) of lipoxin (LX) A4 (LXA4) and LXB4 in extracts of human polymorphonuclear granulocytes, as well as leukotriene (LT) B4 (LTB4) from human polymorphonuclear leukocytes148,149. However, it is possible to utilize UV 54 absorption as supplementary confirmation for HPLC coupled with mass spectrometry (HPLC-MS) measurement outcomes150. HPLC-MS is a powerful analytical technique that can measure and quantify PUFA metabolites and oxylipins in biological samples. In HPLC-MS, the sample is first separated by a chromatography column that differentiates different compounds based on their physical and chemical properties. The separated compounds are then introduced into a mass spectrometer, which detects and identifies the different molecules based on their mass-to-charge ratio (m/z). To measure and quantify PUFA metabolites and oxylipins, the HPLC column is often chosen to separate selected analytes from other components in the sample selectively. Several phases of the HPLC column can be used to separate PUFA metabolites and oxylipins, including normal-phase, reversed-phase, and ion-pairing chromatography. Choosing the appropriate solid phase of the column depends on the physicochemical properties of the compounds of interest and the level of selectivity required for the analysis. HPLC-MS has several advantages for measuring and quantifying PUFA metabolites and oxylipins. HPLC-MS can measure these molecules in a wide range of biological samples, including blood, urine, and tissues. While HPLC-MS is a widely used technique for measuring and quantifying PUFA metabolites and oxylipins, this technique has some limitations. The detection limits for some compounds may be higher than those for other techniques, which can limit the ability to detect low levels of PUFA metabolites or oxylipins in biological samples. Another limitation is that sample preparation can challenge HPLC-MS analysis of PUFA metabolites and oxylipins. Biological samples may contain a complex mixture of compounds, which can interfere with the analysis of the target compounds. Sample preparation may require extraction and purification steps, which can be time-consuming and may introduce variability. The 55 specificity of HPLC-MS can also be a limitation. While HPLC can separate compounds based on their physical properties, it may be hard to distinguish between compounds with very similar structures. This can limit the ability to measure specific metabolites or to distinguish between isomers or different oxidation products. Altogether, while HPLC-MS is a widely used technique for measuring and quantifying PUFA metabolites and oxylipins, there are some limitations to this technique that should be considered when designing experiments and interpreting results. 9.2.2. Chiral chromatography Chiral chromatography coupled with mass spectrometry (Chiral LC-MS) is a specialized technique that can measure and quantify the enantiomers of PUFA metabolites and oxylipins in biological samples. Enantiomers are molecules that are mirror images of each other and have the same chemical and physical properties but rotate polarized light in opposite directions. Chiral chromatography separates enantiomers based on their interactions with a chiral stationary phase, which is a column packed with a chiral molecule that interacts selectively with one enantiomer over the other 151. In this approach, the chiral column separates the enantiomers, and the mass spectrometer detects and identifies the separated compounds based on their mass-to-charge ratio (m/z) and fragmentation patterns. Chiral LC-MS has several advantages for measuring and quantifying PUFA metabolites and oxylipins. It allows for accurately measuring individual enantiomers, which have different biological activities. For example, 12(R)-HETE is reported to be 10–20 times more potent as a chemoattractant for human neutrophils than 12(S)-HETE. 12(R)- HETE, has been reported to increase DNA synthesis and plays a significant role in psoriasis, while increased levels of 12(S)-HETE were found in patients with essential hypertension152. An obstacle encountered in oxylipin analysis is the multitude of isomers generated during non-enzymatic oxidation. Various alternative approaches have been devised to address this concern. One such 56 strategy involves the utilization of specialized chiral columns, including immobilized polysaccharide or protein-based columns151. For example, a Chiralpak AD-RH column (150 × 2.1 mm, 5 μm; Chiral Technologies, West Chester, PA, USA) was employed for chiral HPLC-Tandem Mass Spectrometry (MS-MS) lipidomic profiling of 18-R/S-HEPEs and 18R/S-resolvin (Rv) E2 stereoisomer pairs from human serum153,154. This methodology facilitated the detection of previously unrecognized EPA-derived 18S resolvins, a revelation that required chiral separation. Analyzing biological samples made it possible to identify 18S-HEPE as a precursor to this series of resolvins. Chiral chromatography proves essential for separating established isomer pairs or groups and enables the revelation of novel compounds with potential bioactivity. An example is maresin 2, which exhibited anti-inflammatory and pro-resolving properties155. While chiral chromatography/mass spectrometry is a powerful technique for measuring and quantifying chiral PUFA metabolites and oxylipins, there are also limitations to this technique. These include the necessity for extended equilibration times and employment of an isocratic elution; the mobile phase composition remains constant throughout the analysis: 65:35 methanol: water with 0.1% acetic acid, which results in lengthier analysis periods when compared to other analytical techniques such as GC-MS or HPLC-MS, and reliance on internal standards that may not be easily accessible156, reduced sensitivity (signal-to-noise ratio diminishes with increasing peak width and elution time), restricted high throughput (primarily suited for focused analysis), and heightened susceptibility to alterations in mobile-phase composition157–160. This technique is chiefly applied for the targeted analysis of specific oxylipins due to its limitations. Finally, chiral chromatography/mass spectrometry may not be suitable for oxylipin profiling, which contains hundreds of metabolites, because of practical considerations like complexity and time 57 consumption. Other analytical techniques, such as GC-MS or HPLC-MS, may be more appropriate for these compounds. 10.2.3. Ion mobility spectrometry Ion mobility spectrometry-mass spectrometry (IMS-MS) combines the separation power of ion mobility spectrometry (IMS) with the identification capabilities of mass spectrometry. IMS separates ions based on their size, shape, and charge, which can help to reduce interferences and improve the specificity of the measurement. In IMS-MS, the sample is first ionized, and the resulting ions are introduced into an IMS device. The ions are separated based on their mobility through a buffer gas, usually nitrogen, where the frequency of ion-gas collisions is influenced by their size and shape. The ions are then transferred into the mass spectrometer, where they are separated based on their mass-to-charge ratio. IMS stands as an alternative separation technique with the potential to enhance the identification and characterization of oxylipins within biological samples, particularly In the context of separating isomers and the speed of separation. Integrating IMS between chromatographic and mass spectrometry stages introduces an additional separation dimension, bolstering the reliability of oxylipin identification by incorporating collision-cross-section (CCS) values. The researchers utilized IMS-MS to analyze a panel of lipid mediators, including eicosanoids, endocannabinoids, and specialized pro-resolving mediators. This methodology has effectively segregated 42 isomer pairs or groups using positive or negative ion modes. However, separating enantiomers using IMS remained challenging, as enantiomers have identical CCS values161. In another study, they used a different version of IMS where ions moved through a tube filled with nitrogen gas at low pressure, and ion dimensions and structure influenced their drift time. This mechanism enabled the formation of distinct oxylipin conformers162. However, 58 applying IMS necessitates the construction of supplementary CCS libraries. This involves employing either computationally generated standards or CCS data from existing literature, which introduces complexity to the implementation of this approach. The detection limits for some compounds may be higher than those for other techniques, limiting the ability to detect low levels of PUFA metabolites in biological samples163. Figure 1.13. Drift tube IMS. IMS separates ions based on their mobility through a neutral drift gas under an applied electric field163 10.2.4. Immunoaffinity column chromatography Immunoaffinity column chromatography (IAC) combined with MS is a powerful analytical technique that can measure and quantify PUFA metabolites and oxylipins in biological samples. IAC is a chromatography technique that uses antibodies to selectively isolate specific compounds from a complex sample matrix. The antibodies are immobilized in a column, and the sample is passed through the column, allowing the target compounds to bind to the antibodies. The non- target compounds are then washed away, and the target compounds are eluted from the column for further analysis. MS is a highly sensitive and specific analytical technique that detects and identifies the target compounds eluted from the IAC column164. IAC demonstrates exceptional specificity and selectivity to identify and quantify prostates, isoprostanes (IsoP), and their associated metabolites. Given their typically low abundance (94.5-132.5 pg/ml) in biological fluids 59 like urine, this is a challenge 165. Nevertheless, the utility of IAC is bounded by its sole commercial availability for analyzing 15-F2t-isoprostane, excluding other metabolites derived from n-3 and n- 6 PUFAs166. Researchers explored the potential of IAC to enhance the analysis of trace lipid contents within specific test samples. This dual-extraction protocol, applied to the retrieval of leukotrienes (LTs), offers advantages over conventional solid-phase extractions (SPEs), including mitigated column overload risk, cleaner sample outcomes, and the ability to utilize the flow- through fluid for analyzing compounds not retained by the antibodies. Another research group used Sepharose 4-based IAC columns (4-mL, 1-mL gel resin; Cayman Chemicals, Ann Arbor, MI, USA) for the extraction and subsequent gas chromatography-tandem mass spectrometric (GC– MS/MS) quantification of prostaglandin E1 (PGE1) in human plasma. Nonetheless, the broader adoption of IAC is impeded by the scarcity of readily available antibodies and the intricate nature of antibody production methods, curtailing its widespread applicability164. IAC-MS has several advantages for measuring and quantifying PUFA metabolites and oxylipins. It is highly selective and can isolate specific compounds of interest from a complex biological matrix. It is also highly sensitive, with detection limits in the picomolar range. IAC-MS can measure these molecules in a wide range of biological samples, including blood, urine, and tissues. However, IAC-MS also has limitations. The antibody used for IAC must be specific to the compound of interest, and the antibody binding may be affected by the sample matrix, leading to incomplete isolation of the target compounds. The elution conditions may also affect the stability of the target compounds, leading to degradation or modification. Furthermore, IAC may not be suitable for large-scale sample analysis due to the limited capacity of the antibody column. 60 10.2.5. GC-MS GC is a chromatography technique that separates compounds based on their volatility and polarity. In GC, the sample is vaporized and injected into a column containing a stationary phase, where the compounds are separated based on their vapor pressures and interactions with the stationary phase. The compounds are then detected as they exit the column using a detector, such as a flame ionization detector (FID) or a mass spectrometer. The desire to combine MS with chromatographic methods has persisted, given the heightened sensitivity and specificity of MS compared to alternative chromatographic detectors. The integration of GC with MS (GC-MS) was accomplished in the 1950s, with commercially available instruments emerging in the 1970s. In contemporary times, cost-effective and dependable GC-MS systems have become commonplace in numerous clinical biochemistry laboratories. They play an essential role in various applications167. The most significant advancements in exploring oxylipin levels using GC-MS occurred during the 1980s. For the successful GC analysis of a molecule, the compound must be both volatile and thermally stable. Unfortunately, not all oxylipins fit both criteria. Consequently, the analysis with some oxylipins like dihydroxyeicosatrienoic acids requires the derivatization of carboxyl and hydroxyl groups to enhance their volatility168. Reagents like N, O-bis(trimethylsilyl)- trifluoroacetamide (BSTFA) are used for the silylation of hydroxyl groups, helping their detection through electron ionization (EI)169. To enhance analysis sensitivity, GC is frequently paired with MS detection. This combination allows for measuring multiple analytes within a single sample, significantly lowering routine detection costs147. The integration of GC and MS has found wide application in analyzing oxylipins with urine and plasma samples. The widespread use of GC-MS for quantitatively measuring PGs, TXs, LTs, IsoPs, and other AA metabolites within human urine was underscored170. Subsequently, researchers published various research reports on eicosanoids 61 in plasma, serum, and other biological fluids of healthy individuals, employing validated GC-MS, GC-MS/MS, and LC-MS/MS methods. GC-MS enables the analysis of diverse compounds, such as LTs (LTB4), TXs (TXB2, 11-dh-TXB2), prostacyclin (6keto–PGF1α), prostaglandins (PGF2α, PGE1, PGE2, PGD2), and F2-isoprostanes (15(S)-8-iso-PGF2α)171. Simultaneously, analytical techniques for AA and its metabolites within brain tissues were comprehensively emphasizing the widespread use of GC-MS for the quantification of compounds like PGE2, PGD2, PGF2α, 8,9- dihydroxyeicosatrienoic acid (DiHETrE), 5,6-DiHETrE, 12-HHT, 2-HETE, 3-HETE, 5-HETE, 8,9-HETE, 11,12-HETE, 15-HETE, 8-iso-PGF2α, 9α,11β-PGF2α, 9α,11β-PGF2α, PGE2, PGD2, TXB2, PGF1α, PGF2α, F2-isoprostanes, and AA within brain tissues172. The versatility of GC-MS analysis extends to various other oxylipin types across diverse biological tissues. GC–MS with negative-ion chemical ionization (GC-MS/NICI) was used to analyze EETs, dihydroxyeicosatrienoic acids (DHETs), and 20-hydroxyeicosatetraenoic acid (20-HETE) within coronary venous plasma during coronary artery occlusion and reperfusion in dogs173. GC-MS was integrated after the purification of cell cultures through reverse phase HPLC (RP–HPLC) to investigate the formation of DHETs and HETEs (5,6-DHET, 8,9-DHET, 11,12-DHET, 14,15- DHET, 5-HETE, 8-HETE, 9-HETE, 11-HETE, 12-HETE, and 15-HETE) within human peritoneal macrophage174. GC-MS has several advantages for measuring and quantifying PUFA metabolites and oxylipins, like detection limits in the femtomolar range. GC-MS can also provide information on the structure of the target compounds based on their fragmentation patterns, which can be used to identify specific metabolites or oxylipins. Additionally, GC-MS is widely used in research and clinical settings and has well-established sample preparation and analysis protocols. However, there are some limitations to using GC-MS. One limitation is that GC-MS requires volatile and thermally stable compounds. This may limit 62 the applicability of GC-MS to certain PUFA metabolites or oxylipins, which are too unstable or non-volatile to be analyzed using this technique. In addition, some metabolites and oxylipins may require derivatization to improve their volatility and detectability, which can be time-consuming and may introduce additional sources of variability. While the biological activity of these compounds can depend on their stereochemistry, GC-MS cannot distinguish between the two forms unless chiral columns are used174. 10.2.6. LC-MS/MS The integration of MS with LC (LC-MS) was a logical progression, but advancements in this domain faced constraints for numerous years due to the inherent incompatibility of existing MS ion sources with a continuous liquid stream. Various interfaces were created but proved unreliable, leading to minimal adoption by clinical laboratories. The landscape changed with Fenn's introduction of the ESI source in the 1980s. Manufacturers swiftly incorporated ESI into instruments, revolutionizing protein and peptide biochemistry. Fenn, along with Koichi Tanaka, who developed MALDI, another precious MS ionization technique for biomolecule analysis, received the Nobel Prize in Chemistry in 2002167. Given that LC-MS/MS stands as the most preferred technique for oxylipin analysis, it is crucial to highlight the challenges faced in this regard and the diverse strategies employed by researchers to address them. An inherent advantage of LC-MS/MS is eliminating derivatization requirements, a factor that prevents the introduction of impurities, enhances sensitivity in MS analysis, and concurrently reduces time and costs172. Tandem MS/MS instruments in conjunction with either HPLC or ultra-high-performance liquid chromatography (UHPLC) can analyze multiple analytes175. The primary operating modes of tandem quadrupole MS/MS instruments encompass multiple reaction monitoring (MRM), also known as selected reaction monitoring (SRM). While scanning mode investigates specific mass 63 ranges in either the first or second analyzer, MRM mode produces product ions from pre-selected precursor ions following collision-induced dissociation (CID), thereby enhancing the specificity of the analysis172. This feature facilitates the identification of structurally similar oxylipin isomers176. Relying solely on analyte retention time and MRM transitions for identification has challenges due to various regioisomers with very close physical and chemical properties resulting in similar retention times. Detection in MRM mode proves advantageous in scenarios where LC fails to separate isomeric compounds, like 8 and 12-HETE that coelute. Conversely, when 9 and 12-HETE exhibit akin MS–MS spectra, chromatographic separation becomes imperative175, like the case of PGD2 and PGE2177as well as 8-iso-PGF2 and PGF2 178, where identical mass fragmentation patterns prevail. To maximize specificity and sensitivity in oxylipin analysis, tandem mass spectrometry (MS/MS) equipped with a triple quadrupole (QqQ) detector in MRM mode proves valuable, especially when dealing with co-eluting metabolites176. The utility of triple- quadrupole mass spectrometers in lipidomics research is underscored by their capacity for precursor ion scanning, neutral loss scanning, and unparalleled quantitative analysis precision and accuracy through MRM mode179. Comprising three sequentially arranged quadrupoles, these instruments employ a linear quadrupole configuration where ions of a specific m/z value traverse the quadrupole rods based on applied dc and rf potentials, enabling mass separation akin to selective ion trapping in an ion trap. While the first (Q1) and third (Q3) quadrupoles operate in mass-selective mode, the second quadrupole (Q2) serves as a total ion containment region and gas collision cell, allowing ions above a minimum cut-off m/z value to pass through, inducing ion- molecule collisions and providing chemical information through fragmentation. Triple-quadrupole instruments excel in tandem mass spectrometry, performing low-energy range MS/MS fragmentations in Q2 during product ion scanning, precursor ion scanning, and neutral loss 64 scanning. Operating primarily in MRM mode, they set Q1 to transmit a selected precursor ion and Q3 for a selected product ion, enabling rapid switching for versatile analyte measurement, yielding a substantial improvement in detection selectivity and impressive accuracy and precision. Despite their advantages, triple-quadrupole instruments have inherent limitations, such as low mass resolving power, hindering precise m/z measurement for lipid identification. Additionally, they lack the capability for MSn experiments beyond n = 2 and exhibit low duty-cycle efficiencies in scanning modes. Hybrid instruments like Qq-TOF and QQ-LIT (QTRAP) mass spectrometers have been developed to address some limitations, enhancing resolving power and allowing more precise m/z determinations for lipid analysis180. Currently, focused metabolomics LC-MS/MS strategies for AA metabolites enable the concurrent quantification of over 100 oxylipins with exceptional sensitivity (limit of detection ranging from 0.01 to 0.21 pg on the column) within a runtime of approximately 25 minutes. In these methodologies, the RP-LC interfaces with a highly responsive triple quadrupole (QqQ) MS system that operates in the negative ESI mode175. By integrating ultra-high performance liquid chromatography (UHPLC) separation with MRM transitions executed on a QqQ system, an extensive set of 184 eicosanoid metabolites could be effectively separated and quantified within a brief 5-minute runtime181. To couple LC with MS, careful consideration is given to solvent choice due to the potential for mobile phase additives and buffers to induce ion suppression. For the analysis of oxylipins by LC-MS/MS, weak acids are typically employed in most mobile phases (e.g., 0.1% formic acid (FA) or acetic acid (AA) to improve chromatographic resolution but hinder the formation of carboxylate anions within the ESI source182. Chen et al., in their efforts to optimize ionization processes for enhanced MS signal, assessed various additives (AA, FA, ammonium formate) in conjunction with the same mobile phase composition (50/50, acetonitrile (ACN)/H2O). Their findings indicated that 0.1% FA 65 exhibited lower variance and higher average responses, offering a more consistent and favorable analysis 183. UHPLC offers several advantages in metabolite analysis, including enhanced resolution, faster analysis times, and increased sensitivity. Furthermore, it contributes to reduced solvent consumption, leading to cost savings172. Adopting 2-μm columns instead of conventional HPLC columns significantly improves the chromatographic separation of eicosanoids in human plasma. This change has substantially reduced the analysis duration from the previous range of 20 to 60 minutes down to a more efficient 4 to 12 minutes183. Brose et al. conducted experiments with RP- UHPLC columns, such as BEH C18 (150 × 2.1 mm, 1.7 μm), BEH HILIC (100 × 2.1 mm, 1.7 μm), ACQUITY (Waters, Milford, MA, USA), CSH C18 (150 × 2.1 mm, 1.7 μm), HSS T3 (150 × 2.1 mm, 1.8 μm), and HSS C18 (150 × 2.1 mm, 1.8 μm). They utilized gradients consisting of acidified (ACN)/water, methanol (MeOH)/isopropanol (IPA)/water, and MeOH/water for the quantification of PGs using a Q-TOF instrument (Synapt G2-S; Waters, Milford, MA) with an ESI source. All available PGD2 and PGE2, such as iso-PG, were effectively separated on the HSS T3 column using an ACN/water mixture acidified with 0.1% formic acid (FAc) as a gradient. This separation method also can separate endogenous brain iso-PGs, except PGE2/ent-PGE2 and 8iso-PGE2/15R-PGE2. Notably, this approach resulted in sharper peaks compared to the previously employed LC-MS/MS method, achieving a rapid separation within 4 minutes184. The matrix effect is one of the challenges in quantifying oxylipin by LC-MS/MS, and it refers to the alteration of ionization efficiency for analytes in biological samples due to matrix constituents 185. This phenomenon occurs when other molecules from the samples co-elute with the analytes and impact the ionization process at the electrospray interface. It's worth noting that matrix effects vary depending on the analytes; the most polar compounds tend to exhibit more significant ion 66 suppression, while less polar molecules are less affected by it. These interfering matrix components can originate from the current sample, a previously injected sample, or the column overload186. To effectively mitigate or eliminate the influence of matrix effects, several methods are employed, such as adjustments to the sample extraction methodology, improved chromatographic separation techniques, and the utilization of stable isotope-labeled internal standard (IS)185. These standards enhance the accuracy and reliability of their quantitative measurements by minimizing the impact of matrix effects on the analysis of biological samples. Deuterium-labeled standards can present certain drawbacks, such as exhibiting different retention times compared to analytes, unintended signal amplification, or a potential weakening of ionization. On the other hand, while 13C-labeled standards might theoretically offer advantages for analysis, they are not widely available in the commercial market160. It's crucial to carefully select the appropriate amount of IS since the interaction between the analyte and IS is significantly affected by the concentration, impacting accuracy187. Despite the generally high quality of non- certified standards, it is advisable to calculate correction factors to adjust concentrations and compensate for variations during utilization188. SPE procedure effectively removes or reduces the presence of interfering matrix components that can cause ion suppression or enhancement during the ionization process, helping to minimize the co-elution of matrix components that can lead to matrix effects. Optimized SPE protocols can achieve high and reproducible recoveries of the target analytes, reducing the impact of matrix effects observed in crude biological extracts on quantitative measurements. These procedures involve using Oasis HLB cartridges, Oasis MAX SPE, and incorporating both non-polar and polar washing steps before eluting oxylipins. It's important to note that the extent of the matrix effect can vary depending on the type of tissue under examination and the complexity and richness of the matrix itself189. For instance, the matrix effect was 67 particularly significant in colon tissue analysis, resulting in a signal loss of approximately 50% for PUFA metabolites190. In contrast, when determining urinary levels of oxylipins (e.g., LTB4), the matrix effect led to only a 10% signal loss191. 10.2.7. Imaging mass spectrometry Imaging mass spectrometry (IMS) is a powerful technique enabling the spatial mapping of different lipid distributions in tissue sections based on m/z analysis. The two predominant ion formation methods employed in IMS are MALDI and secondary ion mass spectrometry (SIMS). MALDI typically utilizes N2 UV lasers and infrared laser beams, along with matrices such as 2,5- dihydroxybenzoic acid (DHB), 2,4,6-trihydroxyacetophenone (THAP), and 2- mercaptobenzothiazole (MBT). A novel matrix coating system, the oscillating capillary nebulizer (OCN), has been introduced to enhance matrix homogeneity. Despite its ability to generate intact lipid molecules, MALDI has limitations, including restricted lateral resolution (20- 50 μm), high chemical noise at low m/z, and prolonged acquisition times. In SIMS, the primary ion beam has evolved, providing cleaner spectra at low m/z without needing a matrix. However, one of the challenges with SIMS is the fragmentation of analytes during the ionization process. The high- energy primary ion beam used in SIMS can cause significant fragmentation of biological molecules, making it difficult to obtain complete structural information about the analytes. This limitation can impact the ability to resolve and identify the complete molecular structures present in biological samples. To address this challenge, a massive argon cluster ion source can help reduce analyte fragmentation and resolve analyte fragmentation during ionization. Efforts have been made to combine the strengths of MALDI and SIMS to address their drawbacks, which might offer a promising tool for lipid imaging analyses192. In one study, they introduced a novel method for visualizing fatty acids in mouse retinal samples using silver nanoparticles (AgNPs) as a matrix for 68 MALDI-IMS analyses. The mouse retina, with its intricate eight-layered structure, proved conducive to this approach. Employing a high spatial resolution with a 10 μm scan pitch, they conducted MALDI-IMS on mouse retinal sections, reconstructing ion images for seven fatty acids. Notably, the ion images revealed a six-zonal distribution of free fatty acids, offering insights into their localization within different regions of the retina. Palmitic acid (16:0) exhibited a widespread distribution across all regions except the pigment epithelium (PEM), while LA, oleic acid (18:1), and AA showed equal distribution in the PEM. Stearic acid (18:0) was present in specific regions, including the ganglion cell layer (GC), inner plexiform layer (IP), outer plexiform layer (OP), and PEM. Interestingly, EPA displayed a distribution pattern like palmitic acid. DHA was concentrated in the PEM, revealing its potential significance in visual function, survival, and the prevention of photoreceptor apoptosis. The application of AgNPs in MALDI-IMS proved effective in studying the distribution of long-chain PUFAs and understanding their role in various pathologies, such as age-related macular degeneration and diabetic retinopathy, highlighting the versatility of this technique for investigating heterogeneous biological structures like mouse retinal sections193. 11. Conclusion This chapter has underscored the critical role of PUFA metabolites in the pathology of various diseases, laying the foundation for the importance of oxylipins in medical research and diagnostics. The detailed historical analysis of dietary shifts in omega-3 and omega-6 fatty acid ratios over the centuries has illuminated potential links to prevalent health issues and established a context for the metabolic transformations that have paralleled changes in human diet and lifestyle. The biosynthesis of oxylipins has been elaborated upon, with a comprehensive examination of the metabolic pathways, including COX and LOX, and their physiological impact. The exploration into cytochrome P450-derived oxylipins and epoxide hydrolase has provided a deeper 69 understanding of the complex biochemistry underlying oxylipin functions and their receptors. The chapter has also highlighted the necessity for continual discovery and refinement of bioactive chemicals and biomarkers, detailing the evolution of biomarker identification and its pivotal role in advancing disease diagnostics. Using oxylipins as valuable biomarkers has set the stage for their potential utilization in clinical settings. As we delve deeper into the intricacies of lipid metabolism and its implications for health and disease, the pioneering field of lipidomics emerges as a transformative frontier in biomedical research. The first chapter has laid substantial groundwork, shedding light on the dynamic interplay of lipids within biological systems and the critical insights they offer into cellular functions, disease mechanisms, and potential therapeutic targets. Lipidomics, as a vital component of the 'omics' era, represents a paradigm shift in our approach to understanding lipid-related pathways and networks. It encompasses the large-scale study of pathways and networks of cellular lipids in biological systems, integrating the roles of lipids in cellular processes with their structural diversity and complexity. The challenges discussed in this chapter, from the identification and quantification of lipid species to the elucidation of their functional roles, underscore the need for advanced analytical technologies and interdisciplinary strategies. The comprehensive review of state-of-the-art analytical techniques, such as mass spectrometry coupled with sophisticated separation technologies, highlights the evolution and capabilities of current methodologies in addressing the complexities of lipid profiling. These technological advancements are not merely incremental; they represent leaps forward in our capacity to decipher the lipidome's role in health and disease. Lipidomics' promise lies in its ability to provide a holistic view of lipid physiology and pathophysiology, offering a more granular understanding of disease states and fostering the development of personalized medicine. 70 BIBLIOGRAPHY (1) Saghatelian, A.; McKinney, M. K.; Bandell, M.; Patapoutian, A.; Cravatt, B. F. A FAAH- Regulated Class of N-Acyl Taurines That Activates TRP Ion Channels. Biochemistry 2006, 45 (30), 9007–9015. (2) Klose, C.; Surma, M. A.; Simons, K. Organellar Lipidomics--Background and Perspectives. Curr. Opin. Cell Biol. 2013, 25 (4), 406–413. (3) Sud, M.; Fahy, E.; Cotter, D.; Dennis, E. A.; Subramaniam, S. LIPID MAPS-Nature Lipidomics Gateway: An Online Resource for Students and Educators Interested in Lipids. J. Chem. Educ. 2012, 89 (2), 291–292. (4) Muro, E.; Atilla-Gokcumen, G. E.; Eggert, U. S. Lipids in Cell Biology: How Can We Understand Them Better? Mol. Biol. Cell 2014, 25 (12), 1819–1823. (5) Li-Beisson, Y.; Nakamura, Y.; Harwood, J. Lipids: From Chemical Structures, Biosynthesis, and Analyses to Industrial Applications. Subcell. Biochem. 2016, 86, 1–18. (6) Bou Khalil, M.; Hou, W.; Zhou, H.; Elisma, F.; Swayne, L. A.; Blanchard, A. P.; Yao, Z.; Bennett, S. A. L.; Figeys, D. Lipidomics Era: Accomplishments and Challenges. Mass Spectrom. Rev. 2010, 29 (6), 877–929. (7) Israelachvili, J. Relationship between Chain Protrusions and Surfactant Critical Micelle Concentration Values. Langmuir 1992, 8 (5), 1501–1501. (8) Lasic, D. D. Liposomes; Elsevier Science: London, England, 1993. (9) Deamer, D. The Role of Lipid Membranes in Life’s Origin. Life (Basel) 2017, 7 (1), 5. (10) Murphy, D. J. The Dynamic Roles of Intracellular Lipid Droplets: From Archaea to Mammals. Protoplasma 2012, 249 (3), 541–585. (11) Guha, S.; Calarco, S.; Gachet, M. S.; Gertsch, J. Juniperonic Acid Biosynthesis Is Essential in Caenorhabditis Elegans Lacking Δ6 Desaturase (Fat-3) and Generates New ω-3 Endocannabinoids. Cells 2020, 9 (9), 2127. (12) Alves-Bezerra, M.; Cohen, D. E. Triglyceride Metabolism in the Liver. Comprehensive Physiology. Wiley December 12, 2017, pp 1–22. https://doi.org/10.1002/cphy.c170012. (13) Aneja, R.; Bhati, A.; Hamilton, R. J.; Padley, F. B.; Steven, D. A. Determination of the Fully Saturated Triglyceride Composition of Fats. J. Am. Oil Chem. Soc. 1980, 57 (4), 135–139. (14) Laufs, U.; Parhofer, K. G.; Ginsberg, H. N.; Hegele, R. A. Clinical Review on Triglycerides. Eur. Heart J. 2020, 41 (1), 99–109c. (15) Farmer, J. A. Triglycerides and the Risk of Coronary Artery Disease. Curr. Atheroscler. Rep. 2008, 10 (2), 104–105. 71 (16) Bremer, E. G.; Schlessinger, J.; Hakomori, S. Ganglioside-Mediated Modulation of Cell Growth. Specific Effects of GM3 on Tyrosine Phosphorylation of the Epidermal Growth Factor Receptor. J. Biol. Chem. 1986, 261 (5), 2434–2440. (17) Chen, X.; Shi, C.; Wang, Y.; Yu, H.; Zhang, Y.; Zhang, J.; Li, P.; Gao, J. The Mechanisms of Glycolipid Metabolism Disorder on Vascular Injury in Type 2 Diabetes. Front. Physiol. 2022, 13, 952445. (18) Ryckman, A. E.; Brockhausen, I.; Walia, J. S. Metabolism of Glycosphingolipids and Their Role in the Pathophysiology of Lysosomal Storage Disorders. Int. J. Mol. Sci. 2020, 21 (18), 6881. (19) Miller, W. L.; Auchus, R. J. The Molecular Biology, Biochemistry, and Physiology of Human Steroidogenesis and Its Disorders. Endocr. Rev. 2011, 32 (1), 81–151. (20) Stanczyk, F. Z. Production, Clearance, and Measurement of Steroid Hormones. Glob. Libr. Women S Med. 2009. https://doi.org/10.3843/glowm.10278. (21) Nozawa, Y. Phospholipid Signaling and Cell Function. In Handbook of Neurochemistry and Molecular Neurobiology; Springer US: Boston, MA, 2009; pp 297–311. (22) Peretó, J.; López-García, P.; Moreira, D. Ancestral Lipid Biosynthesis and Early Membrane Evolution. Trends Biochem. Sci. 2004, 29 (9), 469–477. (23) Nyame, K.; Hims, A.; Aburous, A.; Laqtom, N. N.; Dong, W.; Medoh, U. N.; Heiby, J. C.; Xiong, J.; Ori, A. Abu-Remaileh M. Glycerophosphodiesters Inhibit Lysosomal Phospholipid Catabolism in Batten Disease. Mol Cell 2024, 84 (7). (24) Spiegel, S.; Milstien, S. Sphingosine-1-Phosphate: An Enigmatic Signalling Lipid. Nat. Rev. Mol. Cell Biol. 2003, 4 (5), 397–407. (25) Hannun, Y. A.; Obeid, L. M. Principles of Bioactive Lipid Signalling: Lessons from Sphingolipids. Nat. Rev. Mol. Cell Biol. 2008, 9 (2), 139–150. (26) Kolesnick, R. N.; Krönke, M. Regulation of Ceramide Production and Apoptosis. Annu. Rev. Physiol. 1998, 60 (1), 643–665. (27) Quinville, B. M.; Deschenes, N. M.; Ryckman, A. E.; Walia, J. S. A Comprehensive Review: Sphingolipid Metabolism and Implications of Disruption in Sphingolipid Homeostasis. Int. J. Mol. Sci. 2021, 22 (11), 5793. (28) Chen, Y.; Liu, Y.; Sullards, M. C.; Merrill, A. H., Jr. An Introduction to Sphingolipid Metabolism and Analysis by New Technologies. Neuromolecular Med. 2010, 12 (4), 306–319. (29) Jones, E. E.; Dworski, S.; Canals, D.; Casas, J.; Fabrias, G.; Schoenling, D.; Levade, T.; Denlinger, C.; Hannun, Y. A.; Medin, J. A.; Drake, R. R. On-Tissue Localization of Ceramides and Other Sphingolipids by MALDI Mass Spectrometry Imaging. Anal. Chem. 2014, 86 (16), 8303–8311. 72 (30) Wallis, J. G.; Watts, J. L.; Browse, J. Polyunsaturated Fatty Acid Synthesis: What Will They Think of Next? Trends Biochem. Sci. 2002, 27 (9), 467. (31) Emken, E. A.; Adlof, R. O.; Gulley, R. M. Dietary Linoleic Acid Influences Desaturation and Acylation of Deuterium-Labeled Linoleic and Linolenic Acids in Young Adult Males. Biochim. Biophys. Acta 1994, 1213 (3), 277–288. (32) Pawlosky, R. J.; Hibbeln, J. R.; Novotny, J. A.; Salem, N., Jr. Physiological Compartmental Analysis of Alpha-Linolenic Acid Metabolism in Adult Humans. J. Lipid Res. 2001, 42 (8), 1257– 1265. (33) Childs, C. E.; Kew, S.; Finnegan, Y. E.; Minihane, A. M.; Leigh-Firbank, E. C.; Williams, C. M.; Calder, P. C. Increased Dietary α-Linolenic Acid Has Sex-Specific Effects upon Eicosapentaenoic Acid Status in Humans: Re-Examination of Data from a Randomised, Placebo- Controlled, Parallel Study. Nutr. J. 2014, 13 (1), 113. (34) Simopoulos, A. P. Genetic Variants in the Metabolism of Omega-6 and Omega-3 Fatty Acids: Their Role in the Determination of Nutritional Requirements and Chronic Disease Risk. Exp. Biol. Med. (Maywood) 2010, 235 (7), 785–795. (35) Schmitz, G.; Ecker, J. The Opposing Effects of N-3 and n-6 Fatty Acids. Prog. Lipid Res. 2008, 47 (2), 147–155. (36) Mozaffarian, D.; Katan, M. B.; Ascherio, A.; Stampfer, M. J.; Willett, W. C. Trans Fatty Acids and Cardiovascular Disease. Obstet. Gynecol. Surv. 2006, 61 (8), 525–526. (37) Mensink, R. P.; Zock, P. L.; Kester, A. D. M.; Katan, M. B. Effects of Dietary Fatty Acids and Carbohydrates on the Ratio of Serum Total to HDL Cholesterol and on Serum Lipids and Apolipoproteins: A Meta-Analysis of 60 Controlled Trials. Am. J. Clin. Nutr. 2003, 77 (5), 1146– 1155. (38) Kris-Etherton, P. M.; Grieger, J. A.; Etherton, T. D. Dietary Reference Intakes for DHA and EPA. Prostaglandins Leukot. Essent. Fatty Acids 2009, 81 (2–3), 99–104. (39) Thiam, A. R.; Farese, R. V., Jr; Walther, T. C. The Biophysics and Cell Biology of Lipid Droplets. Nat. Rev. Mol. Cell Biol. 2013, 14 (12), 775–786. (40) Pol, A.; Gross, S. P.; Parton, R. G. Review: Biogenesis of the Multifunctional Lipid Droplet: Lipids, Proteins, and Sites. J. Cell Biol. 2014, 204 (5), 635–646. (41) Zechner, R.; Zimmermann, R.; Eichmann, T. O.; Kohlwein, S. D.; Haemmerle, G.; Lass, A.; Madeo, F. FAT SIGNALS--Lipases and Lipolysis in Lipid Metabolism and Signaling. Cell Metab. 2012, 15 (3), 279–291. (42) Eaton, S. Control of Mitochondrial Beta-Oxidation Flux. Prog. Lipid Res. 2002, 41 (3), 197– 239. 73 (43) Krahmer, N.; Farese, R. V., Jr; Walther, T. C. Balancing the Fat: Lipid Droplets and Human Disease. EMBO Mol. Med. 2013, 5 (7), 973–983. (44) Welte, M. A.; Gould, A. P. Lipid Droplet Functions beyond Energy Storage. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 2017, 1862 (10 Pt B), 1260–1272. (45) Goñi, F. M. The Basic Structure and Dynamics of Cell Membranes: An Update of the Singer- Nicolson Model. Biochim. Biophys. Acta 2014, 1838 (6), 1467–1476. (46) de Kroon, A. I. P. M.; Rijken, P. J.; De Smet, C. H. Checks and Balances in Membrane Phospholipid Class and Acyl Chain Homeostasis, the Yeast Perspective. Prog. Lipid Res. 2013, 52 (4), 374–394. (47) Mondal, M.; Mesmin, B.; Mukherjee, S.; Maxfield, F. R. Sterols Are Mainly in the Cytoplasmic Leaflet of the Plasma Membrane and the Endocytic Recycling Compartment in CHO Cells. Mol. Biol. Cell 2009, 20 (2), 581–588. (48) Laude, A. J.; Prior, I. A. Plasma Membrane Microdomains: Organization, Function and Trafficking (Review). Mol. Membr. Biol. 2004, 21 (3), 193–205. (49) Oates, J.; Watts, A. Uncovering the Intimate Relationship between Lipids, Cholesterol and GPCR Activation. Curr. Opin. Struct. Biol. 2011, 21 (6), 802–807. (50) Li, Z.-L.; Buck, M. Computational Modeling Reveals That Signaling Lipids Modulate the Orientation of K-Ras4A at the Membrane Reflecting Protein Topology. Structure 2017, 25 (4), 679-689.e2. (51) Mazhab-Jafari, M. T.; Marshall, C. B.; Smith, M. J.; Gasmi-Seabrook, G. M. C.; Stathopulos, P. B.; Inagaki, F.; Kay, L. E.; Neel, B. G.; Ikura, M. Oncogenic and RASopathy-Associated K- RAS Mutations Relieve Membrane-Dependent Occlusion of the Effector-Binding Site. Proc. Natl. Acad. Sci. U. S. A. 2015, 112 (21), 6625–6630. (51) Bigay, J.; Antonny, B. Curvature, Lipid Packing, and Electrostatics of Membrane Organelles: Defining Cellular Territories in Determining Specificity. Dev. Cell 2012, 23 (5), 886–895. (53) Park, M.-J.; Sheng, R.; Silkov, A.; Jung, D.-J.; Wang, Z.-G.; Xin, Y.; Kim, H.; Thiagarajan- Rosenkranz, P.; Song, S.; Yoon, Y.; Nam, W.; Kim, I.; Kim, E.; Lee, D.-G.; Chen, Y.; Singaram, I.; Wang, L.; Jang, M. H.; Hwang, C.-S.; Honig, B.; Ryu, S.; Lorieau, J.; Kim, Y.-M.; Cho, W. SH2 Domains Serve as Lipid-Binding Modules for PTyr-Signaling Proteins. Mol. Cell 2016, 62 (1), 7–20. (54) Kundu, M.; Mahata, B.; Banerjee, A.; Chakraborty, S.; Debnath, S.; Ray, S. S.; Ghosh, Z.; Biswas, K. Ganglioside GM2 Mediates Migration of Tumor Cells by Interacting with Integrin and Modulating the Downstream Signaling Pathway. Biochim. Biophys. Acta 2016, 1863 (7 Pt A), 1472–1489. (55) Okazaki, Y.; Saito, K. Roles of Lipids as Signaling Molecules and Mitigators during Stress Response in Plants. Plant J. 2014, 79 (4), 584–596. 74 (56) Sunshine, H.; Iruela-Arispe, M. L. Membrane Lipids and Cell Signaling. Curr. Opin. Lipidol. 2017, 28 (5), 408–413. (57) Kris-Etherton, P. M.; Harris, W. S.; Appel, L. J.; AHA Nutrition Committee. American Heart Association. Omega-3 Fatty Acids and Cardiovascular Disease: New Recommendations from the American Heart Association. Arterioscler. Thromb. Vasc. Biol. 2003, 23 (2), 151–152. (58) Correction to: Dietary Fats and Cardiovascular Disease: A Presidential Advisory from the American Heart Association. Circulation 2017, 136 (10), e195. (59) Simopoulos, A. P. The Importance of the Ratio of Omega-6/Omega-3 Essential Fatty Acids. Biomed. Pharmacother. 2002, 56 (8), 365–379. (60) Singh, R. B.; Demeester, F.; Wilczynska, A. The Tsim Tsoum Approaches for Prevention of Cardiovascular Disease. Cardiol. Res. Pract. 2010, 2010, 824938. (61) de Lorgeril, M.; Renaud, S.; Mamelle, N.; Salen, P.; Martin, J. L.; Monjaud, I.; Guidollet, J.; Touboul, P.; Delaye, J. Mediterranean Alpha-Linolenic Acid-Rich Diet in Secondary Prevention of Coronary Heart Disease. Lancet 1994, 343 (8911), 1454–1459. (62) Evolutionary Aspects of Nutrition and Health. Diet, Exercise, Genetics and Chronic Disease; Simopoulos, A. P., Ed. (63) Humphries, S. E. Evolutionary Aspects of Nutrition and Health: Diet, Exercise, Genetics and Chronic Disease. World Review of Nutrition and Dietetics, Vol. 84. Edited by A. p. Simopoulos. Basel: Karger. 1999. Pp. 145. US $172.25, SW.Fr. 198, DM 237. Ann. Hum. Genet. 1999, 63 (4), 377–381. (64) Eaton, S. B.; Konner, M. Paleolithic Nutrition. N. Engl. J. Med. 1985, 312 (5), 283–289. (65) Crawford, M. A.; Wang, Y.; Lehane, C.; Ghebremeskel, K. Fatty Acid Ratios in Free-Living and Domestic Animals. In Modern Dietary Fat Intakes in Disease Promotion; Humana Press: Totowa, NJ, 2010; pp 95–108. (66) Sinclair, A. J.; Slattery, W. J.; O’Dea, K. The Analysis of Polyunsaturated Fatty Acids in Meat by Capillary Gas‐liquid Chromatography. J. Sci. Food Agric. 1982, 33 (8), 771–776. (67) Alt, K. W.; Al-Ahmad, A.; Woelber, J. P. Nutrition and Health in Human Evolution-Past to Present. Nutrients 2022, 14 (17), 3594. (68) Research: From Cells to Animals to Humans. Prog. Lipid Res 2011, 50, 152–192. (69) Buczynski, M. W.; Dumlao, D. S.; Dennis, E. A. Thematic Review Series: Proteomics. An Integrated Omics Analysis of Eicosanoid Biology. J. Lipid Res. 2009, 50 (6), 1015–1038. (70) Mayeux, R. Biomarkers: Potential Uses and Limitations. NeuroRx 2004, 1 (2), 182–188. 75 (71) Pearson, C. E. Repeat Associated Non-ATG Translation Initiation: One DNA, Two Transcripts, Seven Reading Frames, Potentially Nine Toxic Entities! PLoS Genet. 2011, 7 (3), e1002018. (72) Aronson, J. K.; Ferner, R. E. Biomarkers-A General Review. Curr. Protoc. Pharmacol. 2017, 76 (1), 9.23.1-9.23.17. (73) Watanabe, L. M.; de Souza Pinhel, M. A.; Noronha, N. Y.; Nonino, C. B. Creatine Kinase as a Biomarker. In Biomarkers in Disease: Methods, Discoveries and Applications; Springer International Publishing: Cham, 2022; pp 1–23. (74) Rao, A. R.; Motiwala, H. G.; Karim, O. M. A. The Discovery of Prostate‐specific Antigen. BJU Int. 2008, 101 (1), 5–10. (75) Chocholoušková, M.; Jirásko, R.; Vrána, D.; Gatěk, J.; Melichar, B.; Holčapek, M. Reversed Phase UHPLC/ESI-MS Determination of Oxylipins in Human Plasma: A Case Study of Female Breast Cancer. Anal. Bioanal. Chem. 2019, 411 (6), 1239–1251. (76) Earles, S. M.; Bronstein, J. C.; Winner, D. L.; Bull, A. W. Metabolism of Oxidized Linoleic Acid: Characterization of 13-Hydroxyoctadecadienoic Acid Dehydrogenase Activity from Rat Colonic Tissue. Biochim. Biophys. Acta 1991, 1081 (2), 174–180. (77) Nayeem, M. A. Role of Oxylipins in Cardiovascular Diseases. Acta Pharmacol. Sin. 2018, 39 (7), 1142–1154. (78) Xia, F.; He, C.; Ren, M.; Xu, F.-G.; Wan, J.-B. Quantitative Profiling of Eicosanoids Derived from N-6 and n-3 Polyunsaturated Fatty Acids by Twin Derivatization Strategy Combined with LC-MS/MS in Patients with Type 2 Diabetes Mellitus. Anal. Chim. Acta 2020, 1120, 24–35. (79) Camunas-Alberca, S. M.; Moran-Garrido, M.; Sáiz, J.; Villaseñor, A.; Taha, A. Y.; Barbas, C. The Role of Oxylipins and Their Validation as Biomarkers in the Clinical Context. Trends Analyt. Chem. 2023, 164 (117065), 117065. (80) Harbeck, N.; Penault-Llorca, F.; Cortes, J.; Gnant, M.; Houssami, N.; Poortmans, P.; Ruddy, K.; Tsang, J.; Cardoso, F. Breast Cancer. Nat. Rev. Dis. Primers 2019, 5 (1), 66. (81) Malý, M.; Hajšl, M.; Bechyňská, K.; Kučerka, O.; Šrámek, M.; Suttnar, J.; Hlaváčková, A.; Hajšlová, J.; Kosek, V. Lipidomic Analysis to Assess Oxidative Stress in Acute Coronary Syndrome and Acute Stroke Patients. Metabolites 2021, 11 (7), 412. (82) Ye, D.; Zhang, D.; Oltman, C.; Dellsperger, K.; Lee, H.-C.; VanRollins, M. Cytochrome P- 450 Epoxygenase Metabolites of Docosahexaenoate Potently Dilate Coronary Arterioles by Activating Large-Conductance Calcium-Activated Potassium Channels. J. Pharmacol. Exp. Ther. 2002, 303 (2), 768–776. (83) Eriksdotter, M.; Vedin, I.; Falahati, F.; Freund-Levi, Y.; Hjorth, E.; Faxen-Irving, G.; Wahlund, L.-O.; Schultzberg, M.; Basun, H.; Cederholm, T.; Palmblad, J. Plasma Fatty Acid 76 Profiles in Relation to Cognition and Gender in Alzheimer’s Disease Patients during Oral Omega- 3 Fatty Acid Supplementation: The OmegAD Study. J. Alzheimers. Dis. 2015, 48 (3), 805–812. (84) Lebbadi, M.; Julien, C.; Phivilay, A.; Tremblay, C.; Emond, V.; Kang, J. X.; Calon, F. Endogenous Conversion of Omega-6 into Omega-3 Fatty Acids Improves Neuropathology in an Animal Model of Alzheimer’s Disease. J. Alzheimers. Dis. 2011, 27 (4), 853–869. (85) Kalmijn, S. Fatty Acid Intake and the Risk of Dementia and Cognitive Decline: A Review of Clinical and Epidemiological Studies. J. Nutr. Health Aging 2000, 4 (4), 202–207. (86) Borkowski, K.; Pedersen, T. L.; Seyfried, N. T.; Lah, J. J.; Levey, A. I.; Hales, C. M.; Dammer, E. B.; Blach, C.; Louie, G.; Kaddurah-Daouk, R.; Newman, J. W.; Alzheimer’s Disease Metabolomics Consortium. Association of Plasma and CSF Cytochrome P450, Soluble Epoxide Hydrolase, and Ethanolamide Metabolism with Alzheimer’s Disease. Alzheimers. Res. Ther. 2021, 13 (1), 149. (87) Freund-Levi, Y.; Eriksdotter-Jönhagen, M.; Cederholm, T.; Basun, H.; Faxén-Irving, G.; Garlind, A.; Vedin, I.; Vessby, B.; Wahlund, L.-O.; Palmblad, J. Omega-3 Fatty Acid Treatment in 174 Patients with Mild to Moderate Alzheimer Disease: OmegAD Study: A Randomized Double-Blind Trial. Arch. Neurol. 2006, 63 (10), 1402–1408. (88) Wang, W.; Zhu, J.; Lyu, F.; Panigrahy, D.; Ferrara, K. W.; Hammock, B.; Zhang, G. Ω-3 Polyunsaturated Fatty Acids-Derived Lipid Metabolites on Angiogenesis, Inflammation and Cancer. Prostaglandins Other Lipid Mediat. 2014, 113–115, 13–20. (89) Wang, W.; Yang, J.; Nimiya, Y.; Lee, K. S. S.; Sanidad, K.; Qi, W.; Sukamtoh, E.; Park, Y.; Liu, Z.; Zhang, G. ω-3 Polyunsaturated Fatty Acids and Their Cytochrome P450-Derived Metabolites Suppress Colorectal Tumor Development in Mice. J. Nutr. Biochem. 2017, 48, 29– 35. (90) Zhang, J.; Yang, J.; Duval, C.; Edin, M. L.; Williams, A.; Lei, L.; Tu, M.; Pourmand, E.; Song, R.; Graves, J. P.; DeGraff, L. M.; Wong, J. J.-L.; Wang, Y.; Sun, Q.; Sanidad, K. Z.; Wong, S.; Han, Y.; Zhang, Z.; Lee, K. S. S.; Park, Y.; Xiao, H.; Liu, Z.; Decker, E. A.; Cui, W.; Zeldin, D. C.; Zhang, G. CYP Eicosanoid Pathway Mediates Colon Cancer-Promoting Effects of Dietary Linoleic Acid. FASEB J. 2023, 37 (7), e23009. (91) Ye, J.; Ghosh, S. Omega-3 PUFA vs. NSAIDs for Preventing Cardiac Inflammation. Front. Cardiovasc. Med. 2018, 5, 146. (92) Peterson, L. D.; Jeffery, N. M.; Thies, F.; Sanderson, P.; Newsholme, E. A.; Calder, P. C. Eicosapentaenoic and Docosahexaenoic Acids Alter Rat Spleen Leukocyte Fatty Acid Composition and Prostaglandin E2 Production but Have Different Effects on Lymphocyte Functions and Cell-Mediated Immunity. Lipids 1998, 33 (2), 171–180. (93) Thies, F.; Nebe-von-Caron, G.; Powell, J. R.; Yaqoob, P.; Newsholme, E. A.; Calder, P. C. Dietary Supplementation with Gamma-Linolenic Acid or Fish Oil Decreases T Lymphocyte Proliferation in Healthy Older Humans. J. Nutr. 2001, 131 (7), 1918–1927. 77 (94) Calder, P. C. N-3 Polyunsaturated Fatty Acids, Inflammation, and Inflammatory Diseases. Am. J. Clin. Nutr. 2006, 83 (6 Suppl), 1505S-1519S. (95) Reddy, J. K.; Hashimoto, T. Peroxisomal Beta-Oxidation and Peroxisome Proliferator- Activated Receptor Alpha: An Adaptive Metabolic System. Annu. Rev. Nutr. 2001, 21 (1), 193– 230. (96) Wang, Y.-X. PPARs: Diverse Regulators in Energy Metabolism and Metabolic Diseases. Cell Res. 2010, 20 (2), 124–137. (97) Grygiel-Górniak, B. Peroxisome Proliferator-Activated Receptors and Their Ligands: Nutritional and Clinical Implications--a Review. Nutr. J. 2014, 13 (1), 17. (98) Barquissau, V.; Ghandour, R. A.; Ailhaud, G.; Klingenspor, M.; Langin, D.; Amri, E.-Z.; Pisani, D. F. Control of Adipogenesis by Oxylipins, GPCRs and PPARs. Biochimie 2017, 136, 3– 11. (99) Guimarães, R. C.; Gonçalves, T. T.; Leiria, L. O. Exploiting Oxidized Lipids and the Lipid- Binding GPCRs against Cardiometabolic Diseases. Br. J. Pharmacol. 2021, 178 (3), 531–549. (100) Klingenberg, M. Pigments of Rat Liver Microsomes. Arch. Biochem. Biophys. 1958, 75 (2), 376–386. (101) Garfinkel, D. Studies on Pig Liver Microsomes. I. Enzymic and Pigment Composition of Different Microsomal Fractions. Arch. Biochem. Biophys. 1958, 77 (2), 493–509. (102) Guengerich, F. P. Mechanisms of Cytochrome P450-Catalyzed Oxidations. ACS Catal. 2018, 8 (12), 10964–10976. (103) Meunier, B.; de Visser, S. P.; Shaik, S. Mechanism of Oxidation Reactions Catalyzed by Cytochrome P450 Enzymes. Chem. Rev. 2004, 104 (9), 3947–3980. (104) Li, Z.; Jiang, Y.; Guengerich, F. P.; Ma, L.; Li, S.; Zhang, W. Engineering Cytochrome P450 Enzyme Systems for Biomedical and Biotechnological Applications. J. Biol. Chem. 2020, 295 (3), 833–849. (105) Drew, H. R.; Wing, R. M.; Takano, T.; Broka, C.; Tanaka, S.; Itakura, K.; Dickerson, R. E. Structure of a B-DNA Dodecamer: Conformation and Dynamics. Proc. Natl. Acad. Sci. U. S. A. 1981, 78 (4), 2179–2183. (106) Guengerich, F. P.; Waterman, M. R.; Egli, M. Recent Structural Insights into Cytochrome P450 Function. Trends Pharmacol. Sci. 2016, 37 (8), 625–640. (107) Black, S. D. Membrane Topology of the Mammalian P450 Cytochromes. FASEB J. 1992, 6 (2), 680–685. (108) Šrejber, M.; Navrátilová, V.; Paloncýová, M.; Bazgier, V.; Berka, K.; Anzenbacher, P.; Otyepka, M. Membrane-Attached Mammalian Cytochromes P450: An Overview of the 78 Membrane’s Effects on Structure, Drug Binding, and Interactions with Redox Partners. J. Inorg. Biochem. 2018, 183, 117–136. (109) Guengerich, F. P. Human Cytochrome P450 Enzymes. In Cytochrome P450; Springer International Publishing: Cham, 2015; pp 523–785. (110) Preissner, S. C.; Hoffmann, M. F.; Preissner, R.; Dunkel, M.; Gewiess, A.; Preissner, S. Polymorphic Cytochrome P450 Enzymes (CYPs) and Their Role in Personalized Therapy. PLoS One 2013, 8 (12), e82562. (111) Xu, L.-H.; Fushinobu, S.; Ikeda, H.; Wakagi, T.; Shoun, H. Crystal Structures of Cytochrome P450 105P1 from Streptomyces Avermitilis: Conformational Flexibility and Histidine Ligation State. J. Bacteriol. 2009, 191 (4), 1211–1219. (112) Xu, L.-H.; Du, Y.-L. Rational and Semi-Rational Engineering of Cytochrome P450s for Biotechnological Applications. Synth. Syst. Biotechnol. 2018, 3 (4), 283–290. (113) Shoieb, S. M.; El-Ghiaty, M. A.; Alqahtani, M. A.; El-Kadi, A. O. S. Cytochrome P450- Derived Eicosanoids and Inflammation in Liver Diseases. Prostaglandins Other Lipid Mediat. 2020, 147 (106400), 106400. (114) El-Sherbeni, A. A.; Aboutabl, M. E.; Zordoky, B. N. M.; Anwar-Mohamed, A.; El-Kadi, A. O. S. Determination of the Dominant Arachidonic Acid Cytochrome P450 Monooxygenases in Rat Heart, Lung, Kidney, and Liver: Protein Expression and Metabolite Kinetics. AAPS J. 2013, 15 (1), 112–122. (115) Isobe, Y.; Itagaki, M.; Ito, Y.; Naoe, S.; Kojima, K.; Ikeguchi, M.; Arita, M. Comprehensive Analysis of the Mouse Cytochrome P450 Family Responsible for Omega-3 Epoxidation of Eicosapentaenoic Acid. Sci. Rep. 2018, 8 (1). https://doi.org/10.1038/s41598-018-26325-4. (116) Konkel, A.; Schunck, W.-H. Role of Cytochrome P450 Enzymes in the Bioactivation of Polyunsaturated Fatty Acids. Biochim. Biophys. Acta 2011, 1814 (1), 210–222. (117) Fer, M.; Dréano, Y.; Lucas, D.; Corcos, L.; Salaün, J.-P.; Berthou, F.; Amet, Y. Metabolism of Eicosapentaenoic and Docosahexaenoic Acids by Recombinant Human Cytochromes P450. Arch. Biochem. Biophys. 2008, 471 (2), 116–125. (118) Westphal, C.; Konkel, A.; Schunck, W.-H. Cytochrome P450 Enzymes in the Bioactivation of Polyunsaturated Fatty Acids and Their Role in Cardiovascular Disease. Adv. Exp. Med. Biol. 2015, 851, 151–187. (119) Arnold, C.; Markovic, M.; Blossey, K.; Wallukat, G.; Fischer, R.; Dechend, R.; Konkel, A.; von Schacky, C.; Luft, F. C.; Muller, D. N.; Rothe, M.; Schunck, W.-H. Arachidonic Acid- Metabolizing Cytochrome P450 Enzymes Are Targets of ω-3 Fatty Acids. J. Biol. Chem. 2010, 285 (43), 32720–32733. 79 (120) Friedberg, T.; Holler, R.; Löllmann, B.; Arand, M.; Oesch, F. The Catalytic Activity of the Endoplasmic Reticulum-Resident Protein Microsomal Epoxide Hydrolase towards Carcinogens Is Retained on Inversion of Its Membrane Topology. Biochem. J. 1996, 319 (1), 131–136. (121) Harris, T. R.; Hammock, B. D. Soluble Epoxide Hydrolase: Gene Structure, Expression and Deletion. Gene 2013, 526 (2), 61–74. (122) Decker, M.; Arand, M.; Cronin, A. Mammalian Epoxide Hydrolases in Xenobiotic Metabolism and Signalling. Arch. Toxicol. 2009, 83 (4), 297–318. (123) Borhan, B.; Jones, A. D.; Pinot, F.; Grant, D. F.; Kurth, M. J.; Hammock, B. D. Mechanism of Soluble Epoxide Hydrolase. J. Biol. Chem. 1995, 270 (45), 26923–26930. (124) Bogyo, M.; Rudd, P. M. New Technologies and Their Impact on “omics” Research. Curr. Opin. Chem. Biol. 2013, 17 (1), 1–3. (125) Copyright. In Fundamentals of Advanced Omics Technologies: From Genes to Metabolites; Comprehensive analytical chemistry; Elsevier, 2014; p iv. (126) Dennis, E. A. Lipidomics Joins the Omics Evolution. Proc. Natl. Acad. Sci. U. S. A. 2009, 106 (7), 2089–2090. (127) Alex Brown, H. Lipidomics: When Apocrypha Becomes Canonical. Curr. Opin. Chem. Biol. 2012, 16 (1–2), 221–226. (128) Hu, C.; van der Heijden, R.; Wang, M.; van der Greef, J.; Hankemeier, T.; Xu, G. Analytical Strategies in Lipidomics and Applications in Disease Biomarker Discovery. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2009, 877 (26), 2836–2846. (129) Köfeler, H. C.; Fauland, A.; Rechberger, G. N.; Trötzmüller, M. Mass Spectrometry Based Lipidomics: An Overview of Technological Platforms. Metabolites 2012, 2 (1), 19–38. (130) Wislet-Gendebien, S.; Visanji, N. P.; Whitehead, S. N.; Marsilio, D.; Hou, W.; Figeys, D.; Fraser, P. E.; Bennett, S. A. L.; Tandon, A. Differential Regulation of Wild-Type and Mutant Alpha-Synuclein Binding to Synaptic Membranes by Cytosolic Factors. BMC Neurosci. 2008, 9 (1), 92. (131) Ruiz-Gutiérrez, V.; Pérez-Camino, M. C. Update on Solid-Phase Extraction for the Analysis of Lipid Classes and Related Compounds. J. Chromatogr. A 2000, 885 (1–2), 321–341. (132) Persson, E.; Löfgren, L.; Hansson, G.; Abrahamsson, B.; Lennernäs, H.; Nilsson, R. Simultaneous Assessment of Lipid Classes and Bile Acids in Human Intestinal Fluid by Solid- Phase Extraction and HPLC Methods. J. Lipid Res. 2007, 48 (1), 242–251. (133) Sobolik, K. D.; Gremillion, K. J.; Whitten, P. L.; Watson, P. J. Sex Determination of Prehistoric Human Paleofeces. Am. J. Phys. Anthropol. 1996, 101 (2), 283–290. 80 (134) Lin, D. S.; Connor, W. E. Fecal Steroids of the Coprolite of a Greenland Eskimo Mummy, AD 1475: A Clue to Dietary Sterol Intake. Am. J. Clin. Nutr. 2001, 74 (1), 44–49. (135) Fowler, E.; Cheng, N. Comparison of Radioimmunoassay and ELISA Methods for Detection of Antibodies to Chromatin Components. J. Immunol. Methods 1983, 62 (3), 297–303. (136) Basu, S. Radioimmunosassay of 8-Iso-Prostaglandin F2α: An Index for Oxidative Injury via Free Radical Catalysed Lipid Peroxidation. Prostaglandins Leukot. Essent. Fatty Acids 1998, 58 (4), 319–325. (137) Basu, S. Radioimmunoassay of 15-Keto-13, 14-Dihydro-Prostaglandin F2alpha: An Index for Inflammation via Cyclooxygenase Catalysed Lipid Peroxidation. (138) Larré, S.; Tran, N.; Fan, C.; Hamadeh, H.; Champigneulles, J.; Azzouzi, R.; Cussenot, O.; Mangin, P.; Olivier, J. L. PGE2 and LTB4 Tissue Levels in Benign and Cancerous Prostates. Prostaglandins Other Lipid Mediat. 2008, 87 (1–4), 14–19. (139) Aslan, M.; Aslan, I.; Özcan, F.; Eryılmaz, R.; Ensari, C. O.; Bilecik, T. A Pilot Study Investigating Early Postoperative Changes of Plasma Polyunsaturated Fatty Acids after Laparoscopic Sleeve Gastrectomy. Lipids Health Dis. 2014, 13 (1), 62. (140) Neupert, W.; Oelkers, R.; Brune, K.; Geisslinger, G. A New Reliable Chemiluminescence Immunoassay (CLIA) for Prostaglandin E2 Using Enhanced Luminol as Substrate. Prostaglandins 1996, 52 (5), 385–401. (141) Jeremy, J. Y.; Okonofua, F. E.; Thomas, M.; Wojdyla, J.; Smith, W.; Craft, I. L.; Dandona, P. Oocyte Maturity and Human Follicular Fluid Prostanoids, Gonadotropins, and Prolactin after Administration of Clomiphene and Pergonal. J. Clin. Endocrinol. Metab. 1987, 65 (3), 402–406. (142) Ylikorkala, O.; Tenhunen, A. Follicular Fluid Prostaglandins in Endometriosis and Ovarian Hyperstimulation**Supported by the Finnish Cancer Society and Yrjö Jahnsson Foundation. Fertil. Steril. 1984, 41 (1), 66–69. (143) Morgan, A. H.; Hammond, V. J.; Morgan, L.; Thomas, C. P.; Tallman, K. A.; Garcia-Diaz, Y. R.; McGuigan, C.; Serpi, M.; Porter, N. A.; Murphy, R. C.; O’Donnell, V. B. Quantitative Assays for Esterified Oxylipins Generated by Immune Cells. Nat. Protoc. 2010, 5 (12), 1919– 1931. (144) Wang, D.; DuBois, R. N. Measurement of Eicosanoids in Cancer Tissues. Methods Enzymol. 2007, 433, 27–50. (145) Jaffe, B. M.; Behrman, H. R.; Parker, C. W. Radioimmunoassay Measurement of Prostaglandins E, A, and F in Human Plasma. J. Clin. Invest. 1973, 52 (2), 398–405. (146) Gatti, R.; Gotti, R.; Cavrini, V.; Soli, M.; Bertaccini, A.; Carparelli, F. Stability Study of Prostaglandin E1(PGE1) in Physiological Solutions by Liquid Chromatography (HPLC). Int. J. Pharm. 1995, 115 (1), 113–117. 81 (147) Maskrey, B. H.; O’Donnell, V. B. Analysis of Eicosanoids and Related Lipid Mediators Using Mass Spectrometry. Biochem. Soc. Trans. 2008, 36 (Pt 5), 1055–1059. (148) Herrmann, T.; Steinhilber, D.; Knospe, J.; Roth, H. J. Determination of Picogram Amounts of Lipoxin A4 and Lipoxin B4 by High-Performance Liquid Chromatography with Electrochemical Detection. J. Chromatogr. 1988, 428, 237–245. (149) Herrmann, T.; Steinhilber, D.; Roth, H. J. Determination of Leukotriene B4 by High- Performance Liquid Chromatography with Electrochemical Detection. J. Chromatogr. 1987, 416, 170–175. (150) Hartung, N. M.; Mainka, M.; Kampschulte, N.; Ostermann, A. I.; Schebb, N. H. A Strategy for Validating Concentrations of Oxylipin Standards for External Calibration. Prostaglandins Other Lipid Mediat. 2019, 141, 22–24. (151) Schneider, C.; Yu, Z.; Boeglin, W. E.; Zheng, Y.; Brash, A. R. Enantiomeric Separation of Hydroxy and Hydroperoxy Eicosanoids by Chiral Column Chromatography. Methods Enzymol. 2007, 433, 145–157. (152) Yeung, J.; Holinstat, M. 12-Lipoxygenase: A Potential Target for Novel Anti-Platelet Therapeutics. Cardiovasc. Hematol. Agents Med. Chem. 2011, 9 (3), 154–164. (153) Oh, S. F.; Pillai, P. S.; Recchiuti, A.; Yang, R.; Serhan, C. N. Pro-Resolving Actions and Stereoselective Biosynthesis of 18S E-Series Resolvins in Human Leukocytes and Murine Inflammation. J. Clin. Invest. 2011, 121 (2), 569–581. (154) Oh, S. F.; Vickery, T. W.; Serhan, C. N. Chiral Lipidomics of E-Series Resolvins: Aspirin and the Biosynthesis of Novel Mediators. Biochim. Biophys. Acta 2011, 1811 (11), 737–747. (155) Deng, B.; Wang, C.-W.; Arnardottir, H. H.; Li, Y.; Cheng, C.-Y. C.; Dalli, J.; Serhan, C. N. Maresin Biosynthesis and Identification of Maresin 2, a New Anti-Inflammatory and pro- Resolving Mediator from Human Macrophages. PLoS One 2014, 9 (7), e102362. (156) Mesaros, C.; Blair, I. A. Targeted Chiral Analysis of Bioactive Arachidonic Acid Metabolites Using Liquid-Chromatography-Mass Spectrometry. Metabolites 2012, 2 (2), 337– 365. (157) Bayer, M.; Mosandl, A.; Thaçi, D. Improved Enantioselective Analysis of Polyunsaturated Hydroxy Fatty Acids in Psoriatic Skin Scales Using High-Performance Liquid Chromatography. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2005, 819 (2), 323–328. (158) Fuchs, D.; Hamberg, M.; Sköld, C. M.; Wheelock, Å. M.; Wheelock, C. E. An LC-MS/MS Workflow to Characterize 16 Regio- and Stereoisomeric Trihydroxyoctadecenoic Acids. J. Lipid Res. 2018, 59 (10), 2025–2033. (159) Lee, S. H.; Blair, I. A. Targeted Chiral Lipidomics Analysis of Bioactive Eicosanoid Lipids in Cellular Systems. BMB Rep. 2009, 42 (7), 401–410. 82 (160) Massey, K. A.; Nicolaou, A. Lipidomics of Oxidized Polyunsaturated Fatty Acids. Free Radic. Biol. Med. 2013, 59, 45–55. (161) Kyle, J. E.; Aly, N.; Zheng, X.; Burnum-Johnson, K. E.; Smith, R. D.; Baker, E. S. Evaluating Lipid Mediator Structural Complexity Using Ion Mobility Spectrometry Combined with Mass Spectrometry. Bioanalysis 2018, 10 (5), 279–289. (162) Hinz, C.; Liggi, S.; Mocciaro, G.; Jung, S.; Induruwa, I.; Pereira, M.; Bryant, C. E.; Meckelmann, S. W.; O’Donnell, V. B.; Farndale, R. W.; Fjeldsted, J.; Griffin, J. L. A Comprehensive UHPLC Ion Mobility Quadrupole Time-of-Flight Method for Profiling and Quantification of Eicosanoids, Other Oxylipins, and Fatty Acids. Anal. Chem. 2019, 91 (13), 8025–8035. (163) Li, L.; Gu, H.; Lv, Y.; Zhang, Y.; He, X.; Li, P. Ultra-Fast Polarity Switching, Non- Radioactive Drift Tube for the Miniaturization of Drift-Time Ion Mobility Spectrometer. Sensors (Basel) 2022, 22 (13), 4866. (164) Moser, A. C.; Hage, D. S. Immunoaffinity Chromatography: An Introduction to Applications and Recent Developments. Bioanalysis 2010, 2 (4), 769–790. (165) Tsikas, D.; Schwedhelm, E.; Suchy, M.-T.; Niemann, J.; Gutzki, F.-M.; Erpenbeck, V. J.; Hohlfeld, J. M.; Surdacki, A.; Frölich, J. C. Divergence in Urinary 8-Iso-PGF2α (IPF2α-III, 15- F2t-IsoP) Levels from Gas Chromatography–Tandem Mass Spectrometry Quantification after Thin-Layer Chromatography and Immunoaffinity Column Chromatography Reveals Heterogeneity of 8-Iso-PGF2α. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2003, 794 (2), 237–255. (166) Vigor, C.; Bertrand-Michel, J.; Pinot, E.; Oger, C.; Vercauteren, J.; Le Faouder, P.; Galano, J.-M.; Lee, J. C.-Y.; Durand, T. Non-Enzymatic Lipid Oxidation Products in Biological Systems: Assessment of the Metabolites from Polyunsaturated Fatty Acids. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2014, 964, 65–78. (167) Pitt, J. J. Principles and Applications of Liquid Chromatography-Mass Spectrometry in Clinical Biochemistry. Clin. Biochem. Rev. 2009, 30 (1), 19–34. (168) O’Donnell, V. B.; Maskrey, B.; Taylor, G. W. Eicosanoids: Generation and Detection in Mammalian Cells. Methods Mol. Biol. 2009, 462, 5–23. (169) Mueller, M. J.; Mène-Saffrané, L.; Grun, C.; Karg, K.; Farmer, E. E. Oxylipin Analysis Methods. Plant J. 2006, 45 (4), 472–489. (170) Tsikas, D. Application of Gas Chromatography-Mass Spectrometry and Gas Chromatography-Tandem Mass Spectrometry to Assess in Vivo Synthesis of Prostaglandins, Thromboxane, Leukotrienes, Isoprostanes and Related Compounds in Humans. J. Chromatogr. B Biomed. Sci. Appl. 1998, 717 (1–2), 201–245. 83 (171) Tsikas, D.; Zoerner, A. A. Analysis of Eicosanoids by LC-MS/MS and GC-MS/MS: A Historical Retrospect and a Discussion. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2014, 964, 79–88. (172) Puppolo, M.; Varma, D.; Jansen, S. A. A Review of Analytical Methods for Eicosanoids in Brain Tissue. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2014, 964, 50–64. (173) Nithipatikom, K.; DiCamelli, R. F.; Kohler, S.; Gumina, R. J.; Falck, J. R.; Campbell, W. B.; Gross, G. J. Determination of Cytochrome P450 Metabolites of Arachidonic Acid in Coronary Venous Plasma during Ischemia and Reperfusion in Dogs. Anal. Biochem. 2001, 292 (1), 115– 124. (174) Werner, K.; Schaefer, W. R.; Schweer, H.; Deppert, W. R.; Karck, U.; Zahradnik, H. P. Characterization and Identification of Cytochrome P450 Metabolites of Arachidonic Acid Released by Human Peritoneal Macrophages Obtained from the Pouch of Douglas. Prostaglandins Leukot. Essent. Fatty Acids 2002, 67 (6), 397–404. (175) Willenberg, I.; Ostermann, A. I.; Schebb, N. H. Targeted Metabolomics of the Arachidonic Acid Cascade: Current State and Challenges of LC-MS Analysis of Oxylipins. Anal. Bioanal. Chem. 2015, 407 (10), 2675–2683. (176) Petta, T.; Moraes, L. A. B.; Faccioli, L. H. Versatility of Tandem Mass Spectrometry for Focused Analysis of Oxylipids. J. Mass Spectrom. 2015, 50 (7), 879–890. (177) Masoodi, M.; Nicolaou, A. Lipidomic Analysis of Twenty-Seven Prostanoids and Isoprostanes by Liquid Chromatography/Electrospray Tandem Mass Spectrometry. Rapid Commun. Mass Spectrom. 2006, 20 (20), 3023–3029. (178) Kortz, L.; Dorow, J.; Becker, S.; Thiery, J.; Ceglarek, U. Fast Liquid Chromatography- Quadrupole Linear Ion Trap-Mass Spectrometry Analysis of Polyunsaturated Fatty Acids and Eicosanoids in Human Plasma. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2013, 927, 209–213. (179) Larsen, A.; Uran, S.; Jacobsen, P. B.; Skotland, T. Collision-Induced Dissociation of Glycero Phospholipids Using Electrospray Ion-Trap Mass Spectrometry. Rapid Commun. Mass Spectrom. 2001, 15 (24), 2393–2398. (180) Ekroos, K.; Chernushevich, I. V.; Simons, K.; Shevchenko, A. Quantitative Profiling of Phospholipids by Multiple Precursor Ion Scanning on a Hybrid Quadrupole Time-of-Flight Mass Spectrometer. Anal. Chem. 2002, 74 (5), 941–949. (181) Wang, Y.; Armando, A. M.; Quehenberger, O.; Yan, C.; Dennis, E. A. Comprehensive Ultra- Performance Liquid Chromatographic Separation and Mass Spectrometric Analysis of Eicosanoid Metabolites in Human Samples. J. Chromatogr. A 2014, 1359, 60–69. (182) Gandhi, A. S.; Budac, D.; Khayrullina, T.; Staal, R.; Chandrasena, G. Quantitative Analysis of Lipids: A Higher-Throughput LC-MS/MS-Based Method and Its Comparison to ELISA. Future Sci. OA 2017, 3 (1), FSO157. 84 (183) Chen, G.-Y.; Zhang, Q. Comprehensive Analysis of Oxylipins in Human Plasma Using Reversed-Phase Liquid Chromatography-Triple Quadrupole Mass Spectrometry with Heatmap- Assisted Selection of Transitions. Anal. Bioanal. Chem. 2019, 411 (2), 367–385. (184) Brose, S. A.; Baker, A. G.; Golovko, M. Y. A Fast One-Step Extraction and UPLC–MS/MS Analysis for E2/D2 Series Prostaglandins and Isoprostanes. Lipids 2013, 48 (4), 411–419. (185) Viswanathan, C. T.; Bansal, S.; Booth, B.; DeStefano, A. J.; Rose, M. J.; Sailstad, J.; Shah, V. P.; Skelly, J. P.; Swann, P. G.; Weiner, R. Quantitative Bioanalytical Methods Validation and Implementation: Best Practices for Chromatographic and Ligand Binding Assays. Pharm. Res. 2007, 24 (10), 1962–1973. (186) Taylor, P. J. Matrix Effects: The Achilles Heel of Quantitative High-Performance Liquid Chromatography-Electrospray-Tandem Mass Spectrometry. Clin. Biochem. 2005, 38 (4), 328– 334. (187) Araujo, P.; Janagap, S.; Holen, E. Application of Doehlert Uniform Shell Designs for Selecting Optimal Amounts of Internal Standards in the Analysis of Prostaglandins and Leukotrienes by Liquid Chromatography-Tandem Mass Spectrometry. J. Chromatogr. A 2012, 1260, 102–110. (188) Moraes, L. A.; Giner, R. M.; Paul-Clark, M. J.; Perretti, M.; Perrett, D. An Isocratic HPLC Method for the Quantitation of Eicosanoids in Human Platelets. Biomed. Chromatogr. 2004, 18 (1), 64–68. (189) Schoeman, J. C.; Harms, A. C.; van Weeghel, M.; Berger, R.; Vreeken, R. J.; Hankemeier, T. Development and Application of a UHPLC-MS/MS Metabolomics Based Comprehensive Systemic and Tissue-Specific Screening Method for Inflammatory, Oxidative and Nitrosative Stress. Anal. Bioanal. Chem. 2018, 410 (10), 2551–2568. (190) Le Faouder, P.; Baillif, V.; Spreadbury, I.; Motta, J.-P.; Rousset, P.; Chêne, G.; Guigné, C.; Tercé, F.; Vanner, S.; Vergnolle, N.; Bertrand-Michel, J.; Dubourdeau, M.; Cenac, N. LC-MS/MS Method for Rapid and Concomitant Quantification of pro-Inflammatory and pro-Resolving Polyunsaturated Fatty Acid Metabolites. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2013, 932, 123–133. (191) Newman, J. W.; Watanabe, T.; Hammock, B. D. The Simultaneous Quantification of Cytochrome P450 Dependent Linoleate and Arachidonate Metabolites in Urine by HPLC-MS/MS. J. Lipid Res. 2002, 43 (9), 1563–1578. (192) Maroun Bou Khalil, W.; Hou, H.; Zhou, F.; Elisma, L. A.; Swayne, A. P.; Blanchard, Z.; Steffany, A. L.; Bennett, D. Zemin Yao, Steffany A. L. (193) Goto-Inoue, N.; Hayasaka, T.; Zaima, N.; Setou, M. Imaging Mass Spectrometry for Lipidomics. Biochim. Biophys. Acta 2011, 1811 (11), 961–969. 85 IN CHAPTER 2. QUANTITATIVE PROFILING METHOD FOR OXYLIPINS CAENORHABDITIS ELEGANS BY LIQUID CHROMATOGRAPHY COUPLED WITH TANDEM MASS SPECTROMETRY PUBLICATION NOTICE The following chapter is a modified version of a manuscript published by “Elham Pourmand, Fan Zhang, Morteza Sarparast, Jamie K. Alan, and Kin Sing Stephen Lee. Quantitative Profiling Method for Oxylipins in Neurodegenerative Diseases by Liquid Chromatography Coupled with Tandem Mass Spectrometry”. It is available at bioRxiv [Preprint]. 2023, Oct 3:2023.10.02.560544. doi: 10.1101/2023.10.02.560544 86 ABSTRACT Aging is a critical factor for several chronic diseases, including diabetes, neuropathy, hypertension, cancer, and neurodegenerative diseases. Understanding aging mechanisms, which remain largely unexplored, could lead to novel treatments for these conditions. Recent research highlights the importance of cytochrome P450 (CYP)-epoxide hydrolase (EH) metabolites of polyunsaturated fatty acids (PUFAs) as key lipid mediators in various physiological processes. However, their role in aging is not well understood, partly due to the complexity of aging studies. We address this gap using the model organism Caenorhabditis elegans (C. elegans), a well-established tool in aging research, to study the interplay between aging and CYP-EH metabolism of PUFA. Initially, we will ensure the CYP-EH metabolism of PUFA is conserved between humans and C. elegans. Given the structural similarity yet low abundance of the 56 different PUFA CYP-EH metabolites potentially presented in C. elegans, analyzing them in C. elegans presents a significant challenge. This study aims to develop a method using high-performance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS) to quantify these metabolites in C. elegans. Our optimized method enhances sensitivity, enabling us to track each metabolite throughout C. elegans' lifespan using minimal samples. Our results indicate that C. elegans produces PUFA CYP-EH metabolites like mammals, including humans. Furthermore, our method has successfully identified changes in PUFA CYP-EH metabolites profiles resulting from the inhibition of EH in C. elegans. The developed analytical methodology will not only deepen our understanding of the molecular relationship between PUFA CYP-EH metabolism and aging but also shed light on new mechanisms by which aging influences disease via PUFA metabolic pathways. 87 1. Introduction Aging is one of the major risk factors for several diseases, particularly neurodegenerative diseases. For example, the prevalence of Alzheimer’s Disease (AD) will double every 5 years after age 651. The number of patients with AD is expected to triple by 2050 without an effective treatment2. Thus, understanding and modifying the aging process could significantly impact patients with AD and other aging-associated neurodegenerative diseases. Recent research showed that dietary lipids have significant effects on aging and cell senescence3–5, and their downstream oxidized polyunsaturated fatty acid (PUFA) metabolites (namely oxylipins) are key lipid mediators for many disease states, including but not limited to cancer, lupus, cardiovascular6-8, and neurodegenerative diseases9-12. An increase in the expression of enzymes that produce oxylipins, including lipoxygenase (LOX)13, Cyclooxygenase (COX)14, soluble epoxide hydrolase (sEH), and microsomal epoxide hydrolase (mEH), is observed in AD compared to healthy individuals15-17. As a result, endogenous levels of specific oxylipins produced by sEH are upregulated in AD individuals. Moreover, inhibiting the epoxide hydrolases (EH), endogenous enzymes metabolizing epoxy fatty acids to corresponding 1,2-diols, by pharmaceutical inhibitors or a genetic knockout have shown to be beneficial in rescuing diseases, including Parkinson’s disease (PD) and AD 18,19(Figure 1). Our recent study also indicated that dihydroxyeicosadienoic acid (DHED), an epoxide hydrolase metabolite of dihomo-gamma-linolenic acid, induces degeneration of dopaminergic neurons through ferroptosis, which could affect aging11. However, the exact role of oxylipins in neurodegeneration and whether they play any critical role in aging remains elusive 20. Therefore, investigating whether aging has any effects on the oxylipins level and vice versa will help to develop effective preventative measures and treatments for age-related diseases, including neurodegenerative diseases. However, investigating the molecular interactions between oxylipins 88 and aging poses several challenges. First, the mechanisms of aging are understudied and complex. Therefore, conducting studies on an intact animal will likely be beneficial. In addition, aging is a long process, and the experimental settings are difficult to control throughout the aging studies 21- 23. Moreover, oxylipins are largely low in abundance, and there are hundreds of different oxylipins produced in animals. To further complicate the picture, the structures of these oxylipins are very similar, but their functions are very diverse and sometimes in opposition24. Besides, the endogenous levels for some oxylipins can change up to three orders of magnitude depending on the disease status25. Therefore, it is necessary to develop an analytical method to monitor the in vivo level of each oxylipin to understand how the aging process affects oxylipin biosynthesis comprehensively. Yet, this creates difficult analytical challenges. Figure 2.1. Metabolic pathway of polyunsaturated fatty acids. Cytochrome 450 (CYP), epoxide hydrolase (EH), and epoxide hydrolase inhibitor (EHI). These structures are examples of one of the regio/stereoisomers To solve these challenges, this study used Caenorhabditis elegans (C. elegans) as an animal model to investigate the molecular interaction between aging and oxylipin biosynthesis. C. elegans has 89 been used for aging research for decades, owing to its short lifespan, ease of handling and conducting imaging studies, adaptability to high-throughput assays, neuronal structures, and functions homologous to humans, and the presence of a highly conserved genome (>60 to 80%) between C. elegans and humans26. In addition, we developed an analytical method using high- performance liquid chromatography (HPLC) coupled with tandem mass spectrometry to monitor and quantify the oxylipin profile in C. elegans. The developed method is a combination of different analytical techniques, including solid phase extraction (SPE), HPLC, an electrospray ionization (ESI) tandem mass spectrometry to tackle all the challenges associated with qualification and quantification of oxylipins in C. elegans. Our results indicate that we successfully developed a fully validated analytical method to quantify the oxylipin profile in C. elegans. Furthermore, we found that C. elegans has a conserved cytochrome P450-epoxide hydrolase (CYP-EH) metabolic pathway for PUFAs. Our method will eventually facilitate an understanding of the biological functions of oxylipins in aging. 2. Results and discussion 2.1. Mass spectrometry optimization The LC and MS parameters were optimized to improve oxylipin detection because these molecules are present at nM-level concentrations, which is very low. Considering this, the mass spectrometer operating parameters, including collision and cone voltages, were optimized to increase the signal- to-noise ratio while simultaneously decreasing the limit of detection (LOD). In a quadrupole-based mass spectrometer, dwell time, which is the amount of time a mass spectrometer spends measuring the abundances of ions of a particular mass-to-charge ratio during a single sampling period, controls the frequency of data acquisition. Dwell time affects both sensitivity and signal-to-noise ratio. Longer dwell times can increase sensitivity and provide more measurements across a peak. 90 However, the relationship between dwell time and noise is complex, depending on factors such as shot noise, chemical noise, and electronic noise. The optimal dwell time balances these factors to maximize the signal-to-noise ratio while maintaining sufficient data points across chromatographic peaks for accurate quantitation27. Thus, to get the maximum number of measurements with minimum noise, the acquisition was divided into ten functions based on the retention times of the compounds (Table 1). We employed the "auto" feature in the MassLynx V4.2 software, which automatically calculated the dwell time by considering the number of transitions in each function. By utilizing this approach, we ensured sufficient time for acquiring reliable signal-to-noise ratios and an appropriate number of measurements per peak. For LC-MS analysis, it's generally recommended to have at least 10 data points across a chromatographic peak for reliable quantitation. Furthermore, a type I internal standard (IS) was used to normalize the loss of oxylipins due to sample preparation. Type I internal standards are essentially multiply deuterated versions of selected oxylipins that can be easily differentiated by mass spectrometry. The rationale for selecting such internal standards lies in their similarity to the physical properties of the target compounds. This similarity ensures that the loss experienced during extraction is consistent, allowing for the generalization of the loss to other oxylipins, therefore making an accurate normalization. Another advantage of using deuterated oxylipins as internal standards is that their endogenous concentrations in biological samples are negligible owing to the low natural abundance of deuterium. As a result, the signal measured in mass spectrometry analysis is solely derived from the spiked internal standard with known concentration. A list of ISs, their concentrations, and more detail are provided in (Appendix Table S1). To achieve optimal sensitivity and selectivity in this method, a thorough analysis was conducted to identify specific fragments that are generated from the ionized molecules of the sample by 91 collision induced dissociation. Most transitions were selected based on identifying cleavages near double bonds or functional groups like epoxy or hydroxyl modification to provide isomer-selective detection. In certain cases, prioritizing selectivity over sensitivity was deemed crucial, leading to selecting a more distinctive fragmentation ion, even if it was not the most sensitive fragment. This approach ultimately contributes to a more accurate analysis less likely to be influenced by interferences from other oxylipins. For instance, in the case of the CYP metabolite of EPA, 8,9- EpETE, a fragment ion with a m/z 127 was chosen instead of a more sensitive fragment ion with a m/z 255, with the primary aim of enhancing the selectivity of analysis of these regioisomers at the presence of 11,12- EpETE. Selecting m/z 127 is crucial since these two isomers have very close retention times and the exact same molecular weight (Appendix Figure S1). The summary of optimized transitions and mass spectrometric parameters is listed in Table 1. By meticulously selecting suitable multiple reaction monitoring (MRM) transitions, optimizing cone and collision energy, and employing the appropriate LC mobile phases and gradient (Appendix Table S2)28, we achieved successful separation and detection of distinct metabolites of PUFAs and regioisomers (Appendix Figure S2A), with all these considerations, the total number of oxylipins identified in our method is 110 compounds, which is higher compared to other studies29,30. Specifically, the method used by Misheva et al. included 93 oxylipins, while Chhonker et al. reported 66 oxylipins29,30. While these two oxylipin studies predominantly focused on LOX and COX metabolites, our comprehensive method includes LOX and COX enzymes’ metabolites and CYP metabolites like, epoxy and dihydroxy metabolites from all six of the most common PUFA. 92 Table 2.1. Mass spectrometry optimized parameters for each oxylipin metabolite. Acquisition function number (F) and the parameters include multiple reaction monitoring (MRM)(m/z of precursor and product ion), collision potential (CE), cone voltage (CV), retention time (RT)(min), the lower limit of quantification (LLOQ) (nM), the limit of detection (LOD)(nM), the upper limit of quantification (ULOQ)(nM) of oxylipin metabolite Analytes MRM CV CE RT LLOQ LOD F (Volt) (Volt) (min) (nM) (nM) 15,16-DiHODE 12,13-DiHODE 9,10-DiHODE 17,18-DiHETE 14,15-DiHETE 11,12-DiHETE 12,13-DiHOME 8,9-DiHETE 9,10-DiHOME 14,15-DiHETrE 11,12-DiHETrE 20-HEPE 14,15 DHED 9-HOTrE 10,11-DiHDPE 13-HOTrE 18-HEPE 8,9-DiHETrE 7,8-DiHDPE 8-HEPE 12-HEPE 5,6-DiHETrE 13-HODE 5-HEPE 15,16 -EpODE 15-HETE 17,18-EpETE 9,10-EpODE 12,13-EpODE 11-HETE 14,15-EpETE 11,12-EpETE 12-HETE 8,9-EpETE 8-HETE 9-HETE 5,6-EpETE 5-HETE 19,20-EpDPE 8,9-EpEDE 311.20 >223.00 311.20 >183.00 311.20 >201.00 335.20 >247.00 337.20 > 207.00 335.10 >167.00 313.20 >183.00 335.20 >127.00 313.20 >201.00 337.20 >207.00 337.20 >167.00 317.20 >243.00 339.10 >209.00 293.20 >171.00 361.20 >153.00 293.20 >195.00 317.10 >215.00 337.20 >127.00 361.20 >113.00 317.10 >155.00 317.10 >179.00 337.20 >145.00 295.20 >195.00 317.10 >115.00 293.10 >235.00 319.20 >219.00 317.20 >215.00 293.10 >171.00 293.10 >183.00 319.20 >167.00 317.20 >207.00 317.20 >167.00 319.10 >179.00 317.20 >127.00 319.10 >155.00 319.00 >151.00 317.20 >189.00 319.20 >115.00 343.20 >299.00 321.20 >157.00 6.31 6.33 6.35 6.63 6.97 7.15 7.30 7.50 7.70 8.00 8.60 8.65 8.84 8.84 8.91 9.02 9.03 9.18 9.64 9.91 9.91 10.06 10.49 10.58 10.91 10.97 11.03 11.11 11.37 11.48 11.60 11.77 11.82 11.99 11.95 12.26 12.35 12.72 12.80 15.05 0.063 1.250 0.063 0.139 0.063 0.063 0.063 0.250 0.063 0.063 0.250 0.625 0.125 0.312 0.125 0.625 0.625 0.250 0.625 0.208 0.312 0.250 1.000 0.625 0.500 0.312 0.416 0.104 2.500 0.125 0.167 0.250 0.500 1.250 0.500 1.250 0.500 0.167 0.625 0.416 22 22 16 16 16 16 22 20 22 16 22 16 22 16 16 16 16 22 22 16 10 16 16 16 16 10 10 16 16 16 10 10 16 16 16 16 10 16 10 16 7 7 7 7 8 8 8 8 8 8 2 2 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 5 6 6 6 6 6 6 6 6 6 5 44 36 52 39 33 36 20 36 20 33 51 28 28 40 46 36 44 15 44 36 44 52 20 36 34 45 20 28 34 28 27 15 44 20 44 28 20 28 27 28 93 0.021 0.416 0.021 0.046 0.021 0.021 0.021 0.083 0.021 0.021 0.083 0.208 0.042 0.104 0.042 0.208 0.208 0.083 0.208 0.069 0.104 0.083 0.334 0.069 0.083 0.104 0.139 0.035 0.833 0.042 0.055 0.083 0.167 0.416 0.167 0.139 0.167 0.055 0.208 0.139 ULO Q (nM) 50 1000 25 1000 25 25 50 100 50 50 200 500 100 125 100 250 500 200 250 100 50 100 400 500 200 250 200 50 2000 100 200 100 400 200 400 500 400 200 250 500 Table 2.1 (cont’d) 12,13-EpOME 14,15-EpETrE 9,10-EpOME 16,17 EpDPE 13,14-EpDPE 10,11-EpDPE 7,8-EpDPE 8,9-EpETrE 14,15 -EpEDE 5,6-EpETrE 11,12-EpETrE 11,12-EpEDE 295.20 >195.00 319.20 >219.00 295.20 >171.00 343.20 >274.00 343.10 >193.00 343.10 >153.00 343.10 >109.00 319.20 >155.00 321.10 >221.00 319.20 >191.00 319.20 >167.00 321.20 >181.00 6 5 5 5 5 5 5 5 5 5 5 5 20 33 20 20 30 36 44 27 36 20 27 36 16 10 16 10 10 10 16 10 16 10 10 16 13.02 13.15 13.31 13.40 13.56 13.73 14.09 14.25 14.58 14.61 14.93 15.14 0.312 0.625 0.312 1.250 0.208 0.250 1.250 1.250 0.104 2.500 0.208 0.416 0.104 0.208 0.104 0.416 0.069 0.028 0.416 0.416 0.035 0.832 0.069 0.139 250 100 50 500 100 40 1000 500 50 1000 250 500 2.2. Method validation 2.2.1. LOQ and linearity Limit of detection (LOD) and limit of quantification (LOQ) are essential parameters that indicate the lowest analyte concentrations that can be reliably measured by an analytical method. In this study, LOD is defined as the amount of a sample required to produce a signal-to-noise ratio (S/N) of 3 or greater, while the LOQ demands an S/N of 10 or higher to be acceptable. Four consecutive runs of calibration standards (Appendix Table S3) were analyzed within the same day to determine these critical thresholds. This approach allowed for the establishment of LOQs ranging from 0.063 nM to 10 nM and LODs spanning from 0.021 nM to 5 nM, where the analytes were evaluated at a final volume of 100 μL ethanol (75% v/v with water containing a 10 nM concentration of CUDA which is the type II internal standard) (Table 2.1). CUDA is a synthetic fatty acid metabolite mimic, which shares similar chemical and physical properties with oxylipin, making it a dependable control and internal standard. Moreover, CUDA is stable, nonendogenous, and commercially available. The least favorable LOD and LOQ were observed for 12,13-DiHODE (0.416, 1.25 nM), 5,6- DiHETE (5, 10 nM), and 12,13-EpODE (0.833, 2.5 nM), whereas the best sensitivity or lowest LOD and LOQ belong to these three compounds 15,16-DiHODE, 14,15-DiHETE, and 9,10- 94 DiHOME which both have LOD at 0.021 and LOQ at 0.063 nM) (Table 1). The linearity of the method was three orders of magnitude for most compounds when it is evaluated within the concentration range between the lower limit of quantification (LLOQ) and the upper limit of quantification (ULOQ), as determined by the calibration curves constructed for each analyte (Appendix Figures S2A, and S3). 2.2.2. Assessing the accuracy and precision of the method To evaluate the accuracy and precision of the method, we prepared five replicates of quality control samples, including the lower limit of quantification (LLOQ), the lower quality control (LQC), middle-quality control (MQC), and higher quality control (HQC). For intra-day accuracy assessment, the percent difference between the mean concentration of the analytical run and the expected concentration was determined. Five replicates were injected three times within the same day. This approach allowed for a robust evaluation of the method’s consistency and reliability during a single day of operation. Inter-day accuracy was also evaluated by injecting five replicates once daily for three consecutive days. This assessment provided insights into the method’s performance over multiple days, highlighting its stability and dependability across an extended period. Intra-day precision was evaluated as the relative standard deviation of three rounds of injection on the same day. While inter-day precision represented the relative standard deviation of the measurements (n=5, for each QC concentration) injected on three different days. Both precision values were calculated, and most of them were found to be within the acceptable criteria (≤ 25%) (Table 2.2). Less than 20 compounds showed accuracy and precision out of the accepted range for at least one of the QCs, except for some of the LLOQ samples. The compounds that exhibited the least favorable accuracy and precision in the analysis were 5,6-DiHETE and 5,6-DiHETrE. This outcome was anticipated due to the compounds’ susceptibility to intramolecular lactonization 95 (Appendix Figure S4)31. Furthermore, compounds such as 9-HODE, and 13-HODE displayed precision and accuracy values that fell outside the acceptable range. It is worth mentioning that the accuracy and precision of certain prostaglandins and lipoxins were also found to be lower than the acceptable criteria, which explains why these compounds were excluded from our analysis (Table 2.2). However, this observation does not raise concern, as these compounds are not naturally present in C. elegans. This is further supported by the fact that C. elegans lacks homologs for LOX and COX enzymes, which are responsible for the endogenous production of these compounds32. As a result, we expected that these compounds might not be produced in significant amounts. Furthermore, our oxylipin analysis with C. elegans also demonstrated that the metabolites from LOX and COX metabolic pathways were not detected in our method (Appendix Table S4). Therefore, the data related to these metabolites were excluded from this study; a list of all compounds is provided in supporting information (Appendix Table S5). The two figures of merit were acquired for all oxylipins listed in (Table 2.2), except for metabolites from COX and LOX enzymes. Moving forward, we will primarily concentrate on CYP and EH enzyme metabolites. Despite these findings, in general, intra-day precision was found to be better than inter-day precision. Several factors could contribute to this observation, including instrument stability, reagent and sample stability, calibration curve stability, and instrument drift. In this study, it is noteworthy to highlight those certain compounds, including metabolites derived from DGLA, three distinct isomeric forms of EpEDE, and DHED, were successfully synthesized and comprehensively characterized for the very first time. As a result, the incorporation of these oxylipins in our method enables us to determine these important lipids in C. elegans, like DHED which induces neurodegeneration and ferroptosis11, thereby contributing valuable insights to the quantitative lipidomics and targeted metabolomics field. As previously mentioned, our study 96 encompasses epoxy and diol metabolites, including DGLA metabolites (EpEDE and DHED), which were not included in other publications29,30. In terms of accuracy and precision for the epoxy and dihydroxy PUFA, our methodology demonstrated an acceptable range (<20%) of interday and intraday variability across different QC samples. Generally, variation in results is common in oxylipin analysis. A study involving five independent laboratories evaluated the technical variability and comparability of 133 oxylipins using a standardized procedure. Overall, intra-day variability (CV <15%) and inter-day variability (CV <15%) were observed respectively for 85% and 73% of oxylipins33. Several key factors contribute to poor accuracy and precision in oxylipin measurements, including degradation of the compound, incomplete extraction, matrix effects, instrument calibration, instrumental variation, and human error. In this study, we implemented several strategies to mitigate degradation, including the use of antioxidants, purging with argon gas, and storage at -80°C. To address issues associated with incomplete extraction and matrix effects, we strategically employed ten different deuterated ISs. For reducing instrumental variations, we utilized CUDA as a type II internal standard and performed daily calibrations. The accuracy and precision of our method are perfectly comparable with other studies (add two new references), the only oxylipins that did not achieve accuracy or precision levels better than 20% are highlighted in bold in Table 2.2. 97 Table 2.2. The accuracy and precision of the method described in this study. Accuracy and precision of oxylipins, Inter- and Intra-day precision represents the relative standard deviation (RSD) of the measurements (n=5). The intra-day accuracy was determined as the percent difference between the mean concentration per analytical run and the expected concentration. N.D. refers to the concentration below LOQ. Accuracy (Acc.) and precision of eicosanoids metabolites, Interday. precision (Pres.) represents the relative standard deviation of the measurements (n=5). The intra-day accuracy was determined as the percent difference between the mean concentration per analytical run and the expected concentration. The LLQC, the LQC, MQC concentrations are explained in Appendix 2.2 . The bold numbers are accuracy or precision higher than 20, and not detected compounds are shown as (N.D.) LLQC MQC LQC Compound 9,10-DiHODE 5,15-DiHETE 17,18-DiHETE 14,15-DiHETE 11,12-DiHETE 12,13-DiHOME 8,9-DiHETE 9,10-DiHOME 14,15-DiHETrE 19,20-DiHDPE 5,6-DiHETE 16,17-DiHDPE 13,14-DiHDPE 11,12-DiHETrE 20-HEPE 9-HOTrE 14,15 DHED 18-HEPE 8,9-DiHETrE 11,12 DHED 19-HETE 15-HEPE 20-HETE 8,9 DHED 7,8-DiHDPE 12-HEPE 8-HEPE 5,6-DiHETrE 22-HDHA 13-HODE 5-HEPE Intra-day Interday Acc. Prec. Acc Prec. Acc. Prec. Acc. Prec. Acc. Prec Acc. Prec. Intra-day Intra-day Interday Interday 9 8 7 8 8 8 9 13 6 7 31 7 10 9 8 13 5 6 7 6 4 7 7 6 6 6 5 4 6 7 6 15 4 1 10 27 7 23 7 25 7 38 13 22 7 20 21 27 10 22 8 75 38 -11 7 -15 11 -14 9 -28 10 -18 7 -18 7 -20 7 -17 8 -8 5 -13 5 6 -5 7 -6 7 -4 3 5 -11 6 -17 5 -3 6 -18 5 40 17 -18 5 20 8 0 9 20 9 22 13 20 15 4 15 21 12 -4 20 20 9 9 18 463 86 7 4 5 -1 6 2 14 -23 10 -24 8 0 17 -21 12 -21 15 -15 8 -8 12 -11 - 10 9 9 3 17 14 -16 12 -12 6 40 84 8 -9 7 24 9 -10 20 -2 19 19 19 11 21 13 20 25 443 2 -3 2 -10 -17 -1 -5 -7 -13 -1 -5 1 8 15 -16 -15 34 9 21 -9 20 6 14 15 16 17 14 17 13 15 88 7 5 7 7 9 6 5 6 11 10 11 12 8 9 7 5 22 6 9 8 27 -20 34 20 6 58 25 19 16 22 300 3 2 15 -19 -11 5 -15 -15 -8 -5 2 0.0 -5 19 -2 -3 65 0.0 302 -1 66 15 11 21 16 84 15 20 21 14 49 8 13 17 16 23 13 21 27 25 14 14 19 9 15 16 12 11 11 22 16 17 -21 21 24 17 23 23 85 23 29 268 -2 -6 21 -11 -3 7 -1 7 -7 11 9 5 1 17 -5 -8 45 7 289 -1 17 15 7 19 21 22 8 93 16 11 54 12 7 16 8 29 14 5 13 16 13 14 15 9 10 10 7 15 14 25 8 -15 -10 26 27 26 28 22 17 27 23 108 -9 -18 -16 -29 -24 -22 -23 -20 -12 -9 -6 -7 -9 3 -16 -17 4 -16 26 -17 98 Table 2.2 (cont’d) 9,10-EpODE 17-HDoHE 12,13-EpODE 15-oxo-ETE 11-HETE 14,15-EpETE 9-oxo-ODE 11,12-EpETE 12-HETE 8-HETE 8,9-EpETE 15(S)-HETrE 12-oxo-ETE 9-HETE 5,6-EpETE 5-HETE 19,20-EpDPE 12,13-EpOME 14,15-EpETrE 9,10-EpOME 16,17-EpDPE 13,14-EpDPE 10,11-EpDPE 5-oxo-ETE 11,12-EpETrE 7,8-EpDPE 8,9-EpETrE 14,15EpEDE 8,9EpEDE 11,12EpEDE 15,16-DiHODE 12,13-DiHODE 10,11-DiHDPE 13-HOTrE 15,16-EpODE 15-HETE 17,18-EpETE N.D. N.D. -2 17 14 -3 ND ND 5 14 -2 -5 12 -13 -10 20 -6. -1 20 75 85 12 -13 -5 11 -12 -9 18 -13 -11 28 -19 2 12 2 2 17 -36 -24 13 -21 -17 26 16 34 12 0.0 0.0 10 -10 -1 22 92 276 11 -2 -3 14 171 553 7 -5 -3 10 -8 -9 11 0.0 -8 10 -9 -7 6 8 5 N.D. 3 N.D. -4 12 -12 - 4 11 5 -3 11 -2 -10 7 6 23 20 13 10 9 13 -10 11 7 -19 -1 28 N.D. N.D. 19 -2 12 5 -8 19 -8 11 6 9 6 8 9 16 9 16 15 12 11 12 12 50 11 12 46 10 113 9 7 5 8 5 9 8 7 7 16 26 9 14 6 22 9 15 -24 -12 -28 -13 -25 -15 -2 -26 -9. -7 -18 -21 -28 -14 25 -18 -31 15 -25 15. -19 -21 -22 -4 -20 -19 -13 -25 -24 -20 -7 -9 -11 -24 -23 -8 -27 19 6 14 7 13 8 7 14 5 8 13 8 8 8 25 5 12 17 10 19 6 9 7 12 6 6 6 13 7 5 6 6 8 6 18 9 19 -24 -8 -26 -15 -25 -12 -1 -20 -2 4 -7 -15 -28 -2 134 -7 -25 15 -27 18 -16 -24 -24 -3 -19 -19 -15 -28 -25 -24 10 7 -6 -20 -25 -6 -24 14 6 11 7 15 7 6 16 7 10 16 9 7 11 49 9 11 17 10 17 7 9 7 11 6 6 9 9 8 5 12 11 7 7 15 6 14 9 5 2 -2 1 7 6 -20 6 5 -16 7 -28 -3 -13 1 -17 -6 -10 9 -7 -13 -13 -14 -1 -6 -1 -3 3 3 10 10 5 -19 -4 16 -1 11 7 10 9 4 10 10 13 6 8 14 7 12 9 65 8 15 15 7 5 6 6 8 13 6 6 6 7 6 7 6 5 7 13 13 7 12 4 4 0 1 -2 2 7 -3 10 9 2 3 -29 8 6 4 -4 10 -14 11 -3 -15 -15 -9 0 -6 -2 -10 -0 -3 13 7 3 -5 -2 9 -1 11 8 8 8 6 10 9 9 7 5 9 10 15 7 91 6 10 10 8 6 6 6 8 14 6 7 5 7 5 6 6 7 6 6 10 8 11 2.2.3. Evaluating recoveries and extraction efficiencies Recoveries were calculated for C.elegans samples spiked with known concentration of IS type I. Recovery percentages were calculated ranging from 72%-108% (n=3). These values indicate that Type I standards were successfully extracted and recovered during the sample preparation process. The recovery percentage of each IS represents the recovery of all oxylipins in the corresponding data acquisition function. The recovery percentage suggests a reliable and efficient methodology 99 for oxylipins. The precision was satisfactory, with all IS RSDs falling below 12%, except for 8,9- EpETrE-d11. Our results demonstrate that the method provides consistent results across replicates, highlighting its suitability for precise quantification of analytes. The recovery result is comparable with other oxylipin studies in human plasma and serum 34, the recovery percentage of 9-HODE- d4, LTB4-d4, and 5-HETE-d8 was higher in our method than in mouse plasma and liver35 (Table 2.3). Table 2.3. Calculated recovery percentage based on type I standard to study the efficiency of solid phase extraction, n=3 Internal standard 6 keto PGF1α -d4 PGE2-d9 PGB2-d4 LTB4-d4 8,9-DiHETrE-d11 9-HODE-d4 15-HETE-d8 5-HETE-d8 8,9-EpETrE-d11 Spiked conc. (nM) 40 4 16 10 16 16 20 40 40 RC % (mean ± SEM) 72.7 ± 7.3 95.3 ± 7.8 101.4 ± 6.3 81.8 ± 4.8 108.7± 3.9 86.5± 2.5 96.2± 4.4 95.3± 3.9 76.5± 12.6 2.3. C. elegans sample analysis To assess the biological variability of the developed method, we analyzed the oxylipins in C. elegans samples, where age-synchronized worms were grown on OP50, and collected on day 1 of adulthood (Appendix. Figure S5). 2.3.1. Comparing extraction performance of two commonly used homogenization solvents Tissue and animal samples, such as C. elegans, require homogenization before sample preparation to ensure efficient extraction of oxylipins. To achieve accurate quantification of PUFA metabolites in worm homogenates, it is essential to assess the solvent composition for optimal oxylipin recovery36. Considering this, two commonly used solvents, phosphate-buffered saline (PBS) as a water-based solvent and isopropanol as an organic solvent, were tested separately in conjunction with bead-based homogenization for preparing C. elegans homogenates. A comparison of the 100 extraction performance of two solvents revealed that PBS gave better extraction efficiency. The concentration of epoxy and dihydroxy metabolites was higher using PBS as a solvent comparing isopropanol (P ≤ 0.05, P ≤ 0.01, P ≤ 0.001, and P < 0.0001). Although a few metabolites demonstrated similar results for both PBS and isopropanol, with no statistically significant differences between them, such as 11,12 DiHETrE (Figure 2A). One potential reason for the lower efficiency of isopropanol as an organic solvent could be the formation of protein precipitates. Some proteins are bound with oxylipins, and their precipitation may interfere with the extraction process leading to lower concentrations of oxylipins. Thus, the choice of homogenization solvent plays a critical role in the success of oxylipin extraction and quantification. 2.3.2. Assessing oxylipin composition before and after hydrolysis in C. elegans sample To gain a comprehensive understanding of the function of PUFA metabolites37, we needed to analyze both free and esterified oxylipins because studies have revealed that several oxylipins also initiate their effects through their phospholipid-esterified product38. In addition, recent studies revealed that the cell membrane can act as a reservoir for free epoxy-FAs, and an external stimulus can trigger the release of these esterified epoxy-FAs, triggering subsequent biological effects39. 101 A) B) Figure 2.2. A) Optimization of solvent for C. elegans homogenization step, using two different solvents PBS and isopropanol. Oxylipin metabolites concentrations (pmol/g), mean ± SEM (n=3). Statistical differences between two different solvent groups were evaluated by multiple unpaired t-tests with *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P < 0.0001, non-significant is not shown). B) Oxylipin composition before and after hydrolysis in C. elegans samples. Control represents free fatty acids metabolites, whereas hydrolysis shows the total amount of oxylipin metabolites. Concentrations are mean ± SEM (n=3). Statistical differences between control and hydrolysis groups were evaluated by multiple unpaired t-tests with *P ≤ 0.05, **P ≤ 0.01, non-significant is not shown) To determine the total oxylipin levels, a well-established KOH/methanol method was used based on a published method 40 (the detailed hydrolysis method is described in Appendix S4.2). After the hydrolysis process, certain oxylipins exhibited significantly higher concentrations than the nonhydrolyzed samples. These included both CYP metabolites, namely 11,12- EpETE, and 14,15- EpEDE, and EH metabolites, specifically 9,10- DiHOME and 14,15- DiHETE. However, for metabolites such as 8-HETE and 11,12-DiHETE, the concentration difference between hydrolyzed samples and control samples was not statistically significant (Figure 2B). The significantly higher concentrations of certain oxylipins (like 11,12-EpETE and 14,15-EpEDE) in hydrolyzed samples suggest that these compounds exist both in free form and as esterified forms within cell membranes. The KOH/methanol hydrolysis effectively releases these bound forms, causing significant changes in concentration (p-value<0.05 and 0.01). In contrast, metabolites like 8-HETE and 11,12-DiHETE, which did not show significant changes, may predominantly exist in 102 free form. These findings highlight the complex dynamics of oxylipin storage and release in response to external stimuli. To determine whether the changes in relative oxylipin profile before and after hydrolysis were due to the breakdown of specific oxylipins during the hydrolysis step, a separate experiment was conducted with an identical hydrolysis protocol using Type I IS only. During this experiment, we noticed that the relative abundance of all different internal standards under hydrolysis conditions is the same, which shows that the relative stability of all oxylipins under the hydrolysis condition is the same (Appendix Table S6). This observation underscores the importance of evaluating oxylipin composition both before and after hydrolysis in C. elegans samples. The relative changes in concentration of different oxylipins before or after hydrolysis might be attributed to different substrate selectivity of enzymes that incorporate oxylipins or hydrolyze the corresponding esterified oxylipins. Further studies could provide insights into the specific factors contributing to these differences and improve our understanding of oxylipin metabolism in C. elegans. 2.3.3. Matrix effect in C. elegans and ionization efficiency Matrix effects (ME) pose significant challenges to quantitative LC-MS/MS analysis by compromising the precision, sensitivity, and accuracy of the methodology. It is essential to thoroughly assess the presence of matrix effects during method development to ensure trustworthy analytical results41, the matrix effect plays a significant role in oxylipin analysis in a variety of biological samples. Unlike other cases, it is not feasible to construct a calibration curve using the same matrix due to the presence of endogenous oxylipins in most biological samples. Therefore, we were compelled to rely on 75% EtOH as the solvent to create our calibration curve. To quantitatively evaluate matrix effects in the post-extraction addition method, we will compare the concentration of the analyte in a standard solution with that of a post-extraction spiked with the 103 analyte at an equal concentration. Our results indicated that compounds like PGB2-d4, 15(S)- HETE-d8, 9-HODE-d4, and 8,9-EpETrE-d11 showed a significant matrix effect of more than 20%. This change indicates that the worm matrix decreases ionization efficiency in ESI; the calibration curve concentration, spiked concentration, and calculated matrix effect for each compound are shown in Table 2.4. It is important to note that to calculate the concentration of deuterated IS, we normalized the peak area to the area of CUDA. The challenge with this approach was the potential matrix effect on CUDA itself. We compared the peak area of CUDA between the standard solution (average area equal 65379.33 and the RSD is 3.27% n=3) and the post-extraction spiked sample (average area equal 70609.68 and the RSD is 6.27% n=3). We found that the difference in area was 8 percent, indicating a minimal matrix effect for CUDA. An extracted ion chromatogram (XIC) of oxylipin in the calibration curve and day one worm samples is provided (Appendix Figure S2). Three common ways to tackle the ME are reducing the injection volume, diluting the sample before injection, and matrix match. The first two strategies are not applicable to oxylipins due to the extremely low concentration of these compounds in biological samples. The lack of worm samples without endogenous concentration of PUFA explains why the matrix match approach is not applicable to C. elegans. Altogether, we decided to normalize our result by the Type I internal standards; this approach is the most appropriate technique available to decrease the ME for quantitative analysis. 104 Table 2.4. The matrix effect (ME) in C. elegans extracts. ME can be quantitatively evaluated by comparing the concentration of the analyte in standard solution (A) to that of a post-extract spiked with the analyte at the same concentration (B) Calibration curve (A) Spiked (B) Matrix effect % (mean± SEM) Compound PGB2-d4 LTB4-d4 8,9-DiHETrE-d11 9-HODE-d4 15(S)-HETE-d8 5-HETE-d8 8,9-EpETrE-d11 15.6 10.0 15.7 16.7 20.5 42.0 42.1 19.7 9.8 18.8 20.6 23.3 56.9 56.5 -26.2± 5.6 2.4± 8.3 -19.9 ± 2.2 -23.5± 9.2 -13.6± 5.2 -35.6 ± 7.5 -34.2 ± 1.9 2.3.4. Preparation of age-synchronized worm Traditional techniques for preventing progeny overgrowth during C. elegans' lifespan and aging often employ 5-fluoro-2′-deoxyuridine (FUDR) for sterilization. While FUDR has been popularly employed to inhibit DNA and RNA synthesis, thereby eliminating progeny contamination, there are significant concerns tied to its use 42,43. For example, the effects of FUDR have been primarily studied in wild-type (N2) worms. Notably, the interactions between FUDR and specific mutant strains have raised questions about its efficacy and reliability 44,45. Studies like those by Aitalhaj et al. and the gas-1 strain underscore potential interferences, possibly skewing results and influencing longevity 46,47. Furthermore, the idea that the impact of FUDR sterilization is linear could be an oversimplification. External signals from the reproductive system have the potential to alter nematode lifespan 48, hinting at potential unexplored dimensions of FUDR effects. Additionally, our previous study raised concerns over the possible interference of FUDR with CYP-EH metabolites. Amid the aforementioned challenges with FUDR, our approach transitioned to a filtration method to achieve age synchronization49,50. This method was meticulously designed to draw a clear boundary between adult worms and their progeny, resulting in a striking separation efficiency of over 99% (Appendix Figure S5). Notably, while there was a minor loss of adults during filtration, 105 primarily because live animals occasionally adhered to the filter pores or passed through during the washing process, this loss was quantified to be less than 15% (Appendix Figure S5). The high efficacy of this method became even more pronounced as the worms aged. Initial stages required daily filtrations, but the filtration frequency diminished as the adult worms advanced in their life cycle and stopped producing progeny. After the initial week, the predominant aim shifted from rigorous progeny segregation to primarily transferring worms to plates with fresh food. 2.3.5. Analysis of C. elegans with and without EHI To determine whether our method can quantify the oxylipin changes in C. elegans, we quantified the oxylipin profiles in C. elegans with and without the treatment with an epoxide hydrolase inhibitor (EHI). Harris et al. showed that 12-(3-((3s,5s,7s)-adamantan-1-yl)ureido) dodecanoic acid (AUDA), blocks the conversion of epoxyoctadecenoic acids to dihydroxyoctadecenoic acid by inhibiting C. elegans epoxide hydrolases51,52. The oxylipin profile of AUDA-treated worm as compared with vehicle control revealed an increase in the levels of epoxy fatty acids metabolites, and a decrease in dihydroxy fatty acids metabolites, with the most pronounced change observed in (i) 12,13-DiHOME among LA metabolites, (ii) 8,9- DHED and 14,15-DHED for DGLA metabolites, (iii)11,12-EpETrE, and 14, 15-DiHETrE for AA metabolites, and (iv) 17,18-EpETE, 11,12-DiHETE, and 14,15-DiHETE for EPA metabolites (Figure 3). Administration of AUDA resulted in a significant increase in ω-3 epoxy fatty acids (p-value <0.05) and a significant decrease in ω-3 dihydroxy fatty acid (p-value<0.001). On the other hand, we only observed a decrease in ω-6 dihydroxy fatty acids (p-value <0.01), while the ω-6 epoxy fatty acids remained unchanged. It is worth mentioning that interpreting the effect of EH inhibitors could not be done correctly by solely tracking the epoxy fatty acids or dihydroxy fatty acids levels independently, as there is a physiological balance between these different 106 oxylipins (Figure 2.1). Therefore, to further explore the main effect of EH inhibition, the epoxy- to-dihydroxy ratio of different PUFA oxylipin metabolites, which is generally considered a marker of EH inhibition in vivo, was studied53,54. Our results show that wildtype worms treated with AUDA exhibit an increase in the epoxy-to-dihydroxy fatty acid ratio ranging from (2- 6) folds change, with the most significant increase related to the 14,15-(EpEDE/DHED) with 5 folds increase, and 11,12-(EpETE/DiHETE) with 6 folds increase in the ratio. We observed differen magnitudes of effects on various epoxy fatty acids -to-dihydroxy fatty acids ratios caused by the Figure 2.3. PUFA metabolites of C. elegans in the first day of adulthood, with and without epoxide hydrolase inhibitor treatment. The concentration of each compound (pmol/g) represents a replicate of independent populations of whole worm lysate that were 10 mg or greater, mean ± SEM (n=3). Statistical differences between two different groups were evaluated by multiple unpaired t-tests with *P ≤ 0.05, **P ≤ 0.01, and non-significant is not shown) 107 supplementation of AUDA. This observation is intriguing, as it implies that the connection between epoxide hydrolase activity and oxylipin levels could be more complex than initially thought. In particular, the presumption is that inhibiting epoxide hydrolase activity would stabilize or increase the endogenous level of all epoxy fatty acids and decrease the corresponding epoxide hydrolase products (dihydroxy fatty acids). (Figure 2.4A-D). This observation could be attributed to the regioselectivity of the C. elegans EHs, the presence of alternative pathways involved in the metabolism of epoxy and dihydroxy fatty acids metabolites, and the potential feedback regulation of other enzymes that play a part in this process. The relatively stable levels of epoxides, despite reduced conversion to diols, may suggest the presence of compensatory mechanisms. These mechanisms could involve increased metabolism of epoxides through alternative pathways or feedback inhibition of epoxide formation. One such pathway is the beta-oxidation of epoxyeicosatrienoic acids (EETs) to chain-shortened epoxy fatty acids55. According to the study by Feng et al., human skin fibroblasts primarily convert EETs into chain-shortened epoxy fatty acids rather than dihydroxyeicosatrienoic acids (DHETs)55. When the conversion to diols by epoxide hydrolase is inhibited by AUDA, the epoxides are likely diverted into this beta-oxidation pathway. This diversion results in the formation of shorter-chain epoxy fatty acids, possibly explaining the observed stability in epoxide levels despite the reduced conversion to diols. Besides, we also compared the level of CYP-EH metabolites related to ω-3 and ω-6 PUFAs (Figure 2.4 E, and F). The wildtype worms show a higher level of ω-3 epoxy fatty acids (600- 800 pmol/g) compared to ω-6 epoxy fatty acids (90-120 pmol/g), while the dihydroxy fatty acids level was almost similar for both ω-3 and ω-6-PUFAs (150-200 pmol/g). Moreover, we found a higher concentration of different EPA metabolites, and a lower concentration of the ALA 108 metabolites compared to the other PUFA metabolites. This pattern aligns with the levels of their respective parent PUFAs found in the worm52. Figure 2.4. (A-D) The ratio of epoxy to dihydroxy PUFAs metabolite of C. elegans in the first day of adulthood, with and without treatment of epoxide hydrolase inhibitor, AUDA. The concentration of each compound (pMol/g) represents a replicate of independent populations of whole worm lysate that were 10 mg or greater, mean ± SEM (n=3). Statistical differences between two different groups were evaluated by multiple unpaired t-tests with *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, and non-significant is not shown). (E -G). Represent the comparison of the total, ω-6, and ω-3 PUFA epoxy, dihydroxy metabolites and hydroxy metabolites in C. elegans on the first day of adulthood, with and without epoxide hydrolase inhibitor treatment. The concentration of each compound (pmol/g) represents a triplicate of independent populations of whole worm lysate that were 10 mg or greater, mean ± SEM (n=3). Statistical differences between two different groups were evaluated by multiple unpaired t-tests with *P ≤ 0.05, **P ≤ 0.01, and non-significant is not shown) 109 We also measured several hydroxy fatty acids in wild-type worms (Figure 2.4G). It should be noted that the hydroxy fatty acids s detected in this research are predominantly generated via three main pathways in mammals: (i) LOX enzymes, (ii) non-enzymatic oxidation, and (iii) CYP enzymes. Given the absence of LOX homologs in C. elegans, the production of hydroxy fatty acids in this organism is likely primarily facilitated by non-enzymatic oxidation, CYP pathways, or an undiscovered enzyme with monohydroxylation activity56-58. This observation highlights the unique metabolic processes in C. elegans and the potential for further study of these mechanisms. It is worth mentioning that like the majority of other animal species, C. elegans also possesses ∆6 and ∆5 desaturase enzymes that, in collaboration with fatty acid elongases, operate on substrates for the synthesis of 20-carbon PUFA like EPA and AA. Nevertheless, C. elegans lacks the specific enzymatic activity required for the elongation of fatty acids to generate 22-carbon PUFAs, such as DHA52, which explains the absence of DHA epoxy or dihydroxy metabolites in the findings of this study. We also found that AUDA treatment alters the hydroxy fatty acids levels in worms, with a statistically significant decrease in 5-HETE (p-value <0.01), 12-HETE, and 15-HETE (p- value<0.05). The inhibition of EH by AUDA stabilizes epoxy fatty acids and prevents their conversion to dihydroxy fatty acids. The accumulated epoxy fatty acid could inhibit enzymes upstream in the PUFA metabolism pathway, such as LOX, decreasing HETE synthesis. Additionally, the stabilization of epoxy fatty acid by AUDA might reduce the availability of PUFA substrates for conversion into HETEs, resulting in lower HETE levels due to substrate competition. To further investigate the effect of AUDA on HETE levels, conduct RNA-seq to measure the expression levels of genes involved in PUFA metabolism, including those encoding LOX and other enzymes. This could reveal whether AUDA treatment leads to transcriptional changes that 110 affect HETE production. Enzyme activity assays and substrate availability studies are the other two possible approaches. Whereas for the hydroxy metabolites like 19-HETE and 20-HEPE, the difference caused by AUDA treatment was not statistically significant. These findings emphasize the complex effects of AUDA treatment on hydroxy fatty acids levels and point to possible indirect influences on other enzymatic pathways or feedback regulation mechanisms. Further research is necessary to determine the potential roles of these changes and to elucidate the precise mechanisms underlying the observed alterations in hydroxy fatty acids levels. 3. Conclusion Our findings confirm the effectiveness of our method in quantifying oxylipin profiles in C. elegans and reveal key insights into the effects of epoxide hydrolase inhibition. Specifically, we observed a pronounced increase in ω-3 epoxy fatty acids and a stable level of ω-6 epoxy fatty acids, suggesting selective inhibition and differential regulatory pathways. Additionally, the varied effects on epoxy-to-dihydroxy fatty acid ratios highlight the complex interplay of enzymatic processes influencing PUFA metabolism. The method could be used to reveal further intricate relationships between enzymatic activity, oxylipin levels, and the roles of different metabolic pathways in C. elegans, pointing to the possibility of complex regulatory mechanisms at play. By providing a comprehensive picture of alterations in oxylipin profiles after AUDA treatment, our work lays the foundation for future investigations into the underlying mechanisms and potential therapeutic applications of epoxide hydrolase inhibitors. This study aimed to develop a reliable and accurate LC-MS/MS method for quantifying the oxylipin profile in the C. elegans animal model. The developed method can be employed in future studies to investigate the effects of various diseases, treatments, or genetic manipulations on the 111 oxylipin profile and to assess potential therapeutic interventions using C. elegans as an animal model. Ultimately, the knowledge gained from these studies may contribute to a better understanding of the role of oxylipins in health and disease and may potentially lead to the development of novel therapeutic strategies. By further refining the method and expanding the range of oxylipins analyzed, researchers can continue to build upon the foundation established in this study and enhance our understanding of the complex interplay between oxylipin metabolism and various biological processes. APPENDIX 1 provides supporting information, including experimental methods and materials, supplemental figures and tables, and the synthesis and characterization of DGLA metabolites. 112 BIBLIOGRAPHY (1) Hou, Y.; Dan, X.; Babbar, M.; Wei, Y.; Hasselbalch, S. G.; Croteau, D. L.; Bohr, V. A. Ageing as a Risk Factor for Neurodegenerative Disease. Nat. Rev. Neurol. 2019, 15 (10), 565–581. (2) Alzheimer’s Association. 2019 Alzheimer’s Disease Facts and Figures. Alzheimers. Dement. 2019, 15 (3), 321–387. (3) Hamsanathan, S.; Gurkar, A. U. Lipids as Regulators of Cellular Senescence. Front. Physiol. 2022, 13, 796850. (4) Johnson, A. A.; Stolzing, A. The Role of Lipid Metabolism in Aging, Lifespan Regulation, and Age-Related Disease. Aging Cell 2019, 18 (6), e13048. (5) Vierck, J. L.; Dodson, M. V. Interpretation of Cell Culture Phenomena. Methods Cell Sci. 2000, 22 (1), 79–81. (6) Zhang, J.; Yang, J.; Duval, C.; Edin, M. L.; Williams, A.; Lei, L.; Tu, M.; Pourmand, E.; Song, R.; Graves, J. P.; DeGraff, L. M.; Wong, J. J.-L.; Wang, Y.; Sun, Q.; Sanidad, K. Z.; Wong, S.; Han, Y.; Zhang, Z.; Lee, K. S. S.; Park, Y.; Xiao, H.; Liu, Z.; Decker, E. A.; Cui, W.; Zeldin, D. C.; Zhang, G. CYP Eicosanoid Pathway Mediates Colon Cancer-Promoting Effects of Dietary Linoleic Acid. FASEB J. 2023, 37 (7), e23009. (7) Favor, O. K.; Chauhan, P. S.; Pourmand, E.; Edwards, A. M.; Wagner, J. G.; Lewandowski, R. P.; Heine, L. K.; Harkema, J. R.; Lee, K. S. S.; Pestka, J. J. Lipidome Modulation by Dietary Omega-3 Polyunsaturated Fatty Acid Supplementation or Selective Soluble Epoxide Hydrolase Inhibition Suppresses Rough LPS-Accelerated Glomerulonephritis in Lupus-Prone Mice. Front. Immunol. 2023, 14, 1124910. (8) Cambiaggi, L.; Chakravarty, A.; Noureddine, N.; Hersberger, M. The Role of α-Linolenic Acid and Its Oxylipins in Human Cardiovascular Diseases. Int. J. Mol. Sci. 2023, 24 (7), 6110. (9) Eriksdotter, M.; Vedin, I.; Falahati, F.; Freund-Levi, Y.; Hjorth, E.; Faxen-Irving, G.; Wahlund, L.-O.; Schultzberg, M.; Basun, H.; Cederholm, T.; Palmblad, J. Plasma Fatty Acid Profiles in Relation to Cognition and Gender in Alzheimer’s Disease Patients during Oral Omega-3 Fatty Acid Supplementation: The OmegAD Study. J. Alzheimers. Dis. 2015, 48 (3), 805–812. (10) Jernerén, F.; Cederholm, T.; Refsum, H.; Smith, A. D.; Turner, C.; Palmblad, J.; Eriksdotter, M.; Hjorth, E.; Faxen-Irving, G.; Wahlund, L.-O.; Schultzberg, M.; Basun, H.; Freund-Levi, Y. Homocysteine Status Modifies the Treatment Effect of Omega-3 Fatty Acids on Cognition in a Randomized Clinical Trial in Mild to Moderate Alzheimer’s Disease: The OmegAD Study. J. Alzheimers. Dis. 2019, 69 (1), 189–197. (11) Sarparast, M.; Pourmand, E.; Hinman, J.; Vonarx, D.; Reason, T.; Zhang, F.; Paithankar, S.; Chen, B.; Borhan, B.; Watts, J. L.; Alan, J.; Lee, K. S. S. Dihydroxy-Metabolites of Dihomo-γ- Linolenic Acid Drive Ferroptosis-Mediated Neurodegeneration. ACS Cent. Sci. 2023, 9 (5), 870– 882. 113 (12) Camunas-Alberca, S. M.; Moran-Garrido, M.; Sáiz, J.; Villaseñor, A.; Taha, A. Y.; Barbas, C. The Role of Oxylipins and Their Validation as Biomarkers in the Clinical Context. Trends Analyt. Chem. 2023, 164 (117065), 117065. (13) Chu, J.; Giannopoulos, P. F.; Ceballos-Diaz, C.; Golde, T. E.; Pratico, D. Adeno-Associated Virus-Mediated Brain Delivery of 5-Lipoxygenase Modulates the AD-like Phenotype of APP Mice. Mol. Neurodegener. 2012, 7 (1), 1. (14) Yermakova, A. V.; Rollins, J.; Callahan, L. M.; Rogers, J.; O’Banion, M. K. Cyclooxygenase- 1 in Human Alzheimer and Control Brain: Quantitative Analysis of Expression by Microglia and CA3 Hippocampal Neurons. J. Neuropathol. Exp. Neurol. 1999, 58 (11), 1135–1146. (15) Ghosh, A.; Comerota, M. M.; Wan, D.; Chen, F.; Propson, N. E.; Hwang, S. H.; Hammock, B. D.; Zheng, H. An Epoxide Hydrolase Inhibitor Reduces Neuroinflammation in a Mouse Model of Alzheimer’s Disease. Sci. Transl. Med. 2020, 12 (573), eabb1206. (16) Marowsky, A.; Burgener, J.; Falck, J. R.; Fritschy, J.-M.; Arand, M. Distribution of Soluble and Microsomal Epoxide Hydrolase in the Mouse Brain and Its Contribution to Cerebral Epoxyeicosatrienoic Acid Metabolism. Neuroscience 2009, 163 (2), 646–661. (17) Hung, T.-H.; Shyue, S.-K.; Wu, C.-H.; Chen, C.-C.; Lin, C.-C.; Chang, C.-F.; Chen, S.-F. Deletion or Inhibition of Soluble Epoxide Hydrolase Protects against Brain Damage and Reduces Microglia-Mediated Neuroinflammation in Traumatic Brain Injury. Oncotarget 2017, 8 (61), 103236–103260. (18) Griñán-Ferré, C.; Codony, S.; Pujol, E.; Yang, J.; Leiva, R.; Escolano, C.; Puigoriol-Illamola, D.; Companys-Alemany, J.; Corpas, R.; Sanfeliu, C.; Pérez, B.; Loza, M. I.; Brea, J.; Morisseau, C.; Hammock, B. D.; Vázquez, S.; Pallàs, M.; Galdeano, C. Pharmacological Inhibition of Soluble Epoxide Hydrolase as a New Therapy for Alzheimer’s Disease. Neurotherapeutics 2020, 17 (4), 1825–1835. (19) Lee, H.-T.; Lee, K.-I.; Chen, C.-H.; Lee, T.-S. Genetic Deletion of Soluble Epoxide Hydrolase Delays the Progression of Alzheimer’s Disease. J. Neuroinflammation 2019, 16 (1), 267. (20) Gan, L.; Cookson, M. R.; Petrucelli, L.; La Spada, A. R. Converging Pathways in Neurodegeneration, from Genetics to Mechanisms. Nat. Neurosci. 2018, 21 (10), 1300–1309. (21) Colman, R. J. Non-Human Primates as a Model for Aging. Biochim. Biophys. Acta Mol. Basis Dis. 2018, 1864 (9), 2733–2741. (22) Mitchell, S. J.; Scheibye-Knudsen, M.; Longo, D. L.; de Cabo, R. Animal Models of Aging Research: Implications for Human Aging and Age-Related Diseases. Annu. Rev. Anim. Biosci. 2015, 3 (1), 283–303. (23) Johnson, I. P. Response: Commentary: Age-Related Neurodegenerative Disease Research Needs Aging Models. Front. Aging Neurosci. 2016, 8, 44. (24) Kell, D. B. Metabolomic Biomarkers: Search, Discovery and Validation. Expert Rev. Mol. Diagn. 2007, 7 (4), 329–333. 114 (25) Schuchardt, J. P.; Schmidt, S.; Kressel, G.; Dong, H.; Willenberg, I.; Hammock, B. D.; Hahn, A.; Schebb, N. H. Comparison of Free Serum Oxylipin Concentrations in Hyper- vs. Normolipidemic Men. Prostaglandins Leukot. Essent. Fatty Acids 2013, 89 (1), 19–29. (26) Watts, J. L.; Browse, J. Genetic Dissection of Polyunsaturated Fatty Acid Synthesis in Caenorhabditis Elegans. Proc. Natl. Acad. Sci. U. S. A. 2002, 99 (9), 5854–5859. (27) Thomas, S. N.; French, D.; Jannetto, P. J.; Rappold, B. A.; Clarke, W. A. Liquid Chromatography-Tandem Mass Spectrometry for Clinical Diagnostics. Nat. Rev. Methods Primers 2022, 2 (1), 96. (28) Yang, J.; Schmelzer, K.; Georgi, K.; Hammock, B. D. Quantitative Profiling Method for Oxylipin Metabolome by Liquid Chromatography Electrospray Ionization Tandem Mass Spectrometry. Anal. Chem. 2009, 81 (19), 8085–8093 (29) Misheva, M.; Kotzamanis, K.; Davies, L. C.; Tyrrell, V. J.; Rodrigues, P. R. S.; Benavides, G. A.; Hinz, C.; Murphy, R. C.; Kennedy, P.; Taylor, P. R.; Rosas, M.; Jones, S. A.; McLaren, J. E.; Deshpande, S.; Andrews, R.; Schebb, N. H.; Czubala, M. A.; Gurney, M.; Aldrovandi, M.; Meckelmann, S. W.; Ghazal, P.; Darley-Usmar, V.; White, D. A.; O’Donnell, V. B. Oxylipin Metabolism Is Controlled by Mitochondrial β-Oxidation during Bacterial Inflammation. Nat. Commun. 2022, 13 (1), 139. (30) Simultaneous Quantitation of Lipid Biomarkers for Inflammatory Bowel Disease Using LC- MS/MS Yashpal S. Chhonker 1 and Daryl J. Murry 1,4,* , Shrey Kanvinde 2. 2. (31) Fulton, D.; Falck, J. R.; McGiff, J. C.; Carroll, M. A.; Quilley, J. A Method for the Determination of 5,6-EET Using the Lactone as an Intermediate in the Formation of the Diol. J. Lipid Res. 1998, 39 (8), 1713–1721. (32) Vrablik, T. L.; Watts, J. L. Polyunsaturated Fatty Acid Derived Signaling in Reproduction and Development: Insights from Caenorhabditis Elegans and Drosophila Melanogaster. Mol. Reprod. Dev. 2013, 80 (4), 244–259. (33) Mainka, M.; Dalle, C.; Pétéra, M.; Dalloux-Chioccioli, J.; Kampschulte, N.; Ostermann, A. I.; Rothe, M.; Bertrand-Michel, J.; Newman, J. W.; Gladine, C.; Schebb, N. H. Harmonized Procedures Lead to Comparable Quantification of Total Oxylipins across Laboratories. J. Lipid Res. 2020, 61 (11), 1424–1436. (34) Pedersen, T. L.; Gray, I. J.; Newman, J. W. Plasma and Serum Oxylipin, Endocannabinoid, Bile Acid, Steroid, Fatty Acid and Nonsteroidal Anti-Inflammatory Drug Quantification in a 96- Well Plate Format. Anal. Chim. Acta 2021, 1143, 189–200. (35) Fu, X.; Yin, H.-H.; Wu, M.-J.; He, X.; Jiang, Q.; Zhang, L.-T.; Liu, J.-Y. High Sensitivity and Wide Linearity LC-MS/MS Method for Oxylipin Quantification in Multiple Biological Samples. J. Lipid Res. 2022, 63 (12), 100302. (36) Höring, M.; Krautbauer, S.; Hiltl, L.; Babl, V.; Sigruener, A.; Burkhardt, R.; Liebisch, G. Accurate Lipid Quantification of Tissue Homogenates Requires Suitable Sample Concentration, 115 Solvent Composition, and Homogenization Procedure-A Case Study in Murine Liver. Metabolites 2021, 11 (6), 365. (37) Salimon, J.; Abdullah, B. M.; Salih, N. Hydrolysis Optimization and Characterization Study of Preparing Fatty Acids from Jatropha Curcas Seed Oil. Chem. Cent. J. 2011, 5 (1), 67. (38) Annevelink, C. E.; Walker, R. E.; Shearer, G. C. Esterified Oxylipins: Do They Matter? Metabolites 2022, 12 (11), 1007. (39) Ray, S.; Kassan, A.; Busija, A. R.; Rangamani, P.; Patel, H. H. The Plasma Membrane as a Capacitor for Energy and Metabolism. Am. J. Physiol. Cell Physiol. 2016, 310 (3), C181-92. (40) Rund, K. M.; Heylmann, D.; Seiwert, N.; Wecklein, S.; Oger, C.; Galano, J.-M.; Durand, T.; Chen, R.; Gueler, F.; Fahrer, J.; Bornhorst, J.; Schebb, N. H. Formation of Trans-Epoxy Fatty Acids Correlates with Formation of Isoprostanes and Could Serve as Biomarker of Oxidative Stress. Prostaglandins Other Lipid Mediat. 2019, 144 (106334), 106334. (41) Matuszewski, B.; Constanzer, M.; Chavez-Eng, C. Strategies for the Assessment of Matrix Effect in Quantitative Bioanalytical Methods Based on HPLC−MS/MS. Anal. Chem 2003, 75 (13), 3019–3030. (42) Gandhi, S.; Santelli, J.; Mitchell, D. H.; Wesley Stiles, J.; Rao Sanadi, D. A Simple Method for Maintaining Large, Aging Populations of Caenorhabditis Elegans. Mech. Ageing Dev. 1980, 12 (2), 137–150. (43) Mitchell, D. H.; Stiles, J. W.; Santelli, J.; Sanadi, D. R. Synchronous Growth and Aging of Caenorhabditis Elegans in the Presence of Fluorodeoxyuridine. J. Gerontol. 1979, 34 (1), 28– (44) Hosono, R.; Mitsui, Y.; Sato, Y.; Aizawa, S.; Miwa, J. Life Span of the Wild and Mutant Nematode Caenorhabditis Elegans. Effects of Sex, Sterilization, and Temperature. Exp. Gerontol. 1982, 17 (2), 163–172. (45) Hosono, R. Sterilization and Growth Inhibition of Caenorhabditis Elegans by 5- Fluorodeoxyuridine. Exp. Gerontol. 1978, 13 (5), 369–374. (46) Aitlhadj, L.; Stürzenbaum, S. R. The Use of FUdR Can Cause Prolonged Longevity in Mutant Nematodes. Mech. Ageing Dev. 2010, 131 (5), 364–365. (47) Van Raamsdonk, J. M.; Hekimi, S. FUdR Causes a Twofold Increase in the Lifespan of the Mitochondrial Mutant Gas-1. Mech. Ageing Dev. 2011, 132 (10), 519–521. (48) Hsin, H.; Kenyon, C. Signals from the Reproductive System Regulate the Lifespan of C. Elegans. Nature 1999, 399 (6734), 362–366. (49) Hunter, S.; Maulik, M.; Scerbak, C.; Vayndorf, E.; Taylor, B. E. Caenorhabditis Sieve: A Low-Tech Instrument and Methodology for Sorting Small Multicellular Organisms. J. Vis. Exp. 2018, No. 137. https://doi.org/10.3791/58014. 116 (50) Cailloce, J.; Husson, F.; Zablocki, A.; Galy, V.; Merlet, J. Fast and Easy Method to Culture and Obtain Large Populations of Male Nematodes. MethodsX 2023, 11 (102293), 102293. (51) Harris, T. R.; Aronov, P. A.; Jones, P. D.; Tanaka, H.; Arand, M.; Hammock, B. D. Identification of Two Epoxide Hydrolases in Caenorhabditis Elegans That Metabolize Mammalian Lipid Signaling Molecules. Arch. Biochem. Biophys. 2008, 472 (2), 139–149. (52) Watts, J. L. Using Caenorhabditis Elegans to Uncover Conserved Functions of Omega-3 and Omega-6 Fatty Acids. J. Clin. Med. Res. 2016, 5 (2). https://doi.org/10.3390/jcm5020019. (53) Inceoglu, B.; Wagner, K.; Schebb, N. H.; Morisseau, C.; Jinks, S. L.; Ulu, A.; Hegedus, C.; Rose, T.; Brosnan, R.; Hammock, B. D. Analgesia Mediated by Soluble Epoxide Hydrolase Inhibitors Is Dependent on CAMP. Proc. Natl. Acad. Sci. U. S. A. 2011, 108 (12), 5093–5097. (54) Inceoglu, B.; Zolkowska, D.; Yoo, H. J.; Wagner, K. M.; Yang, J.; Hackett, E.; Hwang, S. H.; Lee, K. S. S.; Rogawski, M. A.; Morisseau, C.; Hammock, B. D. Epoxy Fatty Acids and Inhibition of the Soluble Epoxide Hydrolase Selectively Modulate GABA Mediated Neurotransmission to Delay Onset of Seizures. PLoS One 2013, 8 (12), e80922. (55) Fang, X.; Kaduce, T. L.; VanRollins, M.; Weintraub, N. L.; Spector, A. A. Conversion of Epoxyeicosatrienoic Acids (EETs) to Chain-Shortened Epoxy Fatty Acids by Human Skin Fibroblasts. J. Lipid Res. 2000, 41 (1), 66–74. (56) Sarparast, M.; Dattmore, D.; Alan, J.; Lee, K. S. S. Cytochrome P450 Metabolism of Polyunsaturated Fatty Acids and Neurodegeneration. Nutrients 2020, 12 (11), 3523. (57) Bazinet, R. P.; Layé, S. Polyunsaturated Fatty Acids and Their Metabolites in Brain Function and Disease. Nat. Rev. Neurosci. 2014, 15 (12), 771–785. (58) Czapski, G. A.; Czubowicz, K.; Strosznajder, J. B.; Strosznajder, R. P. The Lipoxygenases: Their Regulation and Implication in Alzheimer’s Disease. Neurochem. Res. 2016, 41 (1–2), 243– 257. 117 CHAPTER 3. THE EFFECTS OF SYSTEMIC LUPUS ERYTHEMATOSUS (SLE) ON UNSATURATED LIPIDS AND THEIR OXIDATIVE METABOLISM IN THE WESTERN POPULATION 118 ABSTRACT Studies show that supplementing omega-3 fatty acids decreases the severity of SLE symptoms. However, how these dietary PUFAs affect SLE pathogenesis remains to be seen. The oxidized PUFA metabolites called oxylipins are lipid mediators that play a critical role in inflammation. A recent epidemiology study in China identified 26 oxylipins whose serum levels were significantly changed in the SLE patients compared with the healthy controls, suggesting that oxylipins may play a role in SLE pathogenesis. Thus, in this study, we further investigated how 1) severity of SLE, 2) age, 3) race, and 4) sex affect the endogenous oxylipin metabolism in SLE individuals from the Western population. This comprehensive study collected serum samples and health questionnaire data from 372 SLE patients and 159 healthy controls, including female, male, White, and Black populations. We performed oxylipin analysis on these serum samples using high- performance liquid chromatography coupled with an electrospray tandem mass spectrometry (HPLC-ESI-MS\MS) and serum fatty acid level using gas chromatography coupled with mass spectrometry (GC-MS). Our analysis revealed that key oxylipin serum levels of epoxide hydrolase metabolites, including fatty acid diols 14,15-DiHETrE and 17,18-DiHETE are significantly downregulated in subjects with SLE as compared with healthy subjects. Moreover, our machine learning analysis also revealed that cytochrome P450-epoxide hydrolase-derived eicosanoids and current steroid use are highly important in predicting whether subjects have SLE, indicating a strong link with disease activity. Through advanced statistical modeling and machine learning, our study revealed a potential metabolic pathway, CYP-EH metabolism, impacted by SLE pathogenesis and identified potential biomarkers for SLE. 119 1. Introduction Systemic lupus erythematosus (SLE) is a chronic autoimmune condition where the body's immune system mistakenly attacks its tissues, leading to the rampant generation of autoantibodies. This aberrant immune response results in sustained inflammation throughout the body and can cause various complications across numerous organ systems1. SLE is a heterogenic disease affecting multiple organs and a variable disease course. The destructive impact on end-organs and the subsequent mortality among SLE patients primarily result from immune system alterations and prolonged inflammatory effects2,3. SLE can present with a range of pathologies, including nephritis, blood abnormalities, serositis, arthritis, cutaneous rashes, psychosis, and seizures, further complicating diagnosis and prognosis. Autoantibodies, such as antinuclear antibodies (ANA), are useful for SLE detection, disease activity classification, and prognosis monitoring. Moreover, a series of pro-inflammatory or immunomodulatory cytokines, such as IL-1, IL-6, IL- 10, TNF, IFN-I, and BAFF, have shown substantial upregulation in SLE and are closely associated with autoantibody deposition in various organs4,5. However, new biomarkers are needed for several reasons, such as difficulty identifying flares, inability to predict organ involvement patterns, and lack of biomarkers to assess response to experimental therapies6. Onset varies, with some experiencing juvenile lupus before the age of 167. Besides, a notable gender bias is observed in SLE, with a typical female-to-male ratio of 10:1 within the SLE population8. The underlying molecular mechanism of such significant risk associated with sex remains a subject of investigation. It was hypothesized that the sex hormones affect B-cell responses and autoantibody production9,10. Despite decades of research, only one intervention was recently approved by the FDA. In addition, the most common drug, prednisone, can only treat symptoms with a lot of severe 120 side effects. Therefore, there is an unmet medical need to identify new targets for developing novel therapeutics to treat SLE. Recent studies demonstrated that dietary polyunsaturated fatty acid (PUFA) could be critical in SLE pathogenesis11,12. In general, omega-3 PUFAs, abundant in fatty fish, nuts, seeds, and oils, have been shown to possess anti-inflammatory and immunomodulatory properties. Conversely, omega-6 PUFAs, including linoleic acid (LA) and arachidonic acid (AA), are generally considered pro-inflammatory11,12. Oxylipins are the PUFA downstream metabolites, and as lipid mediators, they can have anti-inflammatory or pro-inflammatory properties, depending on their specific structures13. Since the initial clinical trial in 1989, seven significant clinical studies have explored the effects of omega-3 PUFAs on SLE. Except for one14, these studies have consistently reported positive outcomes, such as enhanced endothelial function, reduced disease activity, and lower levels of inflammatory markers after administering omega-3 PUFAs to SLE individuals. The studies highlight that interventions lasting over 12 weeks were crucial for observing these benefits. Daily supplementation with moderate doses of EPA (162 mg) and DHA (144 mg) led to the remission of SLE in ten individuals. Additionally, both EPA and DHA were found to inhibit T- cell proliferation and reduce the production of inflammatory cytokines such as IL-1, IL-2, and TNF-α in laboratory settings and live subjects14–20. While the mechanism of how omega-3 PUFA affects SLE pathogenesis remains unclear, recent studies have demonstrated that oxylipins could partly be responsible for the actions of omega-3 PUFA on SLE pathogenesis. Oxylipins are produced through enzymatic or non-enzymatic oxidations of PUFAs21. Oxylipins predominantly originate from the metabolic pathways involving three groups of enzymes, namely lipoxygenase (LOX), cyclooxygenase (COX), and cytochrome P450 family proteins (CYP)22. Disruptions in oxylipin metabolism are linked to several health issues, such as cancer, cardiovascular diseases, 121 aging, and rheumatoid arthritis 23,24. Several drugs targeting oxylipin metabolic enzymes have been developed24. Oxylipins are crucial in managing physiological and pathological conditions, particularly in controlling inflammation and immune functions, and are highly relevant to SLE25– 27. Several studies have shown that specific oxylipins such as epoxyeicosatrienoic acid (EET), thromboxane A2 (TXA2), prostaglandin E2 (PGE2), and PGD2 are markedly changed and play an essential role in the development of SLE. This indicates that these oxylipins could serve as promising biomarkers for diagnosing SLE and aiding in its clinical management28–30. In this study, we aim to explore the potential impact of SLE pathogenesis on dietary PUFA metabolism, particularly the balance between omega-6 and omega-3 fatty acids. Our research delves into how 1) PUFA metabolite levels, 2) SLE severity, 3) age, 4) race, 5) sex, and 6) supplementation affect oxylipin metabolism in the Western SLE population. Serum samples and questionnaires from 530 individuals were collected, spanning sex, age, and ethnicity. Oxylipin analysis was done using established oxylipin analysis by HPLC-MS/MS, while serum fatty acids were analyzed using GC-MS. Statistical analyses were conducted to discern significant variations in metabolite concentrations across diverse demographic groups. Additionally, various machine learning algorithms and models were employed to categorize and interpret the dataset and to understand the relative importance of each metabolite, their ratio, and demographics in the presence and severity of the SLE. Our result shows that the cytochrome P450 epoxide hydrolase (CYP-EH) metabolism of PUFA is impacted in SLE, and several metabolites from this specific metabolic pathway could be biomarkers for identifying subjects with SLE. Our results may help to clarify the molecular mechanisms and enzymatic pathways by which dietary fatty acids modulate the progression of lupus pathogenesis and help identify potential novel therapeutic targets for treating or preventing SLE in the future. 122 2. Experimental 2.1 Data collection This research utilizes the Michigan Lupus Epidemiology and Surveillance (MILES) Cohort, which consists of adult lupus patients recruited from the MILES Surveillance Registry based in southeastern Michigan31. Data from the baseline cohort visit were utilized for this study. Sociodemographic data was gathered through structured interviews and questionnaires modeled similarly to those used in the NHANES study32. Participants self-reported their race and ethnicity. Patient-reported outcomes were collected and evaluated based on the specific guidelines of each instrument’s instruction. All participants in the cohort gave their written, informed consent. The Systemic Lupus Activity Questionnaire (SLAQ)33, validated for epidemiological research, assessed lupus disease activity over the previous 30 days. The SLAQ includes 24 items with scores ranging from 0 to 47; higher scores represent more severe disease activity levels. 2.2. Oxylipin quantification 2.2.1. Sample preparation For sample preparation, we utilized Waters Oasis-HLB cartridges (Part No. WAT094226, Lot No. 176A30323A; 60 mg sorbent, 3 cc) for solid phase extraction (SPE). To prepare the cartridge for extraction, we followed these steps: first, we washed it with ethyl acetate (2 mL), then with methanol (2× 2 mL), and finally with a mixture of water/methanol (95:5 v/v) containing 0.1% acetic acid (2× 2 mL). Next, we loaded 200 µl of thawed plasma samples onto these HLB cartridges, we spiked samples with 10 μL of a 0.4A Type I internal standards solution (A is defined in Appendix Table 1), and we also added 10 μL of an antioxidant cocktail including triphenylphosphine (TPP) (0.2 mg/ml), butylated hydroxytoluene (BHT) (0.2 mg/ml), and ethylenediaminetetraacetic (EDTA) (1 mg/ml). After loading the samples, we washed the 123 cartridges with a solution of water/methanol (95:5 v/v) containing 0.1% acetic acid (1.5 mL). Following the wash, we subjected the HLB cartridges to low vacuum for roughly 20 minutes to eliminate any remaining water and other solvent residues. Subsequently, we eluted the compounds from the cartridges with methanol (0.5 mL), followed by ethyl acetate (1 mL), into 2 mL Eppendorf tubes containing a glycerol/Methanol mixture (30% v/v, 6 μL) as a trapping solution. The eluents were then concentrated using a speed vacuum concentrator. The resulting residues were reconstituted in ethanol/water (75% v/v, 100 μL) containing the internal standard 12- [[(cyclohexylamino)carbonyl]amino]-dodecanoic acid (CUDA) (10 nM). To ensure thorough mixing, we vortexed the samples for 5 minutes and then subjected them to centrifugal filtration (0.45 μm Millipore). Finally, the filtrates were transferred to amber glass autosampler vials (1.5 mL capacity) with silanized inserts (200 µL capacity), purged with argon gas, capped, and stored at -80°C until HPLC-MS/MS analysis. 2.2.2. LC-MS/MS analysis LC conditions were chosen to separate the eicosanoids to ensure adequate signal intensity and desired peak shape. We used an XBridge BEH C18 2.1x150 mm HPLC column (Waters, Milford, MA; Serial Number: 01723829118314) with mobile phase A consisting of 0.1% acetic acid in water, and mobile phase B was a mixture of acetonitrile and methanol (84:16) containing 0.1% acetic acid at a flow rate of 250 μL/min and a run time of 20 minutes. The gradient details are available in Appendix Table S3. Eicosanoids with lower hydrophobicity, such as PGs, TXS, and LTs, eluted earlier in the process, while more hydrophobic eicosanoids, including HETEs, HEPEs, AA, DHA, and EPA, eluted later. We maintained the autosampler (Waters ACQUITY FTN) at a temperature of 10°C. The column was connected to a Xevo TQ-XS tandem mass spectrometer 124 (Waters Corp., Milford, MA), equipped with a Waters Acquity I-class pump and Waters Acquity column manager. Electrospray was used as the ionization source for negative MRM mode, serving as an interface between HPLC and tandem mass spectrometry. We infused the mass spectrometer for each analyte standard, optimizing MRM transitions and source parameters to achieve the best selectivity and sensitivity. We strategically distributed the compounds into different functions based on retention time to increase dwell time, thereby lowering detection limits, and enhancing sensitivity. We retained the most abundant transitions for each standard throughout this optimization process. We selected more specific or sensitive transitions when distinguishing between closely related isomers or achieving better detection limits was required. These instrumental optimizations addressed two critical challenges in oxylipin analysis: the low endogenous concentration in biological samples and the structural similarity among compounds. A mixture of PUFA's metabolites with known concentrations creates an 8-concentration calibration standard in ethanol (EtOH) Appendix Table S2. This calibration standard contains both Type I and Type II internal standards. Calibration standards are stored in amber vials sealed under argon gas at -80°C to maintain stability. Calibration curves are created by plotting the response of the analyte, which is the ratio of the analyte's area to its internal standard's area against the concentration. Using 1/x weighting factors in the regression indicates that the lower concentrations have more weight in the curve-fitting process. Regression analysis is performed on the calibration curves to determine their goodness of fit. An R^2 value of 0.998 or greater is achieved for each analyte. An R^2 value of 1.0 would indicate a perfect fit, so values close to or greater than 0.998 suggest high precision and reliability in the calibration process. This approach ensures accurate and precise quantification of PUFA's 125 metabolites in the samples using internal standards and well-prepared calibration curves in LC- MS/MS analysis. Deuterated compounds were added to the samples as Type I Internal Standards before the SPE. They are used to calculate the extraction recovery of prostaglandins, diols, epoxides, and other oxylipins. The choice of deuterated compounds is because they have physical and chemical properties like the targeted metabolites, making them suitable for quantification. Type II Internal Standard was a synthetic compound called CUDA, a fatty acid metabolite mimic. It is added at the last step before injection into LC-MS/MS. Its purpose is to normalize changes in volume and other instrumental variations. The overlay of extracted ion chromatograms of different functions of plasma samples was obtained from a 56-year-old Caucasian female, a patient diagnosed with SLE, which is provided in Appendix Figure 1. The LC-MS/MS analysis was conducted at the Michigan State University Mass Spectrometry and Metabolomics Facility. 2.3. PUFAs quantification 2.3.1. Serum fatty acid quantification procedure Blood samples were obtained through standard venipuncture using coagulant-free red-top vacutainers. Serum specimens were processed and aliquoted following standardized protocols and stored at -70°C. The identification and quantification of fatty acid concentrations were performed using GC-MS34. 2.3.2. Reagents and Standards Analytical-grade reagents from Sigma-Aldrich (St. Louis, MO, USA) were employed. Stearic acid-d35 was the internal standard (Sigma-Aldrich; St. Louis, MO). Methyl ester standards, including palmitelaidic, mead, docosatetraenoic (DTA), n-6 docosapentaenoic (DPA n-6), and n- 3 docosapentaenoic (DPA n-3) acids, were acquired from Cayman Chemical (Ann Arbor, MI, 126 USA). Supelco 37 Component FAME Mix (Sigma-Aldrich; St. Louis, MO) was utilized to create standard curves for other fatty acid methyl esters (FAME). 2.3.3. Fatty acid extraction and derivatization The extraction and methylation of fatty acids were carried out in a single step, following the method described by Lepage and Roy 35 with modifications by Masood et al34. In brief, 2 mL of a 1.8:0.2 (v/v) methanol: acetyl chloride solution with 0.01% (w/v) butylated hydroxytoluene, and internal standard was added to 100 µL of serum; the tubes were sealed to avoid methanol evaporation. The mixture was heated for 1 hour at 100°C, cooled to room temperature, and neutralized with 2 mL of 5% (w/v) sodium bicarbonate solution. FAMEs were extracted twice with 2 mL hexane, dried under nitrogen, resuspended in isooctane, and transferred to GC vials. Samples were stored at -20°C until GC-MS analysis. 2.3.4. GC-MS analysis Quantification of FAMEs was performed using a Perkin Elmer 680/600S GC-MS with an Agilent Technologies DB-23, 30-m column. GC temperature parameters were set as follows: initial temperature at 100°C for 0.5 minutes; ramped 8.0 degrees per minute to 200°C, then 2.5 degrees per minute to 220°C, and finally 10 degrees per minute to 240°C, where it was held for 2 minutes. Seven-point calibration curves, generated with purchased FAME standards, were utilized for quantification through selective-ion monitoring. Data were analyzed using TargetLynx v4.0 (Waters Corporation; Milford, MA, USA). FAME values were normalized per mL of serum. The fatty acids analysis was done by Dr. Fenton’s lab, Department of Food Science and Human Nutrition Michigan State University. 127 2.4. Machine learning (ML) The sample size of the data set is 530. After eliminating the cases with missing values, the sample size is reduced to 477. However, 106 predictors include oxylipins, fatty acids, demographics, medication, and supplementations. Some commonly used statistical models, like regression, logistic regression, etc., could result in unstable estimates because when we have many predictors, the standard error of the estimate will be very large. We, therefore, used ML methods to identify the most relevant predictors. The results are more generalizable because we use cross-validation and splitting data into learning and test data sets. In machine learning, the training set is utilized to train and fit the model, enabling it to learn and identify patterns between inputs and expected outcomes. The testing set, conversely, evaluates the model's performance after training, providing an unbiased assessment of its ability to generalize to new, unseen data rather than merely replicating learned patterns. This approach ensures both the accuracy and the general applicability of the model. Each instance in the dataset has an equal chance of being selected for the training or testing set because the data used for training and testing will be randomly36. We ran several analyses to investigate the association between PUFA, oxylipins, etc., and SLE. First, we ran an ML model using the status of SLE as the outcome. This model type is called the classification ML model, considering subjects binary, which means subjects are categorized in either with SLE or without SLE group. We then ran another group of ML models with the severity scores of diseases as the outcome of the model; this type of model is called the regression ML model, considering data as nonbinary and with continuous lupus score (Table 3.4). We also ran several models to compare the oxylipin dynamics between healthy subjects and SLE patients. Before running ML models, we scanned the data for missing values. There were 53 missing values (0.1%). We dropped the cases with missing values and ended up with 477 cases for ML models. 128 It is worth noting that data distribution was not a major issue based on descriptive analysis, so we did not consider the log transformation to improve the normality of data. The second reason to avoid transformation is the difficulty of interpreting data. All analyses and modeling were performed using R v.4.1.1 and MPlus (v8). 2.4.1. ML algorithms Predicting variables such as the oxylipin levels and the demographic variables are called features. We used several ML algorithms to investigate the predicting power of the features. The outcome variables, SLE and SLE severity, are categorical and continuous, respectively. Thus, we used ML algorithms that can handle classification and/or regression tasks. The algorithms included support vector machine (SVM), generalized boosting methods (GBM), regularized logistic regression (GLMNET), random forests (RF), one-layer neural network for classification and prediction (NNET), linear regression (LM), K nearest neighbor (KNN), and extreme gradient boosting models (XGB). Then, we ensembled the algorithms' predictions to gain more robust findings. We explained the ML algorithms in the following paragraphs. SVM was developed by Cortes and Vapnik (1995)37. The algorithm can be used for both classification and regression problems. It is it is a supervised ML model when we know the actual outcome. SVM maps all data points in an n-dimensional space formed by the n features or covariates. Then, it derives the optimal decision hyperplanes to separate the data points into different classes. The projection of the data points involves a kernel function. We selected the kernel among the most popular ones, including linear, radial, and polynomial kernels. The determination of the kernel in use was based on the accuracy of the holdout set. Parameter c needs to be turned in SVM. The parameter controls the balance between the sensitivity of the boundaries among the groups and the misclassification rates. Smaller c will result in less sensitive boundaries 129 and high misclassification rates. We identified the best value of c by trying a range of values (“tuning of the c parameter”). The GBM (Friedman, 2002) algorithm ensembles the predictions from multiple models38. Because these models are slightly better than random guesses, they are called weak learners. Each weaker leaner improves upon the previous one. This algorithm uses a subsample of the data to build the subsequent decision trees to avoid local minima and to approach global minima. There are several turning parameters: number of trees, depth of each tree, learning rate, and minimum observations allowed in the tree’s terminal nodes. This algorithm is suitable for both classification and regression tasks. GLMNET (Friedman, Hastie, & Tibshirani, 2010) penalizes model complexity by constraining the sizes of the coefficients39. We used a model called Elastic Net, a linear combination of ridge regression and the least absolute shrinkage and selection operator (LASSO). These are the two main types of regularizations. The best combination of the two is found through cross-validation. The turning parameter of the algorithm is lambda, which determines the importance of the penalty. GLMNET can be used both for classification and regression purposes. RF (Wright & Ziegler, 2017) is also suitable for classification and regression. It ensembles the predictions of multiple classification trees. It is also a type of ensemble method.40 The multiple classification trees were created using bootstrapped samples of the data and the randomly selected subsets of the predictors. The tuning parameters include the number of selected predictors, splitting rules, and minimal node size. Single-hidden-layer neural network models (Venables, & Ripley, 2002) extract features (or units) using nonlinear regressions41. These features will then be used for predictions. In general, one can perform feature extractions multiple times. Each extraction is called a layer. In the present research, we are considering a single-layer neural network. The turning parameters are the number 130 of hidden units and decay used to avoid over-fitting. It is specialized in classification tasks. A linear regression model (LM) delineates the relations between a numeric outcome variable and several predicting variables with a linear equation. One tweaks the value of the intercept to find the best-fit model. It is an algorithm developed for regression tasks. KNN (Cover & Hart, 1967) predicts the outcome by finding the nearest neighbors42. The observations' closeness was assessed using a distance measure like Euclidean distance. The parameter tuned for the algorithm is the number of neighbors. We tried scenarios ranging from 2 to 50 neighbors. KNN is suitable for regression. An extreme gradient boosting model (XGB, Chen, & Guestrin, 2016) is a special implementation of the gradient boosting method, which builds models by putting so-called “weak learners” models together43. It is like a Random Forest model. The gradient is to minimize the loss function. XGB computes the second partial derivatives of the loss function to minimize it. XGB models are more generalizable, and the computation will be faster than other gradient-boosting methods. This algorithm is for regression tasks. The parameters of the ML algorithms were learned using 80% of the random sample of the data set, which is called the training set. We used 10-fold repeated cross-validation to tune the parameter estimates of the ML algorithms. The performance was evaluated on the remaining 20% of the data, called the test data set. The accuracy, sensitivity, and specificity of models were compared for the classification models. R-squared was used to evaluate the performance of the regression models. We then ensemble the results of the algorithms for both the classification and regression models. To ensemble the classification models, we took a two-step approach. 1) We output the probabilities of the cases having SLE of each algorithm. These probabilities were then used as the input of an RF model with the SLE status as the outcome. The performance of the 131 ensembled model was also evaluated with accuracy, specificity, and sensitivity (table 2). The ensemble of the regression models was carried out using the Caretensumble function in R. We also compared the relative contributions to the prediction of the features using variable importance. For the classification models, the variable importance for the algorithms was then used to weigh the variable importance of each algorithm. For the regression models, the variable's importance was calculated using the vamp function in R. In addition, a logistic regression model was constructed to quantify the effects of the top features (rankings higher than 15). 3. Result 3.1. Descriptive analysis to study the demographic composition of SLE and healthy subjects We ran descriptive analysis for the entire testing and learning data set. For a detailed description of how each dataset was assembled, please refer to the experimental (section 2.4). Within the population, roughly 70% of the subjects had SLE, with 90% of the SLE patient population being female. About 55% of the subjects were White, and 41% were Black, with <5% of the subjects from other races. The average age of the subjects with SLE was 53.52 years old, and the average age of healthy subjects was 52.86. Therefore, the identified changes between SLE patients and healthy subjects are less likely due to age effects. The learning and test data had characteristics like those of the entire data set. More descriptive analyses are provided in Table 3.1. 132 Table 3.1. Descriptive results, characteristics of study participants including sex, race, age, severity score, status of disease across full data, learning, and test sets. Omega-3 index is define as sum of EPA and DHA divided to sum of EPA, DHA, ALA, LA, DGLA, AA SLE status Sex Race Age Control SLE Female Male White Black Other/DK Mean Median SD Kurtosis Skewness Min Max Mean Median SD SLE severity score Kurtosis Skewness Min Max Mean Median SD Omega 3 index Kurtosis Skewness Min Max Full data Learning Test n 134 343 425 52 258 198 21 % 28.1 71.9 89.1 10.9 54.1 41.5 4.4 n 108 275 336 47 209 156 18 % 28.2 71.8 87.7 12.3 54.6 40.7 4.7 n 26 68 89 5 49 42 3 % 27.7 72.3 94.7 5.3 52.1 44.7 3.2 53.6 54.8 13.0 -0.1 -0.3 18.4 91.9 9.2 8.0 8.8 0.07 0.8 0 38 0.058 0.055 0.019 1.586 0.997 0.015 0.139 52.4 53.3 12.3 -0.4 -0.3 22.7 81.5 8.8 8.0 8.1 -0.8 0.5 0 31 0.057 0.055 0.017 0.582 0.485 0.017 0.107 53.3 54.6 12.9 -0.2 -0.3 18.4 91.9 9.1 8.0 8.7 -0.02 0.8 0 38 0.058 0.055 0.018 1.532 0.932 0.015 0.139 133     3.2. Altered PUFA metabolite correlations in SLE individuals: insights into oxylipin metabolic disruption The initial attempt to examine the differences in the profiles of PUFA metabolites between individuals with SLE and healthy individuals involved analyzing the correlations among various PUFA metabolites (Figure 1). We used 3 heatmap analyses to visualize the correlations in the whole data set, the SLE population, and the healthy population. Strong positive correlation with coefficient (+1) and negative correlation (-1). We observed darker shades of red among some PUFA metabolites—in healthy individuals compared to those with SLE, suggesting a relatively stronger correlation in healthy individuals and that SLE may disrupt oxylipin metabolic homeostasis. For example, DHA derivative 17-HDoHE has a relatively stronger correlation with 22-HDHA and 20-HDHA in healthy individuals than SLE individuals. For LA derivatives, the correlation between 9-oxoODE and 13-oxo-ODE was stronger in healthy than SLE individuals. The same trend was seen for the correlation between 13-HODE and 9-HODE and between 12,13- EpOME and 9,10-EpOME. Regarding AA metabolites, the correlation of PGE2 with 11-HETE and 15-HETE, the correlation of 15-HETrE with 12-oxo-ETE, and the correlation of LTB4 with 5-HETE and 5-oxoETE are relatively stronger in healthy individuals than SLE individuals. These relative changes in heatmap correlation suggest that the metabolic pathways converting these PUFA into their specific metabolites are relatively stronger when SLE is not present. These results indicated that the PUFA oxidative metabolism is likely altered in subjects with SLE. Since these oxylipins play a critical role in inflammation and tissue repair, such alternation of oxylipin metabolism could be a result of SLE pathogenesis; therefore, more studies are needed to understand better how PUFA metabolism is affected by SLE pathogenesis and if any of the oxylipins play a role in SLE 134 pathogenesis. On the other hand, we identified other lipid metabolites that maintain their correlation between healthy and SLE subjects. For example, the correlation between 12,13- DiHOME and 9,10-DiHOME was the same among different study groups (Figure 3.1). Notably, a change in metabolites' correlation was not observed for the other three PUFAs, EPA, DGLA, and ALA (Figure 3.1). Understanding these metabolic variations offers valuable insights into the pathophysiological mechanisms of SLE and could lead to identifying potential diagnostic markers or therapeutic targets. 135 A Figure 3.1. Correlation visualization of all PUFA metabolites. Correlations range from +1 to -1. The most robust positive correlations (1+) are depicted in dark red, while negative correlations (-1) are illustrated in dark purple. A) all data (n= 477), B) healthy individual’s data (n=134), and C) SLE individual’s data (n=343) 136 Figure 3.1 (cont’d) B 137 Figure 3.1 (cont’d) C 138 3.3. Alterations in oxylipin concentrations within SLE subjects: Cytochrome P450-sEH metabolism insights from targeted metabolomics We initially employed a heatmap analysis to identify correlations among all detected metabolites. This approach allowed us to visualize the global metabolic changes and highlighted specific metabolites that exhibited significant variation in correlation between the two groups. Subsequently, we conducted a focused analysis of individual metabolites to quantify their concentrations and assess statistical significance precisely. Our targeted analysis observed significant elevations in the concentrations of multiple oxylipins in SLE compared to healthy individuals (Figure 3.2). Figure 3.2. Comparative analysis of oxylipin concentrations between SLE (n=343) and healthy (n=134) individuals (mean+/- SEM). Data were analyzed using two-way ANOVA followed by Tukey's multiple comparisons test. (p-value < 0.0001 ****) Notably, oxylipins such as 14,15-DiHETE, 9,10-DiHOME, 12,13-DiHOME, 13,14-DiHDPE, 19,20-DiHDPE, and 9-HODE were significantly lower in the SLE group. Meanwhile, the concentration of 9,10-EpOME showed the opposite trend. Since most of these metabolites are either substrates or products of CYP-EH metabolism except 9-HODE, these analyses suggested 139 that SLE affects these metabolisms, and such metabolites could be biomarkers for SLE. Conversely, other metabolites, including 15S-HETE, PGD2, 8,9-DiHETE, 15-HEPE, and 10,11- EpDPE, 5-HETE, 9,10-DiHODE, 17,18-DiHETE, show no statistical differences in the serum concentrations, between SLE and healthy individuals, the same result was seen for all six PUFAs (Appendix Figure2). Combining heatmap correlations and individual metabolite assessments, these comprehensive analyses have elucidated distinct oxylipin profiles associated with SLE. 3.4. Comparison of PUFA metabolites’ levels in different demographics: Sex, race, and age To develop a more comprehensive understanding of the differing oxylipin profiles in SLE and healthy individuals, we intend to expand our analysis by incorporating additional demographic factors such as age, sex, and race. This stratified approach will allow us to determine whether these factors impact oxylipin concentrations and whether specific subgroups within the SLE and healthy populations display unique metabolic profiles. 3.4.1. Sex-dependent variations in oxylipin levels among healthy and SLE subjects Our analysis identified several oxylipin levels that are significantly different between sexes. Our results indicated that within healthy subjects, the plasma levels of 14,15-DiHETrE and 12,13- DiHOME are higher in female subjects than males. Healthy female subjects have a higher concentration than SLE females, which aligns with our analysis of SLE subjects and healthy subjects. For metabolites 12,13-DiHOME, 19,20-DiHDPE, 15-HETE, 11-HETE, 9-HODE, 13- HODE, 17,18- DiHETE, 15-HEPE, and PGD2 female with lupus also demonstrated a significantly higher concentration than the male with lupus. Furthermore, as we described before, we observed that plasma levels of several CYP-EH metabolites are significantly affected by SLE pathogenesis. The plasma levels of epoxy fatty acids, 9,10-EpOME, and 10,11-EpDPE are higher in female subjects with SLE in comparison with healthy female individuals, while the plasma levels of 9,10- 140 DiHOME and 19,20-DiHDPE are lower in female subjects with SLE. Altogether, the sex demographic affects several specific oxylipin metabolisms, but not all of them (Figure 3.3). Figure 3.3. Illustrates the comparison of metabolite concentrations between healthy and SLE individuals across various demographics. Data were analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (mean+/- SEM). P-value<0.5 *, P- value<0.01 **, P-value<0.001*** 3.4.2. Race-dependent variations in oxylipin levels among healthy and SLE subjects Our analysis showed that among all the oxylipins, 14,15-DiHETE, 13,14-DiHDPE, and 22-HDHA are upregulated in lupus subjects with White ethnicities compared to lupus subjects with Black ethnicities. 9,10-DiHOME and 12,13- DiHOME, and14,15-DiHETrE demonstrated a notable decrease in concentration in White individuals with lupus in comparison to healthy White 141 individuals, compound 14,15-DiHETrE showed the same trend for people with black ethnicities. Our observations suggested that race minimally affects oxylipin metabolism (Figure 3.3). 3.4.3. Age-dependent variations in oxylipin levels among healthy and SLE subjects To investigate whether aging affects oxylipin metabolism, samples were classified into three distinct age groups: (18-40) years as a young group, (40-60) as a middle-aged group, and older than 60 as the old category. While some compounds like 14,15-DiHETrE and 19,20-DiHDPE showed a significant decrease between healthy and SLE in the old age group, our results indicated that aging generally does not impact oxylipin profile. The only compound with significant change among different age groups was 9-HODE, with a significant concentration decrease in the young lupus category compared to the middle-aged and old categories (Figure 3.3). 3.5. Analyzing the impact of SLE pathogenesis on ratios of epoxy to dihydroxy metabolites across different demographics Our initial analysis revealed that SLE pathogenesis likely affects CYP-EH metabolism, based on the significant change in metabolites of this pathway between healthy and SLE populations (Figures 3.2, and 3.3). Therefore, we also analyze the ratio of the epoxy fatty acid, which is the substrate of EH, to dihydroxy fatty acid, which is the product of EH, that could indicate whether EH activity is impacted in SLE individuals. We initiated our analysis by examining the variations in each epoxy fatty acid/dihydroxy fatty acid ratio between healthy and SLE individuals. The findings indicated that almost all epoxy fatty acid/dihydroxy fatty acid ratios analyzed with SLE exhibited elevated amounts compared to their healthy counterparts, aside from the specific case of the 13,14 EpDPE/13,14DiHDPE, which did not show a significant difference (Figure 3.4). Together, it suggested that either the EHs are downregulated or the metabolism of the dihydroxy fatty acid increases. 142 Figure 3.4. Illustrates the comparison of metabolite concentrations between healthy and SLE individuals across various demographics, including age, race, and sex. Data were analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (mean+/- SEM). P- value<0.5 *, P-value<0.01 **, P-value<0.001***, P-value<0.0001**** Subsequently, like in previous analyses, we assessed the metabolite ratios while considering various demographic variables, including sex, age, and race. Exploring the sex category, we found that the following ratios exhibited significant differences: for AA metabolites, the ratio of 8,9- EpETrE/8,9-DiHETrE, 14,15-EpETrE/14,15-DiHETrE, and 17,18-EpETrE/17,18-DiHETrE, for LA metabolites 9,10-EpOME/9,10-DiHOME and 12,13-EpOME/ 12,13-DiHOME, For the ALA metabolite, the ratio of 9,10-EpODE/ 9,10-DiHODE, and For DHA metabolite, the ratio of 10,11- EpDPE/10,11-DiHDPE, in all cases healthy females had a significantly lower ratio compared to lupus females. 143 Then we explored the racial demography; the ratio of 8,9-EpETrE/ 8,9-DiHETrE, 14,15- EpETrE/14,15-DiHETrE showed healthy individuals of White ethnicity having a lower ratio than lupus individuals of White ethnicity. The ratio of 9,10-EpOME/9,10-DiHOME and 12,13- EpOME/ 12,13-DiHOME showed healthy individuals of Black ethnicity had a significantly lower ratio than lupus individuals of Black ethnicity, for 12,13-EpOME/12,13-DiHOME the trend was same for White population. Finally, we considered the different age categories. 8,9-EpETrE/8,9-DiHETrE, 9,10- EpOME/9,10-DiHOME, and 12,13-EpOME/ 12,13-DiHOME showed old healthy individuals had a lower ratio than old lupus individuals. 14,15-EpETrE/14,15-DiHETrE, 9,10-EpOME/9,10- DiHOME, and 12,13-EpOME/ 12,13-DiHOME, the ratio showed notable distinctions; in the middle-aged group, healthy individuals had a lower ratio than lupus individuals in the same age range. For the other metabolites, the ratio of 15,16-EpODE /15,16-DiHODE displayed differences between young lupus individuals and the other two age groups, with both old lupus individuals and middle-aged lupus individuals having a higher ratio than young lupus individuals. Altogether, from a sex perspective, healthy females consistently exhibited lower ratios of epoxy to dihydroxy fatty acid metabolites compared to lupus females across various compounds. Concerning different races, healthy individuals of White ethnicity consistently had lower ratios of various epoxy to dihydroxy fatty acid metabolites than lupus individuals of White ethnicity. Moreover, healthy individuals of Black ethnicity had a higher ratio compared to healthy individuals of White ethnicity. From different age group points of view, old healthy individuals generally had lower ratios of epoxy to dihydroxy fatty acid metabolites compared to lupus individuals in the same age group. For 15,16-EpODE/15,16-DiHODE, old and middle-aged lupus individuals had higher ratios than young lupus individuals. Our findings highlight the potential 144 change of the sEH enzyme among healthy and SLE groups of different demographics, sexes, ages, and ethnicities. However, it is important to note that no significant differences were observed between males and females, Black and White individuals, or across different age groups, except for the specific case of 15,16-EpODE/15,16-DiHODE. 3.6. Variation in the ratio of omega-6/ omega-3 metabolites across different demographics The plasma ratio of omega-6 to omega-3 PUFA metabolite is critical to human health. Omega-6 and omega-3 fatty acids metabolites affect the body's inflammatory and immune responses. However, an imbalance in the ratio, with excess omega-6 compared to omega-3, can contribute to chronic inflammation25,26. There was no significant change in the ratio of the metabolite between the healthy and SLE groups (Figure 5). We shifted our focus to differences in specific demographic categories, including sex, age, and race (Figure 3.5). Figure 3.5. Represents changes in the ratio of overall omega-3/6 metabolites between healthy and SLE in different demographics, including sex, race, and age. Data were analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (mean+/- SEM). P- value<0.05 *, P-value<0.01 **, P-value<0.001*** Notably, significant changes were observed in the ratios when comparing healthy females to lupus 145 females, lupus females to lupus males, healthy Black individuals to lupus Black individuals, and old individuals with lupus to demographic subgroups, emphasizing the potential influence of factors such as sex, race, and age on the metabolism and distribution of these essential fatty acids metabolites. 3.7. Lupus flare and different metabolite and demographic categories The questionnaire categorized lupus flare into four groups: no flare, mild flare, moderate, and severe flare. We studied the level of PUFAs and their metabolites in different flare severity. For AA and EPA, the concentration in healthy individuals is significantly higher when compared to SLE patients with moderate flare. Still, the concentrations do not correlate with the severity score of SLE patients. Following the PUFA analysis, we also studied metabolites 14,15-DiHETrE, 9,10- DiHODE, and 17,18-DiHETE and the severity of SLE (Figure 3.6). Our primary focus on these three metabolites was driven by results from ML analysis, which identified them as having the highest variable importance; more details are provided in the following section. Although some significant changes were seen, levels of metabolites were not correlated with flare severity. For this analysis, the variation within each group appears quite high, as indicated by the length of the variation bars. 146 Figure 3.6. Metabolite concentration variations across SLE disease severity levels. This figure illustrates the levels of selected metabolites (14,15-DiHETE, 9,10-DiHODE, 17,18-DiHETE, AA, and EPA) in healthy and SLE individuals experiencing no flare, mild flare, moderate flare, or severe flare. Data were analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (mean+/- SEM). P-value<0.05 *, P-value<0.001***, P-value<0.0001**** . While our results suggest that SLE affects the plasma levels of specific PUFA metabolites, the disease's flare severity does not further impact the plasma concentrations of these fatty acids metabolites. 147 3.8. Machine Learning Results 3.8.1. Identifying key metabolic and demographic predictors of lupus presence through classification ML Analysis We test our dataset in multiple machine-learning models to investigate whether specific oxylipins can be used as biomarkers for SLE pathogenesis. The prediction performances of the algorithms were evaluated on the test set or holdout set of 94 cases (20% of the cases). The learning and test data sets are comparable. Correlations among the predictors are relatively low. In general, a correlation coefficient smaller than 0.4 is considered low, between 0.4 and 0.8 is considered moderate, and above 0.8 is considered high; in our analysis, the correlation was around 0.2. The algorithms vary in terms of their performance, the accuracy of ensemble prediction, and the sensitivity and specificity are provided in (Table S4). The variable importance evaluates the contribution to the accuracy of predictions of a variable. We presented both the variable importance of each algorithm and the overall importance. The overall importance was the weighted averages of the variable importance of the algorithms. The weights for the algorithms were calculated based on the importance of the algorithms from the ensembled model (Table S5). Table 3.2. Variable importance of classification ML models 14,15-DiHETrE 9,10-EpODE/9,10-DiHODE 5,6- EpETrE/5,6-DiHETrE Age 14,15-EpETrE/14,15-DiHETrE 12,13-EpOME/12,13-DiHOME AA 13-oxo-ODE Currently using steroid 15,16-EpODE 15-HETE RF 100.000 95.6663 89.0760 96.7038 82.6809 85.1746 66.8020 76.6994 48.0632 71.0927 74.6201 NNET 72.2549 19.7965 19.6276 8.1083 17.2923 11.0020 76.4077 27.4460 66.7743 50.7390 32.9981 Overall 91.9489 88.0904 79.0008 77.6358 70.3567 63.4777 59.9277 57.9089 56.8947 56.0745 55.7494 SVM 94.4000 96.0700 82.2900 7.5900 100.000 93.5438 30.5569 66.8739 70.5205 9.0419 45.6324 148 Table 3.2 (cont’d) 9,10-EpOME/9,10-DiHOME 9,10-DiHOME Omega 3/6 metabolites 19,20-DiHDPE 12,13-DiHOME 12,13-EpOME 11-HETE 17,18-EpETE/17,18-DiHETE 9,10-EpODE 13-HODE 9,10-DiHODE 17,18-DiHETE 9-HOTrE ALA 11,12-EpETrE/11,12-DiHETrE 5-HETE 5,6-DiHETrE 15-oxo-ETE 9,10-EpOME 8,9-EpETrE/8,9-DiHETrE 9-HODE Omega-3 PUFA 19,20-EpDPE/19,20-DiHDPE Omega-6 PUFA 8,9-DiHETrE Sex 11,12-DiHETrE 15,16-DiHODE LA 12,13-EpODE 20-HDHA 5,6-DiHETE EKODE 12,13-EpODE/12,13-DiHODE DHA 13,14-DiHDPE TXB2 15,16.EpODE/15,16-DiHODE trans THF diol 77.9909 71.5386 69.7029 72.0083 67.6005 70.5983 76.3917 79.1017 48.2830 64.3742 67.0712 68.6990 63.3732 64.2459 66.2063 61.2495 66.9009 49.2666 64.2816 64.1278 61.3962 60.8220 64.4369 60.5393 57.7334 34.1224 57.8581 61.2003 57.5216 49.0978 50.4925 46.2278 49.1326 45.1585 55.6531 40.6028 55.9820 58.2936 57.6933 77.2667 75.5263 9.1745 55.3622 74.5317 69.3353 28.3773 70.3796 67.8518 23.8770 42.7151 64.2798 36.2672 8.1634 15.0506 29.1066 56.3816 14.9014 11.7603 37.6513 20.6116 24.3494 50.6547 19.0535 35.6456 36.8556 13.8820 12.0918 12.5891 41.8531 2.5361 39.4497 26.3053 37.5601 26.6534 17.1556 27.8303 6.8125 26.5125 149 13.4805 21.2706 46.5863 54.7075 31.0074 54.6099 41.3957 1.8801 100.000 47.9432 74.5147 42.7308 17.2394 31.5643 58.3382 30.4023 22.1934 54.0858 31.1793 49.0291 46.6905 21.3863 4.9329 28.2938 18.2992 59.1633 29.2673 20.2148 20.0926 17.4264 55.2185 29.3611 30.1783 35.0481 25.8622 45.3769 27.0272 20.7364 13.3942 55.5802 54.5005 53.1569 50.6963 50.2252 50.1077 49.9671 49.4375 49.0975 47.7753 47.7589 47.2197 46.5023 46.2052 45.6626 44.4484 44.0020 43.3893 43.1811 42.7607 42.5778 42.3122 40.1593 40.1455 39.9509 39.5162 39.2049 37.9763 37.9136 37.8517 37.5058 36.9611 36.5340 36.1423 36.0857 36.0684 35.9848 35.6087 35.1219 Table 3.2 (cont’d) 13-HOTrE 9-oxo-ODE EPA DGLA 10,11-EpDPE/10,11-DiHDPE 5,6-EpETE/5,6-DiHETE 11,12-DiHETE 15,S-HETrE 16,17-EpDPE/16,17-DiHDPE 14,15-EpEDE/14,15-DHED 16,17-DiHDPE 11,12-EpETE/11,12-DiHETE PGE2 14,15-EpETE/14,15-DiHETE 14,15-DiHETE 13,14- EpDPE/13,14-DiHDPE 14,15-DHED 14,15-EpETrE 22-HDHA 10,11-DiHDPE 12,13-DiHODE PGF2a LTB4 7,8-EpDPE/7,8-DiHDPE 7,8-DiHDPE 5,6-EpETrE 8,9-EpETE/8,9-DiHETE White PGD2 Black 5-oxo-ETE 17-HDoHE 11,12-EpETE 10,11-EpDPE Statin 15,HEPE PGD1 8,9-DiHETE Fish oil 59.5139 54.6449 54.4885 52.9342 41.0051 37.5305 35.8660 52.3379 47.4762 31.5111 44.9020 26.6406 36.4970 30.6843 27.9464 39.4986 26.9563 34.0563 30.7579 32.2504 23.0455 35.9860 31.0571 24.5647 24.1632 20.6047 15.7208 16.4157 9.3681 15.0744 15.4843 13.1649 9.4880 13.1806 12.4838 13.9706 10.7861 12.2613 12.8878 18.2579 20.1475 16.0865 1.4255 1.5415 29.9354 50.6464 17.5783 13.5836 42.5079 13.6251 39.6900 18.2994 25.0870 26.7446 16.7413 42.5079 35.9191 26.1561 14.5947 17.8352 14.5201 8.9922 33.2670 33.2587 33.4659 15.9291 25.1699 9.9867 18.5811 2.5692 1.9808 19.2607 20.4542 13.2273 0.5138 3.7295 15.6307 7.1523 150 20.3727 20.6800 16.4532 41.0904 56.2709 29.1599 0.1475 20.4025 32.6101 74.9194 28.9380 16.5258 17.5898 43.2739 26.6488 8.3634 16.5670 5.9735 28.2028 27.8239 59.1286 1.4692 11.5736 22.2135 25.1741 36.0684 61.8465 24.0376 65.5024 31.3269 36.7375 45.2333 10.4171 20.6633 20.5952 22.3606 32.3627 13.6102 6.6805 35.0133 32.7357 32.7071 32.2212 31.8550 31.4107 31.2726 31.2400 29.7151 29.4255 27.9667 24.0260 23.8730 23.4976 23.2099 23.1334 22.2746 22.0832 21.9097 21.4096 20.6142 20.2886 19.9702 19.9112 18.9094 18.3516 16.9442 14.8031 13.4558 13.3696 12.5064 12.2270 11.9817 11.3600 10.2255 9.8476 9.6992 9.5596 8.1835 Table 3.2 (cont’d) 8,15-DiHETE 19,20-EpDPE 12-oxo-ETE 8-isoPGF2a Fish oil and flax oil 17,18-EpETE 8,9-EpETE Flax oil 13,14-EpDPE 11,12-EpETrE 8,9-EpETrE RVD1 8,9-EpEDE/8,9-DHED 14,15-EpETE 5,6-EpETE PGD3 5,15-DiHETE 4.1273 0.2321 0.9034 2.7598 10.7161 9.4149 0.0580 2.5775 1.3923 0.4641 1.3592 4.0444 0.4641 0.0000 0.4641 0.4641 0.4641 4.6184 8.5028 1.6095 4.2178 3.9839 1.3340 2.4035 3.5056 2.8984 0.7587 0.1745 1.5345 0.1623 1.7213 0.1619 0.2977 0.0000 34.0813 17.1853 45.4481 26.9400 14.6437 14.5703 17.6479 4.6767 7.5034 14.9989 14.5156 5.8229 11.0193 3.7426 7.5334 4.7617 0.0000 6.7556 6.3976 6.1719 5.5795 4.8840 3.3694 3.2837 2.6230 2.5100 2.1719 1.9096 1.8934 1.4048 1.3206 1.0022 0.7525 0.0493 The machine learning model's analysis of various factors about lupus presence has been detailed in Table 3.2. The compound 14,15-DiHETrE holds the highest importance score (91.95), indicating a strong association with lupus presence. In addition, several epoxy fatty acids (15,16- EpODE with a score of 56.07 and 12,13-EpOME with a score of 50.11) and dihydroxy fatty acid metabolites (9,10-DiHOME with a score of 54.50; 19,20-DiHDPE with a score of 50.70 and 12,13- DiHOME with a score of 50.23) and a couple of specific epoxy to diol ratios (9,10-EpODE/9,10- DiHODE with a score of 88.09; 5,6-EpETrE/5,6-DiHETrE with a score of 79.00; 14,15- EpTrE/14,15-DiHETrE; 12,13-EpOME/12,13-DiHOME with a score of 63.48 and 9,10- EpOME/9,10-DiHOME) are among the most important factors to predict whether individuals have lupus or not. This study suggests that the SLE pathogenesis impacted CYP-EH metabolism. Moreover, we also found that 13-oxo-ODE and 15-HETE could be other lipid metabolites that predict the presence of lupus. Interestingly, the ML analysis indicated that the omega-3/omega-6 151 metabolites ratio could also help differentiate individuals with lupus from healthy ones, and the analysis also demonstrated that omega-6 PUFA, AA ranked higher than the other 5 PUFAs with a score of 59.92. Such a discovery may have future implications because omega-3 fatty acid metabolites are generally considered to be anti-inflammatory and pro-resolving. In contrast, omega-6 fatty acid metabolites such as prostaglandins are largely pro-inflammatory25,26. Apart from fatty acid-related factors, our ML analysis identified the current use of steroids showing a high importance score (56.89), underscoring its significant role in lupus treatment. In addition, demographic factors such as sex (overall score 39.51) are well-represented in the model, reflecting known disease patterns where lupus is more prevalent in females. Since steroid treatment is the standard treatment for lupus flares and age is one of the risk factors for lupus pathogenesis, these results show that our ML analysis can identify factors that can predict patients with lupus. Various classification models were utilized to determine the relative importance of variables in predicting the disease, although they do not indicate the magnitude or direction of the associations. To address this problem, logistic regression was applied to variables with an overall importance score above 50 from the classification results (Table 3.2). Logistic regression models yield estimates and p-values for non-binary datasets (such as the presence or absence of SLE), where the p-value indicates significance, and the estimate reveals the strength and direction of the association with the disease presence. For instance, the compound 14,15-DiHETrE has an estimate of (-0.03427), suggesting a negative association, though its p-value indicates it is insignificant. AA also has a negative association estimate of (-0.0023), but unlike 14,15-DiHETrE, the p-value shows significance. The ratios of 9,10-EpODE/9,10-DiHODE, 14,15-EpETrE/14,15-DiHETrE, and Steroids all have positive significant association based on the estimates and p-values (Table 3.3). 152 Table 3.3. Logistic regression model (Intercept) 14,15-DiHETrE 9,10-EpODE/9,10-DiHODE 5,6-EpETrE/5,6-DiHETrE Age 14,15-EpETrE/14,15-DiHETrE 12,13-EpOME/12,13-DiHOME AA 13-oxo-ODE Steroid 15,16-EpODE 15-HETE 9,10-EpOME/9,10-DiHOME 9,10-DiHOME 3/6.metabolites 19,20DiHDPE 12,13-DiHOME 12,13-EpOME Estimate 0.529561 -0.03427 1.494594 0.026044 -0.00043 8.351276 0.197138 -0.0023 0.084164 2.395426 -0.02283 -0.03421 -0.62942 0.001271 0.645697 0.006201 -0.0026 -0.00347 Std. Error 0.674275 0.019176 0.57351 0.037957 0.008932 3.363219 0.448272 0.001157 0.056771 0.488772 0.03521 0.049219 1.12567 0.001145 1.130911 0.00776 0.002091 0.013722 Exp (Est) 1.698187 0.966315 4.457524 1.026386 0.999566 4235.582 1.217912 0.997703 1.087807 10.97287 0.977429 0.966368 0.532902 1.001272 1.907316 1.00622 0.997407 0.996539 Pr(>|z|) 0.43223 0.07395 0.00916 0.49263 0.96123 0.01302 0.6601 0.04695 0.1382 9.54E-07 0.51673 0.48702 0.57606 0.26697 0.56803 0.42423 0.21431 0.80056 3.8.2. Identifying key metabolic and demographic predictors of lupus severity through regression ML Analysis Regression ML models studied the regression between the severity of the diseases based on severity score (provided in the questionnaire) and different factors, including metabolite level, ratio, and demographics. We evaluated the regression algorithms' performance based on the models' R-squared using the test or holdout data set (Table S6). The overall score of each variable is presented in Table 3.4. The usage of steroids emerged as the most significant predictor, with a high importance score (4.02), suggesting a robust association between steroid treatment and the severity of lupus. This result is expected as steroids are commonly prescribed in lupus patients with severe flares. The regression analysis identified several PUFA metabolites, specifically 5- 153 HETE (score of 2.7), 14,15-DiHETrE (score of 2.4), and 15-S HETrE (score of 2.25), as the most important variables concerning lupus severity. While the ratio of 12,13-EpOME/12,13-DiHOME (score of 2.24) and 9,10 EpOME /9,10 DiHOME (score of 2.23) ranked as the fourth and fifth important variables of the severity of the disease. The significant association of these metabolites and ratios with the disease suggests that they may have roles in the inflammatory pathways contributing to lupus or as indicators of disease activity. The omega 3 index, with a score of 2.03, was one of the most critical variables of the severity of the diseases. The age of the patients was identified as another strong predictor of disease severity (score of 1.88). This aligns with existing knowledge that the immune system and disease progression are influenced by age, which has been observed in autoimmune diseases, including lupus. On the other hand, other demographics like sex and race were not as profound as age, ranking at the bottom of the table. At the lower end of the importance spectrum, variables like fish oil and flax oil as a supplement and Statin had minimal impact on lupus severity within the context of the model. This suggests that, while these factors may influence the disease, they do not play a significant role in its severity as per the current dataset. We ran another study to find the magnitude and direction of variables with the severity of the diseases. For regression models, we selected the linear regression approach to examine the associations of variables more closely. We only focused on variables with an overall score above 1.85 (based on Table 3.4) and the severity of the disease. The choice of linear regression over logistic regression is due to the nature of the analysis, as severity in this context is treated as a continuous, non-binary variable, making linear regression more suitable. Logistic regression was more suitable for the classification model as a binary data set. The linear regression model offers estimates and p-values; the estimate indicates the magnitude and direction of each variable's 154 association with disease severity, while the p-value determines its significance. For example, steroid showed a significant positive estimate (4.857866), and 14,15-DiHETrE showed a significant negative estimate (-0.05403) (Table 3.5). The results present a multifaceted view of the factors influencing lupus severity, with steroid treatment, age, omega-3 index, and specific PUFA metabolites and ratios standing out as significant predictors. These findings provide a foundation for targeted therapeutic approaches and underscore the potential for personalized medicine in managing lupus. Table 3.4. Variable importance of regression ML models Overall LM KNN SVM XGB GLMNET RF Currently using steroid 4.0227 2.7691 4.9804 4.9804 6.6281 14.4484 5-HETE 14,15-DiHETrE 15, S-HETrE 2.7530 0.1782 3.0650 3.0650 3.2854 2.4117 0.6099 2.7018 2.7018 5.1877 2.2517 1.7734 2.1005 2.1005 4.0700 12,13-EpOME/DiHOME 2.2491 1.3446 3.6737 3.6737 2.6084 9,10-EpOME/9,10-DiHOME 2.2363 1.3678 2.7279 2.7279 1.5859 13-oxo-ODE Omega 3 index 9,10-DiHODE 2.0799 1.1877 2.9232 2.9232 1.4003 2.0135 0.6564 1.5841 1.5841 1.6017 1.9140 0.4753 2.6068 2.6068 1.0170 14,15EpEDE/14,15-DHED 1.8911 1.1662 2.6538 2.6538 0.0000 Age 1.8892 0.3810 2.7610 2.7610 0.5946 9,10-EpODE/9,10-DiHODE 1.8210 0.8436 0.0000 0.0000 3.0645 12,13-DiHOME 1.6983 1.1002 2.7669 2.7669 1.0604 9-oxo-ODE 20-HDHA 1.6715 0.1247 2.7583 2.7583 0.2560 1.6412 1.1317 1.3724 1.3724 1.7538 9,10- DiHOME 1.6337 1.3616 2.0873 2.0873 1.5925 Omega 3/6.metabolites 1.5700 1.4802 2.6669 2.6669 0.2432 14,15-EpETrE/DiHETrE 1.5439 1.1152 0.0000 0.0000 1.4119 9,10-EpOME 19,20-DiHDPE EPA 15-HETE 1.5430 0.7877 2.2528 2.2528 3.5192 1.4451 0.6743 1.8960 1.8960 0.9223 1.4206 1.3736 1.0233 1.0233 2.8485 1.3699 2.8428 1.0460 1.0460 1.6911 12,13-EpOME 1.3304 1.2993 2.1296 2.1296 1.9252 8,9-EpETrE/8,9-DiHETrE 1.3259 0.4244 2.3503 2.3503 0.0075 1.0609 0.0614 1.7576 0.0022 0.0061 0.3894 0.0000 1.3241 5.2371 0.0560 0.0590 0.0131 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.1621 0.7054 0.0000 1.4910 1.1650 1.9125 1.8152 1.4653 1.5662 1.7544 1.8129 1.5039 1.6063 1.0341 1.6432 1.8508 1.5713 1.5514 1.3818 1.7801 1.4146 1.5850 1.4976 1.5634 1.5595 1.5228 1.2226 1.1280 155 Table 3.4 (cont’d) 17,18-EpETE/17,18-DiHETE Omega6 9-HODE TXB2 13-HOTrE 15,16-DiHODE 5,6-DiHETrE Omega3 17,18-DiHETE DHA 1.2978 0.7492 0.9900 0.9900 0.2668 1.2743 0.0000 1.0499 1.0499 1.9764 1.2525 1.1221 0.6253 0.6253 0.8879 1.2128 0.1078 1.1174 1.1174 0.0888 1.1964 1.6666 1.3365 1.3365 0.5927 1.1902 0.9635 0.9556 0.9556 2.5082 1.1817 0.0572 1.0398 1.0398 3.0439 1.1688 0.0000 0.6104 0.6104 1.7102 1.1631 2.1986 1.1905 1.1905 0.2134 1.1615 1.7965 0.6917 0.6917 3.0921 7,8-EpDPE/7,8-DiHDPE 1.1036 1.3868 1.6042 1.6042 0.0000 7,8-DiHDPE 13-HODE 8,9-DiHETrE 1.0961 0.4336 1.3616 1.3616 1.3802 1.0669 1.3753 0.9608 0.9608 0.5765 1.0626 1.1286 1.5222 1.5222 1.2123 15,16-EpODE/15,16-DiHODE 1.0557 2.1677 0.6067 0.6067 0.3952 19,20-EpDPE/19,20-DiHDPE 1.0400 1.6874 0.5706 0.5706 0.0318 14,15-DHED 9,10-EpODE 11-HETE AA EKODE 15,16-EpODE 11,12-DiHETrE 1.0042 0.8551 0.0666 0.0666 0.6382 0.9902 0.2408 1.6399 1.6399 0.0015 0.9776 1.1926 0.5832 0.5832 3.5746 0.9770 1.5353 0.8160 0.8160 1.0145 0.9744 0.4471 0.9187 0.9187 0.0578 0.9696 0.1839 0.5739 0.5739 0.5481 0.9673 0.3185 0.0000 0.0000 2.7300 11,12-EpETrE/DiHETrE 0.9662 0.0647 0.5524 0.5524 0.0000 16,17-EpDPE/16,17-DiHDPE 0.9521 1.3893 0.9635 0.9635 0.0000 16,17-DiHDPE 0.9234 0.3647 0.9668 0.9668 0.2504 22-HDHA AA 0.9209 2.0970 1.3686 1.3686 2.0181 0.9192 1.3253 0.5095 0.5095 0.5185 10,11-DiHDPE 0.9093 1.8249 1.4963 1.4963 0.9471 LA 9-HOTrE trans-THF-diol DGLA 12,13-EpODE PGF2a 11,12-DiHETE 11,12-EpETE PGE2 0.9072 0.0693 0.5036 0.5036 4.5399 0.8925 0.4325 0.7392 0.7392 0.4604 0.8610 0.3208 0.3469 0.3469 2.8661 0.8594 0.1845 0.4229 0.4229 2.5356 0.8461 0.8389 1.3646 1.3646 0.0810 0.8409 2.5451 0.5182 0.5182 0.3182 0.7924 0.6704 0.2610 0.2610 0.9193 0.7817 3.6429 0.9267 0.9267 0.8352 0.7788 2.5320 1.4101 1.4101 0.6928 5,6-EpETrE/5,6-DiHETrE 0.7772 0.4626 0.0000 0.0000 0.1514 0.0000 0.0000 0.0000 0.0000 0.4669 0.0000 0.0000 0.0000 0.0000 0.0622 0.8274 3.3366 0.0000 0.0000 0.2124 1.8875 0.0000 4.6574 0.0000 0.1010 0.0000 0.0000 0.0000 0.0000 0.3244 0.0000 1.2321 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0342 0.0000 4.1950 0.0000 0.0000 1.5443 1.3313 1.3520 1.3112 1.4009 1.5437 1.4783 1.4176 1.4297 1.2223 0.8293 0.8668 1.3695 1.0717 1.2497 1.2141 0.9017 0.8378 1.3304 1.2715 1.2444 1.3953 1.3546 1.4133 1.1758 1.1276 0.8563 1.3357 0.8208 1.2283 1.1805 1.2630 1.2311 0.7991 0.8807 0.9240 0.4640 0.6388 1.3506 156 Table 3.4 (cont’d) 15-oxo-ETE 0.7769 1.6642 0.3304 0.3304 0.4219 14,15-EpETE/14,15-DiHETE 0.7628 0.2189 0.3380 0.3380 0.6773 13,14-DiHDPE 5,6-EpETrE 0.7382 0.6927 0.8458 0.8458 0.1611 0.6951 0.8285 1.2564 1.2564 0.1396 12,13-EpODE/12,13-DiHODE 0.6933 1.2919 0.5365 0.5365 0.3445 8,9-EpETE/8,9-DiHETE 0.6586 0.1891 0.3135 0.3135 0.0000 Sex LTB4 0.6143 1.0664 1.1080 1.1080 0.1504 0.6137 0.0254 0.4734 0.4734 0.0591 4.4821 0.0000 0.0000 0.9318 0.0000 0.0000 5.5445 0.0000 12-oxo-ETE 0.5805 3.6695 0.7989 0.7989 0.0630 20.9416 5,6-EpETE /5,6-DiHETE 0.5762 0.9584 0.6698 0.6698 0.0000 14,15-DiHETE 0.5503 0.1491 0.0025 0.0025 0.0216 13,14-EpDPE/13,14-DiHDPE 0.5496 0.4609 0.0000 0.0000 0.0371 11,12-EpETE/11,12-DiHETE 0.5369 1.6851 0.0000 0.0000 0.0000 10,11-EpDPE/10,11-DiHDPE 0.5171 1.3824 0.0000 0.0000 0.5365 8,9-DiHETE 14,15-EpETrE Fish and flax oil 5,6-DiHETE 12,13-DiHODE 10,11-EpDPE 5, oxo-ETE 15-HEPE 0.4921 1.9899 0.3157 0.3157 1.7369 0.4777 1.1357 0.2388 0.2388 1.4486 0.4310 2.3171 0.8123 0.8123 0.0136 0.3910 0.9340 0.0041 0.0041 0.0490 0.3576 0.6327 0.0155 0.0155 1.5063 0.3291 0.2504 0.2516 0.2516 0.0024 0.2921 0.6823 0.1648 0.1648 0.0343 0.2834 0.3439 0.1248 0.1248 0.0000 Flax oil supplement 0.2478 0.7670 0.2662 0.2662 0.0003 19,20- EpDPE 8,15-DiHETE White Black Statin 0.2460 0.1062 0.2798 0.2798 0.0000 0.2313 0.9849 0.4510 0.4510 0.1422 0.2233 0.9016 0.2081 0.2081 0.0036 0.2128 0.5217 0.1525 0.1525 0.0018 0.2077 0.0000 0.2134 0.2134 0.0852 Fish oil as supplement 0.1890 0.1520 0.1907 0.1907 0.0000 PGD1 8,9-EpETrE 8,9-EpETE 5,6-EpETE 17,18-EpETE PGD2 17-HDoHE 14,15-EpETE 13,14-EpDPE 8-iso PGF2a 0.1864 1.6411 0.0439 0.0439 0.0000 0.1864 0.0548 0.3936 0.3936 0.0000 0.1827 0.3411 0.5513 0.5513 0.0000 0.1757 0.5378 0.5247 0.5247 0.0000 0.1741 0.1415 0.2928 0.2928 0.0000 0.1708 0.7856 0.0938 0.0938 0.3356 0.1633 1.8552 0.0315 0.0315 0.0232 0.0709 0.1230 0.1603 0.1603 0.0546 0.0703 0.3595 0.0952 0.0952 0.0000 0.0664 0.3743 0.0816 0.0816 0.0000 157 0.0000 0.0000 0.1103 0.0000 0.0000 0.8106 0.0000 5.5666 0.0000 0.0000 0.0000 0.0000 0.0000 4.2007 0.0000 7.1383 0.3767 0.0000 0.0000 0.0000 0.0000 5.7862 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.8777 0.8616 0.8649 0.5714 0.9325 0.6322 0.4220 0.8360 0.2388 0.6719 0.6956 0.9542 0.9303 0.8844 0.6409 0.6777 0.2459 0.6754 0.5787 0.4493 0.4250 0.4316 0.2251 0.2910 0.0499 0.2770 0.2941 0.2552 0.2352 0.2975 0.0301 0.0461 0.0463 0.1590 0.2410 0.2620 0.0430 0.0745 0.0743 Table 3.4 (cont’d) 8,9-EpEDE/8,9-DHED RVD1 11,12-EpETrE 5,15-DiHETE PGD3 0.0209 0.7843 0.0574 0.0574 0.0000 0.0179 0.1522 0.0018 0.0018 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0053 0.0298 0.0000 0.0000 0.0000 Table 3.5. Linear regression Estimate Std. Error (Estimate/ std. error)= t value Pr(>|t|) (Intercept) Steroid 5-HETE 14,15-DiHETrE 15 S-HETrE 6.818923 2.30983 4.857866 0.848702 0.228252 0.252936 -0.05403 0.026563 -0.14901 0.458913 12,13-EpOME/12,13-DiHOME 0.005666 0.012571 9,10-EpOME/9,10-DiHOME -0.0212 0.025794 13-oxo-ODE Omega3index 9,10-DiHODE 0.051353 0.107485 23.36136 20.35532 -1.0447 0.445583 14,15-EpEDE/14,15-DHED -0.94939 0.476006 Age 0.038489 0.028379 9,10-EpODE/9,10-DiHODE 0.109444 0.038075 2.952 5.724 0.902 -2.034 -0.325 0.451 -0.822 0.478 1.148 -2.345 -1.994 1.356 2.874 0.0033 1.78E-08 0.36726 0.04246 0.74555 0.65235 0.41156 0.63302 0.25164 0.01943 0.04663 0.17561 0.00422 3.8.3. Exploring the impact of SLE on PUFA metabolic pathways: Mechanistic insights from epoxide to diol conversion We ran additional models to investigate the significant changes in PUFA pathways for three PUFAs: LA, DHA, and ALA. We did not build the DGLA model because the metabolites' measurements had zero variance. Models for EPA and AA were abandoned due to inadequate model fits. The second derivative metabolites (diol) are the outcome variables of the model, the first derivative (epoxide), the SLE status, and the interaction between the first derivative and the SLE status are input of model. Table 3.6. presents the statistical analysis of the pathway of LA, shedding light on the activities of the sEH in converting EpOMEs to DiHOMEs and how these activities are altered in the context 158 of SLE. The model fits adequately. The chi-square was 5.117 (df = 4, p = 0.2755). The RMSEA was 0.023. The CFA was 0.998. For the conversion of regioisomers 9,10-EpOME to 9,10- DiHOME, an estimated coefficient is 29.489 and a P-value of less than 0.001. This suggests that the enzyme is actively converting 9,10-EpOME to 9,10-DiHOME. However, when considering the interaction of this pathway with the presence of SLE, there is a significant reduction in this conversion process, indicated by a negative coefficient of -26.598 and a P-value of less than 0.001. (Table 3.6). Table 3.6. PUFA pathway of LA Coefficients 9,10-EpOME => 9,10-DiHOME Group => 9,10-DiHOME 9,10-EpOME * Group => OL9_10DiHOME 12,13-EpOME => 12,13-DiHOME Group => 12,13DiHOME 12,13-EpOME * Group => 12,13-DiHOME Est. 29.489 44.101 -26.598 5.161 -25.707 -5.615 se 3.337 25.794 3.574 0.535 13.417 0.807 9,10-DiHOME with 12,13-DiHOME 18311.307 1341.468 intercept (9,10-DiHOME) intercept (12,13-DiHOME) residual (9,10-DiHOME) residual (12,13-DiHOME) 24.305 83.749 39321.137 15638.444 22.199 10.634 2417.939 962.084 P-value 0.000 0.087 0.000 0.000 0.055 0.000 0.000 0.274 0.000 0.000 0.000 A similar pattern is observed for the pathway from 12,13-EpOME to 12,13-DiHOME, where the conversion is again positive and significant (coefficient of 5.161, P-value of less than 0.001). Yet, the interaction with SLE status shows a significant decrease (coefficient of -25.707, P-value of 0.055) (Table 3.6). Additionally, the direct interaction between 12,13-EpOME and the SLE group status was significantly negative (coefficient of -5.615, P-value of less than 0.001), further confirming the negative impact of SLE on this enzymatic activity. The high Comparative Fit Index (CFA) of 0.998, along with a non-significant Chi-square (P-value = 0.274) and a low RMSEA of 0.023 collectively suggest that the model adequately fits the observed data, indicating that these 159 findings are robust. The significant residuals for 9,10-DiHOME and 12,13-DiHOME (P-value < 0.001) imply that while the model captures key aspects of the pathway dynamics, additional unexplained variabilities exist. The results showed that the pathways from 9,10-EpOME to 9,10- DiHOME and from 12,13-EpOME to 12,13-DiHOME significantly differed between normal and SLE subjects. Both pathways were reduced in SLE subjects. Table 3.7. provides a summary of the ML of the ALA pathway, with a focus on activities of sEH in converting epoxides (EpODEs) to diols (DiHODEs) and how these activities are changed in the context of individuals with SLE. This model also fits adequately; the chi-square was 14.804 (df = 6, p =0.0218). The RMSEA was 0.053, and the CFA was 0.992. The analysis reveals that under standard conditions, sEH effectively catalyzes the conversion from 9,10-EpODE to 9,10- DiHOME, as indicated by a significant positive estimate (Est. = 1.205, P-value = 0.021). However, the presence of SLE appears to inhibit this enzymatic function, as suggested by the negative interaction between the SLE group and the 9,10-EpODE to 9,10-DiHODE conversion (Est. = - 1.061, P-value = 0.040). A similar pattern is observed with the conversion from 15,16-EpODE to 15,16-DiHODE. While the enzyme shows high activity in this conversion (Est. = 2.035, P-value < 0.001), SLE again diminishes this activity, evidenced by the negative interaction term (Est. = - 0.573, P-value < 0.001). The results of DHA showed that the PUFA pathways of DHA were not significantly different between the healthy and SLE subjects (Table S7). 160 Table 3.7. PUFA pathway of ALA Coefficients 9,10-EpODE => 9,10-DiHODE 12,13-EpODE => 9,10-DiHODE 15,16-EpODE => 9,10-DiHODE Group => 9,10-DiHODE 9,10-EpODE * Group => 9,10-DiHODE 12,13-EpODE => 12,13-DiHODE 9,10-EpODE => 12,13-DiHODE 15,16-EpODE => 12,13-DiHODE Group =>12,13-DiHODE Group * 12,13-EpODE => 12,13-DiHODE 15,16-EpODE => 15,16-DiHODE 9,10-EpODE => 15,16-DiHODE 12,13-EpODE=> 15,16-DiHODE Group => 15,16-DiHODE 15,16-EpODE * Group => 15,16-DiHODE 9,10-DiHODE with 12,13-DiHODE 9,10-DiHODE with 15,16-DiHODE 12,13-DiHODE with 15,16-DiHODE intercept (9,10-DiHODE) intercept (12,13-DiHODE) intercept (15,16-DiHODE) residual (9,10-DiHODE) residual (12,13-DiHODE) residual (15,16-DiHODE) Est. 1.205 0.035 0.074 -0.141 -1.061 0.079 -0.164 0.037 0.100 -0.036 2.035 -3.107 0.130 2.011 -0.573 0.038 1.325 0.323 0.395 0.082 2.109 0.473 0.107 34.646 se 0.522 0.013 0.010 0.081 0.516 0.007 0.071 0.005 0.037 0.009 0.113 1.219 0.112 0.688 0.126 0.010 0.186 0.085 0.068 0.030 0.553 0.029 0.007 2.129 P-value 0.021 0.007 0.000 0.080 0.040 0.000 0.020 0.000 0.007 0.000 0.000 0.011 0.249 0.003 0.000 0.000 0.000 0.000 0.000 0.006 0.000 0.000 0.000 0.000 6. Discussion SLE is an autoimmune disease, and its clinical heterogeneity leads to diagnostic challenges. Key factors in SLE pathophysiology include autoantibodies, pro-inflammatory cytokines, and the impact of sex hormones. Recent studies have shown that omega-3 fatty acid supplements could improve the quality of life of SLE patients. Such intervention may be more critical in the Western population because the consumption of omega-6 PUFAs, including LA and AA, which are generally considered pro-inflammatory, has been substantially increased in the Western diet due 161 to the use of vegetable oil, notably in corn and soybean oils over the past century44,11,12. The mechanism of how omega-3 PUFA supplement improves SLE pathogenesis remains unknown. Recent studies suggested that the oxidized metabolites, called oxylipins, from the AA cascade, are key metabolites for many mammalian physiologies, especially inflammation13. For example, DHA is metabolized to specialized pro-resolving mediators (SPMs) essential for quelling inflammation, including resolvins, protectins, and anti-inflammatory epoxide metabolites. These SPMs play a significant role in suppressing inflammatory signals and enhancing the clearance of cell debris, which are critical for controlling chronic inflammation45,46. In contrast, some cells, such as macrophages, produce LA-derived epoxides, commonly called leukotoxins. Leukotoxins are important in lupus because they can exacerbate inflammation and immune responses, contributing to disease progression47. Both epoxy fatty acids are further metabolized to pro-inflammatory dihydroxy fatty acids by the EH enzyme13. This study investigated the impact of SLE pathogenesis on oxylipin metabolism and aimed to identify potential lipid biomarkers for SLE. Our initial analysis of the plasma levels of oxylipins in healthy subjects and those with SLE revealed significant differences in the plasma levels of several specific oxylipins, especially oxylipins generated by CYP-EH metabolism. Specifically, 9,10-DiHOME and 14,15-DiHETrE exhibited a notable decrease in concentration in individuals with SLE compared to healthy individuals. We further studied the importance of these metabolites in the prediction and severity of diseases using ML; each variable (metabolites, demographic information, and other factors) has a score that reflects its relative contribution to the model's ability to classify individuals correctly as binary output with or without SLE. Interestingly, our ML analysis showed the current use of steroids (56.8947) has a high importance score, reflecting its relevance in distinguishing between individuals with and without lupus. Steroid use is common 162 in treating lupus to control inflammation, which may explain its strong association with lupus in our model. Discovering that steroids are among the most significant variables affirms the reliability of our ML model, suggesting we can have confidence in the other variables identified by the same models. Our ML model identified that 14,15-DiHETrE (91.9489) and AA (59.9277) have the highest importance score, suggesting they have the strongest association with the subject of lupus in the western population. Since 14,15-DiHETrE is produced by CYP-EH metabolism, we then focus our analysis on epoxy and dihydroxy fatty acids. The dihydroxy fatty acids metabolites have higher importance scores than the epoxides in our dataset. This suggests that dihydroxy fatty acid metabolites are strongly associated with the disease status or its severity in the context of lupus. The strong association between certain diol metabolites and lupus in our ML model suggests they could potentially serve as biomarkers for the disease or indicators of disease severity. Such findings align with previous rodent studies that have demonstrated the importance of this CYP- EH metabolic pathway in SLE pathogenesis28,48. We, thus, investigated whether this pathway was impacted by SLE pathogenesis. Epoxy fatty acids (substrate of EH)/ dihydroxy fatty acids (product of EH) ratio has been used to evaluate whether this EH metabolism is impacted by diseases; therefore, these ratios were included in our analyses. In this study, we found that SLE patients consistently exhibited higher ratios of several epoxies to dihydroxy and lower levels of dihydroxy fatty acids, indicating a decreased EH activity or expression in these SLE patients. For example, we found that 9,10-EpODE/9,10-DiHODE and 9,10-EpODE/9,10-DiHODE are the second and third most important variables in the classification model with respective important scores of 88.0904 and 79.0008. Moreover, for the regression ML models, we observed 12,13-EpOME /12,13-DiHOME and 9,10- EpOME/9,10DiHOME appeared as the top five most important variables with respective scores 163 of 2.2491 and 2.2363. The importance of those ratios in the ML model confirms what we mentioned before about epoxy and dihydroxy fatty acids as key lipid signaling molecules for inflammation, which also play an important role in SLE pathogenesis. To better study the role of EH, we also conducted mechanistic statistical models to examine the metabolism of PUFAs across LA, ALA, and DHA pathways. That study has provided compelling insights into the metabolic alterations associated with SLE. These pathway models reveal that for LA and ALA metabolites (but not DHA), SLE subjects exhibit significantly reduced activity in sEH metabolism in converting epoxy metabolites to diol. The results from these models underscore the importance of EH in lipid metabolism and highlight its potential role in the pathophysiology of SLE. These findings could lead to new avenues for research and therapy, offering hope for better management and treatment of SLE through targeted intervention in PUFA metabolism. The Omega-3 Index (O3I) is a critical biomarker for assessing omega-3 PUFA status, with a target level of at least 8% suggested for cardiovascular health and longevity benefits; it is the ratio of the combined concentration of EPA and DHA over the sum of all PUFAs (AA, DGLA, LA, EPA, DHA, ALA)49,50. However, typical O3I levels in the U.S. are low, ranging from 4 to 6%51, and the mean O3I in our cohort was 5.8%. Omega-3 PUFAs such as ALA, found in plant oils and nuts, have a limited conversion efficiency in humans to longer-chain omega-3 PUFAs like EPA and DHA, which are abundant in fatty fish52, yet only 20% of Americans meet the recommended intake of two fish servings per week53, and this is still generally insufficient to achieve the "desirable" O3I levels54,55. In general, omega-3 PUFAs possess anti-inflammatory properties that have a physiological impact on lupus disease. For example, research has shown that diets enriched with DHA from fish oil substantially extended the median (658 days) and maximum (848 days) lifespans of lupus-prone NZBWF1 mice. Conversely, mice consuming EPA-enriched fish oil diets 164 exhibited median and maximum lifespans of around 384 and 500 days, respectively56 48, while omega-6 fatty acids could potentially exacerbate inflammation57,58. Interestingly, our analysis revealed that O3I ranked as the eighth important variable in the regression model (Table 4) with a relatively strong association with the severity of SLE (score of 2.0799). To further dissect the relationship between different PUFA and SLE pathogenesis, we also analyze total plasma levels of omega-3 and omega-6 fatty acids. Our ML model also indicated that the total omega-3 and omega-6 fatty acid levels are important to predict whether subjects have SLE, but their ability to predict the severity of SLE is significantly diminished, as indicated by their respective scores. The Score of 1.2743 for the total omega-6 fatty acid level and 1.1688 for the total omega-3 fatty acid level rank top 40 among all the factors we have examined. The O3I and total plasma level of omega-3 and omega-6 fatty acids could be affected by either the endogenous fatty acid biosynthesis or dietary supplements. Therefore, we also analyze the relative importance of omega- 3 fatty acid supplements in predicting the severity score of SLE. Unfortunately, based on our analysis, supplementation with fish oil (rich in DHA), flax oil (rich in ALA), and a combined regimen of both oils have relatively low scores, placing them in the bottom 20 of the importance ranking, suggesting that their supplementation may not have strong association with severity of SLE. However, O3I and total plasma levels of omega-3 and omega-6 FA are highly important in predicting whether subjects have SLE, and they may indicate that the biosynthesis of omega-3 and omega-6 fatty acids is altered by the SLE pathogenesis and should warrant further investigation in the future. In addition, it could affect the basal omega-3 and omega-6 FA levels, which could impact the efficacy of omega-3 fatty acid supplements. Lupus overwhelmingly affects women, with females accounting for around 90% of all cases. While the reasons for this stark gender disparity are not fully understood, sex plays a significant 165 role in the development and progression of lupus59. Compared to women, men with lupus tend to have a later age of disease onset60 ; more severe manifestations like increased risk of cardiovascular disease61 plus lupus nephritis may be more severe in men62. The sex-based differences in lupus highlight the need for tailored approaches to prevention, management, and support for patients of all genders. We found some metabolites, like 12,13-DiHOME, 19,20-DiHDPE, 15-HETE, 11- HETE, 9-HODE, 13-HODE, 17,18- DiHETE, 15-HEPE, and PGD2 in female with lupus demonstrated a significantly higher concentration than the male with lupus. Regarding both ML models, sex located in the middle or bottom half of the ranking does not show a strong association with the status and severity of the diseases. Lupus disease severity and outcomes can vary significantly depending on the patient's age at disease onset. Juvenile-onset lupus is associated with a more severe clinical presentation, including higher rates of nephritis and neuropsychiatric disease compared to adult-onset cases63. Mortality rates are also substantially higher in those diagnosed with lupus at a younger age64. In contrast, lupus that develops in older adults over 50 is typically milder, with less frequent kidney and neurological issues, though these patients may face other age-related challenges like increased organ damage, weight gain, and decreased kidney function63. Despite the differences in active disease manifestations, older lupus patients still experience a lower quality of life, potentially due to the compounding effects of accumulated disease damage over time. In our classification ML study, age is among the top five most important variables, with an overall score of 77.6358, which shows great importance in predicting diseases in the Western population. Age is also strongly associated with disease severity based on the regression ML models with an overall score of 1.8892. Besides ML models, we also studied the effect of SLE in the concentration of different metabolites based on their age group, and we found compound 9-HODE demonstrated a significant 166 concentration decrease in the young lupus category compared to the middle-aged and old categories. The conclusion of the study is that age is a strong predictor of disease severity in SLE according to regression ML models, with age-related differences in metabolite concentrations further highlighting the complex interaction between age and disease impact. Lupus is a disease that exhibits significant racial and ethnic disparities, with incidence and prevalence rates that are substantially higher in minority populations compared to White individuals. Black, South Asian, and Native American/Indigenous lupus patients tend to experience a more aggressive disease course with increased nephritis, neuropsychiatric involvement, accumulated organ damage, and higher mortality65,66. The ML study showed the importance of ethnicity and lupus diseases. In the prediction of disease between races, Although the White population has higher importance (overall score of 14.8031) than the Black (overall score of 13.3696), the change in importance score is not noticeable (less than 10% difference) moreover, in both ML models related to statue and severity of disease, race, showed less importance rank as last few factors in the table. These variables indicate that race does not have a relatively important role in the model's predictions of diseases. Based on metabolite concentration, 14,15-DiHETrE showed a decrease in concentration in lupus vs healthy people for both the Black and White categories; for other compounds like 14,15-DiHETE and 13,14-DiHDPE, we observed lupus persons of White ethnicities have a higher concentration in comparison to lupus individual with Black ethnicities. The observed variation in metabolite concentrations, notably 14,15- DiHETrE, 14,15-DiHETE, and 13,14-DiHDPE, between lupus patients of different ethnicities raises important questions about the metabolic pathways involved in the disease's pathogenesis and progression. 14,15-DiHETrE's decrease in both Black and White lupus patients compared to healthy individuals suggests a potentially diminished role in these regulatory processes, which 167 might affect the disease's severity and manifestations. Furthermore, 14,15-DiHETrE is recognized as a potent peroxisome PPAR alpha and gamma activator. These receptors are integral in lipid metabolism and energy homeostasis, with PPAR alpha primarily expressed in the heart, liver, and kidney and PPAR gamma predominantly in adipose tissue. Given these interactions, the modulation of 14,15-DiHETrE in lupus individuals could reflect adaptive or pathogenic responses to the chronic inflammatory state induced by the disease. Furthermore, the elevated levels of 14,15-DiHETE and 13,14-DiHDPE observed in White lupus patients as opposed to Black patients suggest a possible ethnic variation in lipid metabolism or enzymatic activity related to these metabolites.14,15-DiHETE is known for its involvement in inflammatory modulation24. Higher concentrations in one ethnic group may indicate a differential response to inflammation or a variant regulatory mechanism in lipid pathways. Acknowledging limitations, such as the uneven sample distribution among demographic groups, the lack of detailed information on specific omega-3 and omega-6 PUFAs in diets, and details on lupus flare-like pain and favor, the study underscores the need for future research. 7. Conclusion Our study provides compelling evidence of the intricate relationship between dietary PUFAs, oxylipins metabolites of PUFA, and SLE. It underscores the potential of dietary intervention in modulating disease activity and offers insight into potential biomarker targets. We have demonstrated that oxylipins derived from omega-6 and omega-3 PUFA exhibit significant variations in concentration across different demographic groups, suggesting that sex (12,13- DiHOME, 19,20-DiHDPE, 15-HETE), race (14,15-DiHETE and 13,14-DiHDPE), and age (9- HODE) can influence PUFA metabolism and, consequently, inflammation and the immune response in SLE. Our use of ML models has further elucidated the complex interactions between 168 demographic factors, dietary PUFA, and their association with the status and severity of lupus. The importance scores from the ML model indicate that metabolites like 14,15-DiHETrE and 17,18-DiHETE, current steroid use, and the balance of dietary omega fatty acids could be integral to understanding and potentially managing SLE. During the mechanistic study and based on the changes in the ratio of Epoxy to Dihydroxy metabolites, the CYP-sEH pathway was identified as a significant influencer in lupus disease. Our results have highlighted specific oxylipins with potential as biomarkers for SLE, which could lead to improved diagnostic tools and treatment options tailored to individual patient profiles. While our findings are suggestive, they require validation through additional biological studies and clinical trials. Supporting Information Supporting Information, including experimental methods and materials, supplemental figures, and table, is available in APPENDIX 2. 169 BIBLIOGRAPHY (1) Tsokos, G. C.; Lo, M. S.; Costa Reis, P.; Sullivan, K. E. New Insights into the Immunopathogenesis of Systemic Lupus Erythematosus. Nat. Rev. Rheumatol. 2016, 12 (12), 716–730. (2) Mak, A.; Isenberg, D. A.; Lau, C.-S. Global Trends, Potential Mechanisms and Early Detection of Organ Damage in SLE. Nat. Rev. Rheumatol. 2013, 9 (5), 301–310. (3) Murimi-Worstell, I. B.; Lin, D. H.; Nab, H.; Kan, H. J.; Onasanya, O.; Tierce, J. C.; Wang, X.; Desta, B.; Alexander, G. C.; Hammond, E. R. Association between Organ Damage and Mortality in Systemic Lupus Erythematosus: A Systematic Review and Meta-Analysis. BMJ Open 2020, 10 (5), e031850. (4) Aringer, M. Inflammatory Markers in Systemic Lupus Erythematosus. J. Autoimmun. 2020, 110 (102374), 102374. (5) Rönnelid, J.; Tejde, A.; Mathsson, L.; Nilsson-Ekdahl, K.; Nilsson, B. Immune Complexes from SLE Sera Induce IL10 Production from Normal Peripheral Blood Mononuclear Cells by an FcgammaRII Dependent Mechanism: Implications for a Possible Vicious Cycle Maintaining B Cell Hyperactivity in SLE. Ann. Rheum. Dis. 2003, 62 (1), 37–42. (6) Liu, C.-C.; Kao, A. H.; Manzi, S.; Ahearn, J. M. Biomarkers in Systemic Lupus Erythematosus: Challenges and Prospects for the Future. Ther. Adv. Musculoskelet. Dis. 2013, 5 (4), 210–233. (7) Bader-Meunier, B.; Quartier, P.; Deschênes, G.; Cochat, P.; Haddad, E.; Koné-Paut, I.; Leblanc, T.; Prieur, A. M.; Salomon, R.; Bodemer, C.; Lévy, M.; Group française d’étude du lupus érythémateux disséminé de l’enfant. Childhood-onset systemic lupus erythematosus. Arch. Pediatr. 2003, 10 (2), 147–157. (8) Beeson, P. B. Age and Sex Associations of 40 Autoimmune Diseases. Am. J. Med. 1994, 96 (5), 457–462. (9) Cohen-Solal, J. F. G.; Jeganathan, V.; Grimaldi, C. M.; Peeva, E.; Diamond, B. Sex Hormones and SLE: Influencing the Fate of Autoreactive B Cells. Curr. Top. Microbiol. Immunol. 2006, 305, 67–88. (10) Grimaldi, C. M. Sex and Systemic Lupus Erythematosus: The Role of the Sex Hormones Estrogen and Prolactin on the Regulation of Autoreactive B Cells. Curr. Opin. Rheumatol. 2006, 18 (5), 456–461. (11) Kris-Etherton, P. M.; Taylor, D. S.; Yu-Poth, S.; Huth, P.; Moriarty, K.; Fishell, V.; Hargrove, R. L.; Zhao, G.; Etherton, T. D. Polyunsaturated Fatty Acids in the Food Chain in the United States. Am. J. Clin. Nutr. 2000, 71 (1), 179S-188S. (12) Simopoulos, A. P. The Importance of the Omega-6/Omega-3 Fatty Acid Ratio in Cardiovascular Disease and Other Chronic Diseases. Exp. Biol. Med. (Maywood) 2008, 233 (6), 674–688. 170 (13) Morisseau, C.; Hammock, B. D. Impact of Soluble Epoxide Hydrolase and Epoxyeicosanoids on Human Health. Annu. Rev. Pharmacol. Toxicol. 2013, 53 (1), 37–58. (14) Bello, K. J.; Fang, H.; Fazeli, P.; Bolad, W.; Corretti, M.; Magder, L. S.; Petri, M. Omega-3 in SLE: A Double-Blind, Placebo-Controlled Randomized Clinical Trial of Endothelial Dysfunction and Disease Activity in Systemic Lupus Erythematosus. Rheumatol. Int. 2013, 33 (11), 2789–2796. (15) Clark, W. F.; Parbtani, A.; Huff, M. W.; Reid, B.; Holub, B. J.; Falardeau, P. Omega-3 Fatty Acid Dietary Supplementation in Systemic Lupus Erythematosus. Kidney Int. 1989, 36 (4), 653– 660. (16) Westberg, G.; Tarkowski, A. Effect of MaxEPA in Patients with SLE. A Double-Blind, Crossover Study. Scand. J. Rheumatol. 1990, 19 (2), 137–143. (17) Das, U. N. Beneficial Effect of Eicosapentaenoic and Docosahexaenoic Acids in the Management of Systemic Lupus Erythematosus and Its Relationship to the Cytokine Network. Prostaglandins Leukot. Essent. Fatty Acids 1994, 51 (3), 207–213. (18) Duffy, E. M.; Meenagh, G. K.; McMillan, S. A.; Strain, J. J.; Hannigan, B. M.; Bell, A. L. The Clinical Effect of Dietary Supplementation with Omega-3 Fish Oils and/or Copper in Systemic Lupus Erythematosus. J. Rheumatol. 2004, 31 (8), 1551–1556. (19) Elkan, A.-C.; Anania, C.; Gustafsson, T.; Jogestrand, T.; Hafström, I.; Frostegård, J. Diet and Fatty Acid Pattern among Patients with SLE: Associations with Disease Activity, Blood Lipids and Atherosclerosis. Lupus 2012, 21 (13), 1405–1411. (20) Arriens, C.; Hynan, L. S.; Lerman, R. H.; Karp, D. R.; Mohan, C. Placebo-Controlled Randomized Clinical Trial of Fish Oil’s Impact on Fatigue, Quality of Life, and Disease Activity in Systemic Lupus Erythematosus. Nutr. J. 2015, 14 (1), 82. (21) Kortz, L.; Dorow, J.; Ceglarek, U. Liquid Chromatography-Tandem Mass Spectrometry for the Analysis of Eicosanoids and Related Lipids in Human Biological Matrices: A Review. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2014, 964, 1–11. (22) Gladine, C.; Ostermann, A. I.; Newman, J. W.; Schebb, N. H. MS-Based Targeted Metabolomics of Eicosanoids and Other Oxylipins: Analytical and Inter-Individual Variabilities. Free Radic. Biol. Med. 2019, 144, 72–89. (23) Palla, A. R.; Ravichandran, M.; Wang, Y. X.; Alexandrova, L.; Yang, A. V.; Kraft, P.; Holbrook, C. A.; Schürch, C. M.; Ho, A. T. V.; Blau, H. M. Inhibition of Prostaglandin-Degrading Enzyme 15-PGDH Rejuvenates Aged Muscle Mass and Strength. Science 2021, 371 (6528), eabc8059. (24) Wang, B.; Wu, L.; Chen, J.; Dong, L.; Chen, C.; Wen, Z.; Hu, J.; Fleming, I.; Wang, D. W. Metabolism Pathways of Arachidonic Acids: Mechanisms and Potential Therapeutic Targets. Signal Transduct. Target. Ther. 2021, 6 (1), 94. 171 (25) Dennis, E. A.; Norris, P. C. Eicosanoid Storm in Infection and Inflammation. Nat. Rev. Immunol. 2015, 15 (8), 511–523. (26) Fullerton, J. N.; O’Brien, A. J.; Gilroy, D. W. Lipid Mediators in Immune Dysfunction after Severe Inflammation. Trends Immunol. 2014, 35 (1), 12–21. (27) He, J.; Ma, C.; Tang, D.; Zhong, S.; Yuan, X.; Zheng, F.; Zeng, Z.; Chen, Y.; Liu, D.; Hong, X.; Dai, W.; Yin, L.; Dai, Y. Absolute Quantification and Characterization of Oxylipins in Lupus Nephritis and Systemic Lupus Erythematosus. Front. Immunol. 2022, 13, 964901. (28) Hye Khan, M. A.; Stavniichuk, A.; Sattar, M. A.; Falck, J. R.; Imig, J. D. Epoxyeicosatrienoic Acid Analog EET-A Blunts Development of Lupus Nephritis in Mice. Front. Pharmacol. 2019, 10, 512. (29) Patrono, C.; Ciabattoni, G.; Remuzzi, G.; Gotti, E.; Bombardieri, S.; Di Munno, O.; Tartarelli, G.; Cinotti, G. A.; Simonetti, B. M.; Pierucci, A. Functional Significance of Renal Prostacyclin and Thromboxane A2 Production in Patients with Systemic Lupus Erythematosus. J. Clin. Invest. 1985, 76 (3), 1011–1018. (30) He, J.; Chan, T.; Hong, X.; Zheng, F.; Zhu, C.; Yin, L.; Dai, W.; Tang, D.; Liu, D.; Dai, Y. Microbiome and Metabolome Analyses Reveal the Disruption of Lipid Metabolism in Systemic Lupus Erythematosus. Front. Immunol. 2020, 11, 1703. (31) Korotkova, M.; Jakobsson, P.-J. Persisting Eicosanoid Pathways in Rheumatic Diseases. Nat. Rev. Rheumatol. 2014, 10 (4), 229–241. (32) https://wwwn.cdc.gov/nchs/nhanes/. (accessed 2024-05-13). questionnaires, NHANES datasets, and related documentation. (33) Karlson, E. W.; Daltroy, L. H.; Rivest, C.; Ramsey-Goldman, R.; Wright, E. A.; Partridge, A. J.; Liang, M. H.; Fortin, P. R. Validation of a Systemic Lupus Activity Questionnaire (SLAQ) for Population Studies. Lupus 2003, 12 (4), 280–286. (34) Masood, A.; Stark, K. D.; Salem, N., Jr. A Simplified and Efficient Method for the Analysis of Fatty Acid Methyl Esters Suitable for Large Clinical Studies. J. Lipid Res. 2005, 46 (10), 2299– 2305. (35) Lepage, G.; Roy, C. C. Direct Transesterification of All Classes of Lipids in a One-Step Reaction. J. Lipid Res. 1986, 27 (1), 114–120. (36) Ziegler, A. An Introduction to Statistical Learning with Applications. R. g.James, D.Witten, T.Hastie, and R.Tibshirani (2013). Berlin: Springer. 440 Pages, ISBN: 978‐1‐4614‐7138‐7. Biom. J. 2016, 58 (3), 715–716. (37) Cortes, C.; Vapnik, V. Support-Vector Networks. Mach. Learn. 1995, 20 (3), 273–297. (38) Friedman, J. H. Stochastic Gradient Boosting. Comput. Stat. Data Anal. 2002, 38 (4), 367– 378. 172 (39) Friedman, J.; Hastie, T.; Tibshirani, R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J. Stat. Softw. 2010, 33 (1), 1–22. (40) Wright, M. N.; Ziegler, A. Ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. J. Stat. Softw. 2017, 77 (1). https://doi.org/10.18637/jss.v077.i01. (41) Venables, W. N.; Ripley, B. D. Modern Applied Statistics with S, 4th ed.; Statistics and Computing; Springer: New York, NY, 2002. (42) Nearest Neighbor Pattern Classification" (PDF). IEEE Transactions on Information Theory. (43) Chen, T.; Guestrin, C. XGBoost. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining; ACM: New York, NY, USA, 2016. https://doi.org/10.1145/2939672.2939785. (44) Simopoulos, A. P. Evolutionary Aspects of Diet, the Omega-6/Omega-3 Ratio and Genetic Variation: Nutritional Implications for Chronic Diseases. Biomed. Pharmacother. 2006, 60 (9), 502–507. (45) Serhan, C. N. Systems Approach to Inflammation Resolution: Identification of Novel Anti- Inflammatory and pro-Resolving Mediators. J. Thromb. Haemost. 2009, 7 Suppl 1, 44–48. (46) Fredman, G.; Hellmann, J.; Proto, J. D.; Kuriakose, G.; Colas, R. A.; Dorweiler, B.; Connolly, E. S.; Solomon, R.; Jones, D. M.; Heyer, E. J.; Spite, M.; Tabas, I. An Imbalance between Specialized Pro-Resolving Lipid Mediators and pro-Inflammatory Leukotrienes Promotes Instability of Atherosclerotic Plaques. Nat. Commun. 2016, 7 (1), 12859. (47) Ozawa, T.; Sugiyama, S.; Hayakawa, M.; Taki, F.; Hanaki, Y. Neutrophil Microsomes Biosynthesize Linoleate Epoxide (9,10-Epoxy-12-Octadecenoate), a Biological Active Substance. Biochem. Biophys. Res. Commun. 1988, 152 (3), 1310–1318. (48) Favor, O. K.; Chauhan, P. S.; Pourmand, E.; Edwards, A. M.; Wagner, J. G.; Lewandowski, R. P.; Heine, L. K.; Harkema, J. R.; Lee, K. S. S.; Pestka, J. J. Lipidome Modulation by Dietary Omega-3 Polyunsaturated Fatty Acid Supplementation or Selective Soluble Epoxide Hydrolase Inhibition Suppresses Rough LPS-Accelerated Glomerulonephritis in Lupus-Prone Mice. Front. Immunol. 2023, 14, 1124910. (49) Harris, W. S. The Omega-3 Index: From Biomarker to Risk Marker to Risk Factor. Curr. Atheroscler. Rep. 2009, 11 (6), 411–417. (50) Schuchardt, J. P.; Cerrato, M.; Ceseri, M.; DeFina, L. F.; Delgado, G. E.; Gellert, S.; Hahn, A.; Howard, B. V.; Kadota, A.; Kleber, M. E.; Latini, R.; Maerz, W.; Manson, J. E.; Mora, S.; Park, Y.; Sala-Vila, A.; von Schacky, C.; Sekikawa, A.; Tintle, N.; Tucker, K. L.; Vasan, R. S.; Harris, W. S. Red Blood Cell Fatty Acid Patterns from 7 Countries: Focus on the Omega-3 Index. Prostaglandins Leukot. Essent. Fatty Acids 2022, 179 (102418), 102418. 173 (51) Stark, K. D.; Van Elswyk, M. E.; Higgins, M. R.; Weatherford, C. A.; Salem, N., Jr. Global Survey of the Omega-3 Fatty Acids, Docosahexaenoic Acid and Eicosapentaenoic Acid in the Blood Stream of Healthy Adults. Prog. Lipid Res. 2016, 63, 132–152. (52) Baker, E. J.; Miles, E. A.; Burdge, G. C.; Yaqoob, P.; Calder, P. C. Metabolism and Functional Effects of Plant-Derived Omega-3 Fatty Acids in Humans. Prog. Lipid Res. 2016, 64, 30–56. (53) Terry, A. L.; Herrick, K. A.; Afful, J.; Ahluwalia, N. Seafood Consumption in the United States, 2013-2016. NCHS Data Brief 2018, No. 321, 1–8. (54) McDonnell, S. L.; French, C. B.; Baggerly, C. A.; Harris, W. S. Cross-Sectional Study of the Combined Associations of Dietary and Supplemental Eicosapentaenoic Acid + Docosahexaenoic Acid on Omega-3 Index. Nutr. Res. 2019, 71, 43–55. (55) Jackson, K. H.; Polreis, J. M.; Tintle, N. L.; Kris-Etherton, P. M.; Harris, W. S. Association of Reported Fish Intake and Supplementation Status with the Omega-3 Index. Prostaglandins Leukot. Essent. Fatty Acids 2019, 142, 4–10. (56) Halade, G. V.; Rahman, M. M.; Bhattacharya, A.; Barnes, J. L.; Chandrasekar, B.; Fernandes, G. Docosahexaenoic Acid-Enriched Fish Oil Attenuates Kidney Disease and Prolongs Median and Maximal Life Span of Autoimmune Lupus-Prone Mice. J. Immunol. 2010, 184 (9), 5280–5286. (57) Innes, J. K.; Calder, P. C. Omega-6 Fatty Acids and Inflammation. Prostaglandins Leukot. Essent. Fatty Acids 2018, 132, 41–48. (58) DiNicolantonio, J. J.; O’Keefe, J. The Importance of Maintaining a Low Omega-6/Omega-3 Ratio for Reducing the Risk of Autoimmune Diseases, Asthma, and Allergies. Mo. Med. 2021, 118 (5), 453–459. (59) Barbhaiya, M.; Costenbader, K. H. Environmental Exposures and the Development of Systemic Lupus Erythematosus. Curr. Opin. Rheumatol. 2016, 28 (5), 497–505. (60) Wallace, D. J.; Hahn, B. H. Dubois’ Lupus Erythematosus; Lippincott Williams & Wilkins: Philadelphia, 2002. (61) Urowitz, M. B.; Gladman, D.; Ibañez, D.; Bae, S. C.; Sanchez-Guerrero, J.; Gordon, C.; Clarke, A.; Bernatsky, S.; Fortin, P. R.; Hanly, J. G.; Wallace, D. J.; Isenberg, D.; Rahman, A.; Alarcón, G. S.; Merrill, J. T.; Ginzler, E.; Khamashta, M.; Nived, O.; Sturfelt, G.; Bruce, I. N.; Steinsson, K.; Manzi, S.; Ramsey-Goldman, R.; Dooley, M. A.; Zoma, A.; Kalunian, K.; Ramos, M.; Van Vollenhoven, R. F.; Aranow, C.; Stoll, T.; Petri, M.; Maddison, P.; Systemic Lupus International Collaborating Clinics. Atherosclerotic Vascular Events in a Multinational Inception Cohort of Systemic Lupus Erythematosus. Arthritis Care Res. (Hoboken) 2010, 62 (6), 881–887. (62) Resende, A. L.; Titan, S. M.; Barros, R. T.; Woronik, V. Worse Renal Outcome of Lupus Nephritis in Male Patients: A Case-Control Study. Lupus 2011, 20 (6), 561–567. (63) Lythgoe, H.; Lj, M.; Hedrich, C. M.; Aringer, M. Classification of Systemic Lupus Erythematosus in Children and Adults. Clin. Immunol. 2022, 234 (108898), 108898. 174 (64) Arnaud, L.; Tektonidou, M. G. Long-Term Outcomes in Systemic Lupus Erythematosus: Trends over Time and Major Contributors. Rheumatology (Oxford) 2020, 59 (Suppl5), v29–v38. (65) Buie, J.; McMillan, E.; Kirby, J.; Cardenas, L. A.; Eftekhari, S.; Feldman, C. H.; Gawuga, C.; Knight, A. M.; Lim, S. S.; McCalla, S.; McClamb, D.; Polk, B.; Williams, E.; Yelin, E.; Shah, S.; Costenbader, K. H. Disparities in Lupus and the Role of Social Determinants of Health: Current State of Knowledge and Directions for Future Research. ACR Open Rheumatol. 2023, 5 (9), 454– 464. (66) Lewis, M. J.; Jawad, A. S. The Effect of Ethnicity and Genetic Ancestry on the Epidemiology, Clinical Features and Outcome of Systemic Lupus Erythematosus. Rheumatology (Oxford) 2016, kew399. 175 CHAPTER 4. INVESTIGATING THE EFFECT OF TBHP-INDUCED OXIDATIVE STRESS ON OXYLIPIN AND LIPID PROFILE USING TRIPLE QUADRUPOLE AND HR-MS 176 ABSTRACT Oxidative stress, resulting from an imbalance between the production of reactive oxygen species (ROS) and the biological system's ability to neutralize or repair the resulting damage, is a critical factor in the pathogenesis of various diseases, including chronic inflammatory conditions and neurodegenerative disorders such as Alzheimer's, Parkinson's, and Huntington's disease. Lipid peroxidation, particularly of polyunsaturated fatty acids (PUFAs), compromises membrane integrity and is directly related to oxidative stress as it involves the oxidative degradation of lipids by ROS, leading to cellular damage and the progression of various diseases. Tracking lipid biomarkers under oxidative stress offers insights into disease mechanisms and potential therapeutic targets. This study investigates the impact of oxidative stress induced by tert-butyl hydroperoxide (tBHP) on lipid and oxylipin profiles using the aging model organism Caenorhabditis elegans (C. elegans). Our methodologies include targeted LC-MS/MS and untargeted HR-MS analysis to monitor changes in oxylipin and lipid concentrations. Our findings reveal substantial alterations in oxylipin profiles and lipid concentrations, indicating a complex interplay between lipid metabolism and oxidative stress responses. The upregulation of dihydroxy fatty acid metabolites, known pro-inflammatory agents, alongside unchanged CYP 450 metabolites, suggests a possible enzyme regulation under oxidative stress. These changes imply an adaptive mechanism to maintain cellular homeostasis. This study underscores the necessity of comprehensive lipidomic analyses using both positive and negative ESI modes and more advanced analytical instrumentation like ion mobility spectrometry (IMS) to understand and fully characterize lipid behaviors under oxidative stress. Such approaches lay the groundwork for future investigations into lipid metabolism and oxidative stress-related diseases, potentially identifying new therapeutic targets and diagnostic biomarkers. 177 1. Introduction Oxidative stress happens when there is an imbalance between the production of ROS and the biological systems' capacity to neutralize or repair the oxidative damage. Oxidative stress has been implicated in the pathogenesis of a variety of diseases, including cardiovascular diseases1, cancer2, chronic inflammatory diseases like lupus3, and neurodegenerative disorders such as Alzheimer’s disease (AD), Parkinson's disease (PD), Huntington's disease (HD), amyotrophic lateral sclerosis (ALS), and spinocerebellar ataxia (SCA)4–6. Despite making up just 2% of the human body's weight, the brain uses 20% of its basal oxygen consumption due to its high metabolic rate. The brain is highly susceptible to oxidative damage due to physiological, anatomical, and functional factors7. In AD, oxidative stress is thought to contribute to the formation of amyloid-beta plaques and neurofibrillary tangles. In PD, oxidative stress can promote alpha-synuclein aggregation and the degeneration of dopaminergic neurons. HD is also associated with increased oxidative damage and impaired antioxidant defenses8,9. ROS are highly reactive molecules derived from oxygen, including free radicals such as superoxide (O2 •-), hydroxyl radical (•OH), and non-radical species like hydrogen peroxide (H2O2). Excessive ROS can lead to oxidative damage to biomolecules inside the cells, including protein, DNA, and lipid, leading to various diseases, and accelerating the aging process10. For example, ROS initiates chain reactions that oxidize PUFA in cell membranes, causing membrane dysfunction and forming toxic byproducts like malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE). Protein oxidation occurs when ROS oxidizes amino acid residues within the proteins, leading to protein misfolding, cross-linking, and modulating their functions11. Accumulation of oxidized proteins can contribute to various pathologies. Additionally, ROS can modify DNA bases, leading to breakage of DNA strands and chromosomal aberrations, potentially resulting in 178 genomic instability and mutations if unrepaired11. Mitochondria is one of the main organoids that generates ROS through their major role in cell energy production by synthesizing adenosine triphosphate (ATP) via oxidative phosphorylation. This process initiates with the oxidation of reduced coenzymes, NADH (Nicotinamide Adenine Dinucleotide) and FADH2 (Flavin Adenine Dinucleotide), within the electron transport chain. During this sequence of reactions, complexes I and III of the electron transport chain are the primary sites for generating superoxide radicals. The formation of superoxide radicals involves the single electron reduction of oxygen (O2) mediated by H+ pumps acting on redox-active prosthetic groups found in electron-binding proteins such as reduced coenzymes10,12. This reduction is thermodynamically favorable and is supported by various electron donors in mitochondria, which help drive the reaction forward13. It is estimated that about 1% of the oxygen mitochondria consume is specifically used for producing the superoxide radicals14,15. The production of ROS occurs as a byproduct of normal mitochondrial oxidative phosphorylation but can increase when there is electron leakage from redox centers or enzyme subunits or complexes, taking place either within the mitochondrial matrix or in the intermembrane space. Following its formation, the superoxide radical is rapidly transformed into less reactive non-radical forms of ROS, such as hydrogen peroxide (H2O2), by the action of the superoxide dismutase (SOD) enzyme. This conversion is crucial in controlling ROS levels and preventing cellular damage. Hydrogen peroxide then passes through the membrane and disperses among various cellular compartments12,13,16,17. In addition to mitochondria, peroxisomes and the endoplasmic reticulum (ER) are critical intracellular organelles in producing reactive oxygen. Peroxisomes contribute to ROS production through several pathways, including the β-oxidation of long-chain fatty acids, the catabolism of purines and polyamines, and the metabolism of D- amino acids and PUFAs18. It's important to note that while β-oxidation of very long-chain fatty 179 acids is generally not a major pathway in the brain19, recent research has shown that astrocytes have some capacity to perform β-oxidation of fatty acids20–22. Biological membranes are predominantly made up of lipids and are especially susceptible to oxidative damage due to their rich content of PUFAs. The oxidation of these PUFAs can severely affect the integrity and fluidity of the membrane and lead to the production of various reactive electrophilic species (RES). These RES, resulting from lipid peroxidation, could react with proteins, DNA, and other crucial biomolecules. The presence of lipid peroxidation-derived RES has been linked to the development of a range of inflammatory and degenerative conditions, where they serve as powerful signaling molecules23. These molecules influence various cellular processes, including gene expression, apoptosis, and inflammatory responses. Furthermore, the degradation byproducts of lipid peroxidation, such as aldehydes, ketones, and isoprostanes, are valuable biomarkers for assessing oxidative stress in vivo 23. Tracking these biomarkers sheds light on the underlying mechanisms of diseases and offers the potential for creating diagnostic and therapeutic approaches that specifically address lipid peroxidation and its consequential impacts. This focus on lipid peroxidation could lead to significant advancements in understanding and managing oxidative stress-related conditions. Given the central role of oxidative stress in numerous pathologies, various therapeutic strategies have been explored. These include antioxidant supplementation, targeting specific ROS- generating enzymes, and lifestyle modifications such as exercise and diet24,25. A recent study has also indicated that oxidative stress could produce oxylipins from PUFA, specifically, epoxyeicosatrienoic acids (EETs) from AA in an in vitro system. Such discovery highlights a potential pathway for EETs formation that could occur within cell membranes and impact human 180 physiology26 because EETs prevent mitochondrial dysfunction27, oxidative stress, and neuroinflammation28. However, this observation has not been investigated in vivo. In this chapter, we aim to explore the impact of oxidative stress on lipids and oxylipin profiles, which are implicated in neurodegenerative diseases, as outlined in the previous chapter. We aim to integrate findings from targeted oxylipin analysis and data from untargeted global lipidomics to identify specific lipid regulatory mechanisms that may be pivotal in oxidative-associated disease progression. We seek to discover new metabolic pathways, particularly under stress conditions that induce lipid peroxidation and inflammation, by mapping oxylipins (from targeted analysis) and the broader through untargeted methods. We employ tBHP in our aging animal model, C. elegans, to induce oxidative stress. We are monitoring the relative changes in the concentrations of lipids and oxylipins using two distinct analytical approaches: targeted and untargeted lipidomics. The details of these methods are discussed in the following sections. The targeted LC- MS/MS analysis of oxylipin profiles in C. elegans under oxidative stress has provided a nuanced understanding of the specific metabolic alterations. The significant upregulation of dihydroxy fatty acid, while the CYP 450 metabolites, did not significantly change. For the global lipidomics approach, we tentatively found some compounds from the prenol, and the triglyceride (TG) lipid class significantly degraded under oxidative stress. 2. Research design/strategy 2.1. tBHP-induced oxidative stress in C. elegans Using tBHP to induce oxidative stress in C. elegans is an effective method for studying cellular responses and adaptations to increased oxidative conditions. tBHP, a well-established model compound, is commonly utilized across various biological studies to simulate oxidative stress by generating ROS in cells and tissues. Research involving rat hepatocytes has shown that tBHP can 181 significantly alter cellular redox states, particularly in cells already predisposed to oxidative stress, such as steatosis (fatty) hepatocytes, which refers to liver cells with accumulation of an excessive amount of fat. This condition is known as fatty liver disease or hepatic steatosis29. These cells exhibited higher baseline oxidative stress levels, demonstrated by increased lipid peroxidation and protein carbonylation. Upon tBHP exposure, critical antioxidants like glutathione (GSH) were rapidly depleted, and a notable decrease in mitochondrial membrane potential, indicating severe mitochondrial dysfunction. Additionally, there was an increase in ROS production and cytotoxicity, measured by lactate dehydrogenase leakage, alongside a reduction in the activity of key antioxidant enzymes such as SOD and catalase. This scenario underscores how tBHP can effectively induce oxidative stress, leading to observable physiological changes in cellular model hepatocytes29. Parallel findings in C. elegans have reinforced the utility of tBHP in oxidative stress research. Specifically, the application of tBHP in studies of glucose-6-phosphate dehydrogenase (G6PD) deficiency, a condition known for compromising the cellular redox balance through impaired NADPH production, has been particularly illuminating. tBHP treatment in C. elegans mimicked the oxidative stress conditions observed in G6PD deficiency, leading to similar outcomes such as reduced brood size, increased germ cell apoptosis, and significant embryonic defects30. These effects were associated with increased MDA production and enhanced calcium-independent phospholipase A2 (iPLA) activity, disrupting critical reproductive processes like oogenesis and embryogenesis. Such results provide compelling evidence that tBHP can effectively replicate the oxidative stress and associated phenotypic changes seen in genetic deficiencies affecting the redox balance30. 182 By integrating these insights from both hepatocyte and C. elegans models, it becomes clear that tBHP is a valuable tool for inducing oxidative stress in experimental settings. Its ability to elicit distinct oxidative responses makes it suitable for investigating the molecular and cellular mechanisms underlying oxidative stress and its physiological consequences. This rationale supports using tBHP in C. elegans to elucidate further the pathways involved in oxidative stress management and the potential impacts on organismal health and disease progression. 2.2. Targeted VS. untargeted lipidomics Researchers employ two primary approaches for the in-depth analysis of oxylipins and lipids, depending on the study's objectives. The first approach, the "targeted approach," involves the specific analysis of particular types of oxylipins. To accurately identify compounds, researchers typically use the Multiple Reaction Monitoring (MRM) technique to compare the MRM of the analyte and its retention time with a standard reference. This step is crucial as retention time helps differentiate analytes that share identical MRM transitions. Each analyte is typically assigned unique transitions, which are verified against the method after each injection. The Triple Quadrupole (QqQ) instrument is commonly used for these analyses due to its sensitivity in quantitative analysis. In this study, we used QqQ instrumentation named Xevo TQ-XS from Waters company. Nevertheless, this approach can be limited by the availability of commercial standards, resulting in many studies only analyzing a partial range of known oxylipins and lipids. The untargeted approach is used for a more exhaustive analysis of lipidomic profiles, including all major molecular species of lipids and oxylipins in a sample. This method employs high-resolution mass spectrometry (HR-MS), particularly Quadrupole Time-of-Flight (QTOF) and Orbitap instruments. In untargeted lipidomics, all detected ions are annotated using a database that maps fragmentation patterns. Recently, Watrous et al. developed a novel untargeted LC-MS/MS 183 technique that leverages these fragmentation patterns, gathered from a broad array of standards, to create a network. This network has enabled the classification of over 500 different oxylipins in human plasma, showcasing the potential of this advanced method 31. Integrating both targeted and untargeted metabolomic approaches can yield comprehensive insights into lipid profiles, as exemplified by the work of Wheelock et al. In this study, the researchers employed targeted metabolomics using a QqQ system to measure 36 oxylipins and augmented this with untargeted metabolomics using an Orbitrap analyzer to annotate 219 additional metabolites. This dual approach leverages the strengths of both methodologies to achieve a more complete analysis of lipidomic profiles32. In the field of oxylipin research, the identification of compounds is greatly facilitated by using various databases that can either be accessed manually or integrated into automated systems. These databases are essential for matching mass spectra to known compounds and include several prominent resources. For instance, the LIPID MAPS and Metabolomics WorkBench Database are open-source lipid databases crucial for researchers who need to match observed masses with those cataloged in the LIPID MAPS Structure Database (LMSD). It supports researchers by allowing the calculation of exact masses of lipid ions, predicting MS/MS spectra, and providing detailed structural information such as molecular mass, formula, and shorthand notation33,34. Alongside those databases, LipidBlast serves as an in-silico MS/MS library designed to enhance the identification process in lipidomics by providing an extensive array of predicted tandem mass spectral library of 212,516 MS/MS spectra covering 119,200 compounds from 26 lipid classes35. Additionally, the Lipid Mass Spectrum Analysis Database (LMSAD) offers a wealth of information on mass spectrometry data, structures, and annotations of biologically significant lipids, serving as a valuable reference for lipid compositions in various biological samples36. 184 LipidBank, officially recognized by the Japanese Conference on the Biochemistry of Lipids (JCBL)37, and the METLIN Database, is known for its extensive compilation of individual substance analyses using MS/MS data generated on various mass spectrometry devices38. This study used three databases: Lipid Map, LipdBlast, and Metabolomics WorkBench. 3. Experimental 3.1. Age-synchronized worms To create an age-synchronized worm, we initiated the process by selecting a specific number of healthy, well-nourished day-one adult worms and relocating them to fresh nematode growth media (NGM) plates, seeding each with 50 to 100 worms and OP50 as a bacterial food source. These adult nematodes deposited eggs over a span of approximately 6 to 10 hours. These eggs were then isolated to ensure they hatched separately. After a period ranging from 36 to 48 hours, the plates were cleansed using s-basal solution, and the contents were decanted onto a 40 mm cell strainer positioned atop a 50 mL centrifuge tube. This filtration process was employed to distinguish the larger L4 larvae, which were retained on the filter, from smaller larvae, bacterial residues, or any other extraneous contaminants allowed to pass through. Subsequently, the L4 larvae retained on the filter were thoroughly rinsed with 75 to 100 mL of s-basal and collected into a 1.7 mL centrifuge tube. This tube was centrifuged at 325 x g using a benchtop centrifuge for 30 seconds. Post-centrifugation, the supernatant s-basal was carefully aspirated, leaving a concentrated pellet of L4 larvae. These larvae were then reconstituted in a fresh s-basal solution and carefully placed onto either the control or supplemented NGM plates that had been pre-seeded with OP50, facilitating the recovery post-filtration. 185 3.2. Worms’ treatment with tBHP To assess the effects of tBHP on C. elegans, we first collected worm samples and divided them into two groups. Each group was transferred to separate 15 mL tubes. The control sample was maintained in an S-basal buffer, while the treated sample was exposed to a 5 mM tBHP solution in an S-basal buffer. Both samples were placed in a shaker and incubated at room temperature for 30 minutes. Following incubation, the tBHP-treated and control worms were washed three times with an S-basal buffer. This washing process involved adding S-basal buffer to each tube, centrifuging at 1500x g for 2 minutes, and carefully removing the supernatant to eliminate residual tBHP. The washed samples were then transferred to small vials; later, both control and treated samples were divided into two portions, and oxylipin and global lipidomic sample preparation was applied according to the standard protocol. 3.3. Sample preparation for global lipidomics (LC-HR-MS) The global lipidomics protocol for 10 mg of C. elegans worms involves several steps. First, combine 100 µl of PBS with 10 µl of an antioxidant mixture that includes triphenylphosphine at 0.2 mg/ml in ethanol, butylated hydroxytoluene at 0.2 mg/ml in methanol, and EDTA at 1 mg/ml in water. Next, add 5 µl of an internal standard solution, SPLASH® LIPIDOMIX, containing 15:0- 18:1-d7-PC, 18:1- d7-cholesterol, and 15:0-18:1- d7-15:0 TG at a concentration of 1 mg/ml, sourced from Avanti Polar Lipids, and mix by vortex. Proceed by adding a few metal beads to the sample, flash-freezing with liquid nitrogen, and homogenizing thoroughly. Carefully transfer the homogenized sample to a glass vial. Add 375 µl of ice-cold methanol and 1250 µl of ice-cold methyl tert-butyl ether (MTBE) to the vial and mix thoroughly with a vortex mixer. Incubate this mixture for 1 hour at 4 °C on an orbital shaker set at 200 rpm. 186 After incubation, aid phase separation by adding 750 µl of deionized distilled water (ddH2O) and vortex mixing again. Place the mixture back on the orbital shaker for an additional 10 minutes. Then, centrifuge the mixture for 10 minutes at 4 °C with 1000g. Carefully collect the upper organic layer into a low-binding Eppendorf tube and evaporate the solvent using a speed vacuum concentrator. Reconstitute the dry residue in 100 µl of isopropanol, vortex for 10 minutes, and pass through a centrifuge filter. Transfer the filtrate to a clean vial with a glass insert. Finally, purge the extract with argon gas to prevent oxidation and store it at -80°C until the Q-Exactive injection. All solvents used during the extraction were HPLC grade. 3.4. Global lipidomic, separation technique, and instrumentation The reverse phase ultra-high-performance liquid chromatography (RP-UHPLC) analysis was conducted using a Vanquish Horizon system (Thermo Fisher Scientific, Germering, Germany) equipped with a BEH C18 column (UHPLC, C18, 1.7 µm, 2.1x100mm, Part number 186002352, Waters, Milford, MA, USA). Lipid separation was achieved through gradient elution utilizing solvent A (ACN/ddH2O, 60:40, v/v) and solvent B (i-PrOH/ACN, 90:10 v/v,), each containing ten mM ammonium formate and 0.1% (v/v) formic acid. This process was carried out at a column temperature of 55 °C with a flow rate of 0.4 mL/min, following a specified gradient for solvent B starting with 20% and after two minutes, reaching 43% at 12 minutes, 54% at 12.1 minutes, 70% at 18 minutes, 99% at 18.1 minutes, and 20% at 20 minutes. The positive ESI scan range was m/z 100.0-1500.0, the resolution was 35000 at m/z 200, the spray voltage was 3.5 kV. For negative ESI, parameters were the same except polarity, which is negative, and the spray voltage was 2.5 kV. Each control, tBHP-treated sample, pooled and blank, was analyzed in both positive and negative ESI modes to cover different classes of lipids, while the focus of analysis in this chapter 187 is on negative ESI runs. The analyses performed using data-dependent MS/MS and for collision energy elevated energy stepped of 30-40-50 normalized units was selected. 3.5. Detail information about filtering and normalizing data using progenesis and EZinfo We initiated our analysis by importing negative ESI run data into Progenesis QI software, where we commenced by examining the alignment of peak retention times under the 'Review Alignment' tab. For alignment reference, we selected pooled extracts, resulting in peaks displayed in green at the center of the map, where many of the lipids were concentrated. In the 'Experimental Design' section, we designed two groups: one treated with tBHP and a control group. This setup enabled Progenesis QI to perform statistical comparisons between the two groups. Within the 'Peak-Picking' tab, we accounted for potential losses of compounds during the extraction process by normalizing all peak areas, specifically using PC 15:0/18:1(d7) due to its compatibility with negative ESI ionization. The software provided various functionalities to refine our analysis. We could access crucial data such as p-values, retention times, exact masses, adduct masses, maximum abundance, fold changes, peak width, highest mean, lowest mean, and min CV through the' Review Compounds' tab. We first sorted the compounds by retention time. Compounds with a retention time of less than 0.5 minutes were tagged and later excluded during filtering, as no significant lipids eluted within this timeframe. We later tagged and excluded compounds with a maximum abundance under 10,000 to minimize the complexity of the data set and speed data processing. Our focus then shifted to compounds exhibiting a fold change greater than five and a p-value under 0.05 between the control and tBHP-treated groups, and we generated separate tags for them. Using EZinfo software’s subsequent steps, signal normalization was adjusted according to sample weight, and we conducted a Principal Component Analysis (PCA). Several adjustments were 188 necessary to optimize our statistical analysis using EZinfo software. Under the 'Report' tab of EZinfo, we selected the 'Condition' option to label and color the data in the loading plot. This adjustment allowed EZinfo to segregate the data based on the control and treatment groups, enhancing the clarity and interpretability of results (Figure 4.2). PCA graph showed two different clusters for control and treatment. Additionally, we modified the scaling to 'Pareto' since this scaling method normalizes data to the standard deviation of each measurement, which is particularly useful in weighting lower-abundance compounds to the statistical analysis. Furthermore, we set the data transformation to 'Automatic'. Typically, lipidomic data are not normally distributed, and applying a logarithmic transformation helps scale the distributions to approach normality. 3.6. Targeted lipidomics method to quantify oxylipins Chapter 2-Experimental section provides a detailed explanation of targeted lipidomics, including sample preparation, quantification, and instrumentation. 4. Results 4.1 Untargeted lipidomics: Data analysis using Progenesis QI, EZinfo, Metabolomics Workbench, and Xcalibur 4.1.1. Assigning detected lipids to lipid classes in control samples Our analysis involved subjecting C. elegans extracts to electrospray ionization (ESI) in positive and negative modes to assess the abundances of members of various lipid classes (Figure 4.1). Phosphatidylcholine (PC), Phosphatidylglycerol (PG), Phosphatidylethanolamine (PE), Sphingomyelin (SM), Diacylglycerol (DG), Cholesteryl ester (CE), and Triacylglycerol (TG) are effectively ionized in the positive mode. Conversely, Phosphatidylinositol (PI), PC, PG, Phosphatidylserine (PS), PE, and SM are detected in the negative ESI mode39. Phospholipids with 189 a quaternary nitrogen atom, such as PC and SM, form abundant [M+H] + positive ions. Acidic phospholipids, including PI, PS, and phosphatidic acid (PA), typically generate stable negative ions in the form of [M-H]-. Notably, some lipid classes, specifically PC, PG, PE, and SM, were detected in both ionization modes40. This phenomenon is not uncommon, as many lipids can form both anions and cations, depending on their functional groups and the experimental ion source conditions. However, they may prefer ionization efficiency in one mode over the other. We employed Progenesis QI software to perform peak detection, chromatographic alignment, and adduct combination, followed by a data-driven search within the LipidBlast library, using negative ESI mode for our analysis. While the search identified numerous lipids, many exhibited identification scores below the acceptable threshold of 40. The criteria for these scores include the accuracy of the exact mass, isotopic pattern, fragmentation pattern, retention time, and ion mobility. Our setup lacked ion mobility capabilities, and matching retention times with database entries is notably challenging with LC separation methods, unlike with GC. 190 Figure 4.1. Chromatogram of control sample with two ionization modes. The top one uses a negative ESI mode, and the bottom one uses a positive ESI mode The LipidBlast database search facilitated the comparison of our experimental spectra against the database spectra. We selected 10 with acceptable (> 40) identification scores from the identified compounds that exhibited at least one matching fragment with the database spectra. Notably, these compounds are considered the best matches and the most informative, with all ten belonging to the PE lipids, both of which form [M-H]-1 ions in negative ionization mode. The mass error between the exact mass and experimental mass for all selected compounds was less than 0.005 Da. Based on the measured experimental mass, a molecular formula is predicted. Yet, this molecular formula could match many isomers. For instance, the compound with an exact m/z of 502.241 was suggested to be either PE or GPEtn with the same molecular formula C25H46NO7P, showcasing the complexity of lipid identification where different isomers share the same formula, but the ID score for GPEtn was 39, falling below the acceptable range. The retention times in RP-LC can be 191 used to differentiate classes of lipids based on their aqueous solubility. PE diesters, being larger and more hydrophobic than corresponding PE monoesters, elute later in a reversed-phase separation. For example, a PE diester with a neutral molecular weight of 745.5648 exhibits a longer retention time of 14.35 minutes relative to the PE monoesters. Conversely, a PE monoester, which is smaller and less hydrophobic with m/z 462.2631, has a shorter retention time of 1.64 minutes. This variation primarily reflects the hydrophobicity and molecular size differences between the two compounds since the main functional groups are similar (Table 4.1). Formula Compound ID Table 4.1. Compounds annotated based on LipidBlast search of negative-ion data. These are selected identified compounds based on the identification score above 40. The fragment ion m/z column refers to a C. elegans compound’s fragment ion m/z, that matches the database fragment Fragment ion m/z 281.2479 395.3887 255.2323 305.2479 239.2010 265.2166 241.2166 253.2166 267.2323 Experimental m/z 745.5648 855.6728 737.5010 502.2941 436.2470 462.2631 438.2629 450.2630 715.5530 PE (18:0/18:1) PE (18:2/26:0) PE (16:0/20:5) PE (20:3/0:0) PE (15:1/0:0) PE (17:2/0:0) PE (15:0/0:0) PE (16:1/0:0) PE (18:0/17:1) C41H80NO8P C49H94NO8P C41H72NO8P C25H46NO7P C20H40NO7P C22H42NO7P C20H42NO7P C21H42NO7P C40H78NO7P RT (min) 14.35 15.62 10.14 2.31 1.61 1.64 2.06 1.78 14.45 M-H M-H M-H M-H M-H M-H M-H M-H M-H 43.5 40.0 43.7 57.8 51.5 52.3 58.9 58.5 45.7 IDScore Adduct 4.1. Identifying lipids that show differential change between control and treatment conditions 4.2. 1. Progenesis QI software filtered the results based on the fold change and p-value for tBHP- treated and control samples After conducting a preliminary lipid search on control samples to identify a few classes of lipids present in C. elegans, we studied the effects of oxidative stress induced by tBHP treatment on the global lipidomics of C. elegans. This investigation aims to understand how oxidative stress alters the lipid profile in C. elegans, potentially revealing critical insights into the biochemical responses 192 to oxidative damage at the lipid level. We went through different filtration and data normalization steps, as explained in detail in the experimental section. In the final stages of our analysis, we focused on identifying compounds using databases like ChemSpider and LipidBlast, accessed through the 'Identify Compound' tab within the Progenesis QI software. We opted to filter out compounds with retention times shorter than 0.5 minutes (excluding 220 compounds) and peaks with an abundance of less than 11000 (excluding 1161 compounds). We focused our analysis on compounds with a fivefold change difference in signals between treatments,1322 compounds; from compounds that remained, we selected compounds with p-values less than 0.05. This approach allowed us to concentrate on 59 compounds that showed substantial differences between the control and tBHP-treated groups, facilitating a more focused and potentially insightful analysis of our data. 4.2.2. PCA analysis by EZinfo narrowed down the result into the six most influential compounds in differentiation between control and treatment To facilitate further statistical analysis, such as PCA, we exported our filtered data from Progenesis to EZinfo software. After adjusting parameters, explained in the experimental section, the PCA plot demonstrated distinct clustering of triplicate control and tBHP-treated C. elegans samples (Figure 4.2). 193 Figure 4.2. PCA plot demonstrating distinct clustering of triplicate control and tBHP-treated C. elegans samples based on lipidomic profiling in negative ionization mode. The x-axis is labeled "[t1]" which is the first principal component (PC1), and the y-axis is labeled "[t2]" which is the second principal component (PC2) EZinfo generated a PCA loadings plot, where each spot represents a compound with specific m/z and RT. We selected six compounds that were furthest from the origin. These compounds are considered the most important as they are more abundant and contribute more than those nearer the plot origin to influence the separation of tBHP-treated samples from control samples (Figure 4.3). 194 Figure 4.3. Loading plot analysis of HR-MS data for C. elegans samples with and without tBHP treatment, 6 lipids are selected based on this graph The six selected compounds were transferred back to Progenesis QI (Table 4.2), and we conducted a library search for them within Progenesis, as previously described. This systematic approach ensured the accuracy of our data analysis and highlighted potential biomarkers or essential compounds that could be pivotal in understanding the biochemical impact of the treatments applied in our study. All six compounds gave p-values less than 0.05. Unfortunately, none of the compounds identified via Chem Spider achieved a score higher than 40, rendering them below our annotation acceptance threshold, and LipidBlast failed to identify any lipids. The lack of acceptable lipid identification indicates the need for more extensive identifications of C. elegans lipids and database enhancements to the Metabolomics WorkBench, which is explained in detail in the following section. 195 Table 4.2. List of six most important compounds in negative ionization mode, based on PCA loadings, in differentiating between control samples based on EZinfo PCA analysis Retention time (min) m/z Fold change difference P-value Highest mean Compound 1 Compound 2 Compound 3 Compound 4 Compound 5 Compound 6 2.71 3.33 3.36 2.94 2.05 1.97 421.2179 339.2330 525.2803 466.1322 405.1864 433.1817 64 16.9 19.6 75.2 19.2 32.4 0.01 0.0094 0.0055 0.039 0.0014 0.0035 Control Control Control Control Control Control 4.2.3. Metabolomics WorkBench provided multiple structures for each m/z Given the LipidBlast library's inability to successfully identify the six compounds of interest, we adopted a more refined approach using the Metabolomics Workbench for an advanced search. Using that database, we needed to set specific mass tolerances to enhance the accuracy of our database search. We utilized the formate adduct [M+HCOO-] of internal standard with an exact mass of 797.6057, whereas ChemDraw calculated the exact mass as 797.6043. This resulted in a calculated mass error of 1.76 ppm, equivalent to 0.0014 Da. Notably, an error of less than 0.005 Da for orbitrap mass analyzers is generally acceptable, and our calculations comfortably fell within this threshold. To expand our search, we considered different possible adducts for each of the six compounds, including [M-H]-, [M-CH3]-, and [M+formate]-. The logic behind considering formate adduct was that both ammonium formate and formic acid were present in the mobile phase, and some lipids lacking acidic functional groups may ionize by formate adduct formation. This approach generated multiple potential identifications, with up to 15 suggestions for some m/z values examined. A preliminary review of each suggestion determined that three of the six compounds, based on PCA, were not lipids; for example, a compound with m/z 433.1817 was suggested to be a peptide, benzenediols, or an organo-heterocyclic. The Compound with m/z 196 421.2179 is probably a peptide or an oxygen-containing organic compound. Lastly, a compound with m/z 466.1322 may likely be a peptide or polyketide. By eliminating those three m/z, this database search effectively narrows our focus to the remaining three compounds, including compounds with m/z 339.2330, 525.2797, and 405.1862. 4.2.4. Xcalibur, MS2 provided fragmentation information for each compound To elucidate the structure of the remaining three compounds, we analyzed the chromatograms, Data-Dependent Acquisition (DDA), MS2 spectra, and the fragmentation patterns of each m/z value using Thermo Xcalibur software. This comprehensive approach aimed to identify a structure that aligns with the observed fragmentation patterns. For the compound with an m/z 339.2330, the most significant fragmentation observed was at m/z 163.1132. Other smaller fragments were deemed noise due to their low intensity and uniform height, which suggested a lack of meaningful contribution to the compound's structure. Notably, the mass defect analysis indicates the number of hydrogens; since hydrogen is the only element with a significant positive mass defect, it is the source of the mass defect in each compound. In other words, by dividing the mass defect of the compound into the mass defect of hydrogen (0.0078), we can find an approximation of the number of H in that compound. Based on the MS2 spectrum (Figure 4.4), we found the changes in the number of hydrogens could not account for the observed mass defects among the other fragment ions. This analysis confirmed that only the fragment at m/z 163.1132 is pivotal for identifying the compound, and all others are noise. We used MassLynx to find the elemental composition of the compound with m/z 163.1132. In case we use the most common elements like C, H, N, and O, only one formula, (C11H15O-) matches the molecular weight; the number of H in this formula suggests the compound would have a significant number of rings or double bonds, probably aromatic rings. 197 Figure 4.4. Extracted ion chromatogram and negative-ion MS2 spectrum related to compound with m/z 339.2330 The Metabolomics Workbench suggested two classes of lipids that matched the exact mass of 339.2330 steroids and prenols (Figure 4.5). However, steroids typically do not form an (M-H) molecular ion due to the absence of acidic hydrogen. Thus, do not ionize well in negative ESI, excluding steroid-like structures from our considerations. Conversely, the prenol class, specifically phenolic compounds, matched the observed retention time and showed the ability to generate an [M-H]- ion, aligning more closely with our experimental observations. Nevertheless, after examining the potential fragmentation patterns of all suggested prenol structures, the key fragment at m/z 163 observed in the spectra did not match the expected fragmentation of any of the proposed structures. That led to a lack of fragmentation evidence to support identifying tentative prenol structure. It is known that phenolic oxygen in prenol can undergo rearrangements to form a six- membered ring with a double bond and fuse to the aromatic ring, which makes it difficult to fragment. Given this context, we tentatively identified a compound with m/z 339.2330 as a prenol 198 lipid with the molecular formula C23H32O2. However, the presence of numerous possible isomers Figure 4.5. The compounds suggested by the metabolomic workbench with m/z 339.2330 and mass tolerance <0.005, The top structures are sterol lipids, and the bottom structure is prenol lipids and the limited information from our instrumental analysis means that a definitive structural determination remains elusive. Thus, while we are reasonably confident that the compound is a prenol lipid, the exact structural isomer cannot be conclusively identified without additional analytical data. For the second compound with m/z 405.1862, the initial examination involved reviewing structures that matched the exact mass within the acceptable mass tolerance of the instrument. Per the Metabolomics Workbench suggestions, all the structures included a chlorine (Cl) atom. So, a critical aspect of the analysis involved checking for the presence of a Cl isotopic pattern, the [M+2]- peak at m/z 407, which would indicate the presence of a Cl isotope atom (Figure 4.6). The absence of this peak in our spectra led to the conclusion that Cl is not a compound component, 199 thus ruling out all structures containing chlorine suggested by the Metabolomics Workbench (Figure 4.7). Figure 4.6. Extracted ion chromatogram and negative-ion MS2 spectrum related to compound with m/z 405.1862 To further refine our understanding of the compound’s composition, we employed the elemental composition tool on MassLynx software. This tool helps determine the likely elemental composition by considering the (M+1) carbon isotope peak, which primarily reflects the number of carbon atoms due to the natural abundance of 13C. Isotope analysis narrowed the molecule's possible number of carbon atoms to around 32; we used a wider range of (25-34) carbon and used that range to filter our options during the search for elemental components. Only one formula, C29H25O2 -, matched the exact mass with the appropriate carbon atoms. Considering that the compound forms an [M-H]- ion in negative ESI, this suggests the compound has acidic properties, 200 leading to a revised neutral elemental formula of C29H26O2. This composition, with more carbons relative to hydrogens, suggests a structure with numerous rings and aromatic systems. Such a structure is uncommon for traditional lipids, which usually contain long chains of hydrocarbons with fewer aromatic rings and more hydrogen than carbon. Figure 4.7. The compounds suggested by the Metabolomic Workbench are all prenol lipids with m/z 405.1864 and mass tolerance <0.005, each containing one Cl atom The mass spectrometry analysis of the compound with an m/z 525.2797 did not reveal a definitive fragmentation pattern (Figure 4.8). All observed fragments displayed very low intensity, resembling noise rather than meaningful structural information. Additionally, there are fragments with uncommon mass losses of 50 and 10 Da, further supporting the theory that these signals are not true fragments but noise. Given the above findings and the absence of a fragmentation pattern, we could still progress our analysis by excluding certain suggested structures from the Metabolomics Workbench list. For example, the possibility of phospholipids forming formate adducts was discounted (Figure 4.9). 201 Phospholipids typically do not form these adducts as they are sufficiently acidic to form [M-H]- ions. This consideration allowed us to rule out two suggested structures involving phospholipid formate adducts. The remaining suggested structures included PG 14:0 and PG 16:0 with 14 and 16-carbon chain esters. PG14, however, is uncommon in the lipid profiles typically analyzed, which led us to focus on the structure of PG 16:0. However, the possibility of other isomers cannot be dismissed, and the lack of a supportive fragmentation pattern means that while we can be reasonably confident about the triacylglycerols lipid class and molecular formula, determining the exact structural isomer requires additional data. Further analytical approaches would be necessary to identify the exact structure of this lipid conclusively. Techniques such as nuclear magnetic resonance (NMR) spectroscopy or ion mobility spectrometry provide more structural information but require purification of larger quantities. These methods would help to confirm the presence of specific structural features and eliminate the ambiguity currently faced due to the lack of clear fragmentation data. Hence, while we can tentatively classify this compound within a specific lipid category, precise structural identification awaits further investigation. 202 Figure 4.8. Chromatogram and MS2 spectrum related to compound with m/z 525.2797 Figure 4.9. The compounds suggested by the Metabolomics Workbench with m/z 525.2797 and mass tolerance <0.005. The top two structures are Glycerophospholipids. The one on the right is PG 16:0/2:0 the one on the left is PG 14:0/4:0. The bottom structures are in formate adduct form, and the bottom left structure is PA 10:0/10:0, and the bottom right structure is PA 16:0/4:0 203 4.2. Targeted lipidomics indicates a significant change in the concentration of LOX and EH metabolites Our LC-MS/MS analysis delineated the impact of tBHP treatment on oxylipin profiles in C. elegans; for lipoxygenase metabolites, tBHP treatment led to a pronounced increase in the concentrations of hydroxyeicosatetraenoic acids (HETEs), hydroxyeicosapentaenoic acids (HEPEs), as well as 9-hydroxyoctadecadienoic acid (9-HODE) in the tBHP-treated group, implying an upregulation of LOX enzymes' activity in response to oxidative stress. The increase could also be due to non-enzymatic oxidation; this is suggested by the increase in all regioisomers of hydroxylated fatty acids from the same fatty acid rather than specific regioisomers. Moreover, C. elegans does not appear to have direct homologs of mammalian lipoxygenases or cyclooxygenases. The nematode's genome does not encode enzymes that are structurally similar to mammalian LOX or COX. Dihydroxyeicosatrienoic acids (DiHETEs), including 11,12- DiHETE and 12,13-DiHOME, were significantly elevated in the tBHP-treated samples. These diols are known metabolites produced by the EH, suggesting that this EH metabolic pathway is likely altered by the tBHP-induced oxidative stress, upregulating the dihydroxy fatty acids. Typically, if only EH was induced, besides producing more dihydroxy fatty acids, the epoxy fatty acids, the substrate of the EH, should also be downregulated. However, the concentrations of epoxy fatty acids like epoxyeicosatrienoic acids (EpETEs) and epoxyoctadecamonoenoic acids (EpOMEs) did not show significant changes upon treatment with tBHP. Such results indicated while the endogenous levels of dihydroxy fatty acid increase, the production of the epoxy fatty acids is likely to be increased; therefore, the endogenous levels of epoxy fatty acids were maintained (Figure 4.10). 204 Figure 4.10. Comparative analysis of oxylipin concentrations in C. elegans (mean+/- SEM): control VS tBHP treatment – Two-Way ANOVA (Tukey) significance indicated by ** p- value<0.01, **** p-value<0.0001" 5. Discussion and future direction Using tBHP as an oxidative stress inducer in C. elegans has provided valuable insights into the cellular adaptations and responses to increased oxidative conditions. In this study, we used both targeted (TQXS as a triple quadrupole and MRM) and untargeted lipidomics (Q-Exactive as an orbitrap and DDA). Our results showed significant alterations in the oxylipin profile and three lipids’ concentrations following tBHP treatment, indicating a complex interplay between lipid metabolism and oxidative stress response mechanisms. 205 The lipid peroxidation products generated under oxidative stress may influence the intricate balance of lipid profiles in C. elegans. We used UHPLC coupled with HR-MS instrumentation plus various software and databases, including Progenesis QI, EZinfo, LipidMap, Metabolomics WorkBench, and Xcalibur to tentatively identify three m/z, including (339.2330, 405.1862, and 525.2797), with at least fivefold difference and P-value <0.05 between the control and treatment groups all those compounds showed higher concentration in control in compare of treatment group. We successfully identified the molecular formulas of the compounds and the lipid classes. However, as previously discussed, the absence of strong fragmentation patterns prevented us from determining the exact structures and specific isomers. This is consistent with the function of prenol lipids in cellular defense. For instance, alpha-tocopherol, a prenol lipid that is a potent biological antioxidant, is the most active form of vitamin E in humans. It can act as an antioxidant to alleviate oxidative stress induced by the GPX4 inhibitor RSL341. Our study showed oxidative stress degrades this compound. However, the precise role of prenol lipids in managing oxidative stress within C. elegans is yet to be fully understood and warrants further investigation. Moreover, the modulation of glycerophospholipids levels, probably a compound such as PG 16:0, in response to oxidative stress indicates a shift in lipid metabolism, which may be related to changes in membrane composition or energy storage strategies. These alterations could either be part of a direct response to oxidative damage or represent an adaptation to maintain cellular homeostasis during stress recovery 42. As noted earlier, while HR-MS effectively determined the exact mass of each compound, our current analytical equipment failed to deliver clear fragmentation patterns and definitive structural identification for each lipid metabolite. This limitation was due to the low signal-to-noise levels and the presence of numerous isomers for each lipid compound. The low signal-to-noise problem 206 might be solved by concentrating the worm extracts to obtain better MS/MS spectra. To overcome the presence of numerous isomers challenge, integrating IMS into our analytical setup could lead to significant advancements in the future43. Adding IMS and the collision cross-section (CCS) values would be particularly beneficial for distinguishing between isomeric species, thereby enhancing our ability to elucidate the structures of lipid metabolites. CCS is a physical property of ions representing the effective area presented by an ion when it collides with neutral gas molecules in an IMS experiment. The CCS value reflects the size and shape of an ion in its three-dimensional conformation as it travels through a gas-filled chamber under the influence of an electric field44. CCS is valuable for isomer detection because isomers—molecules with the same chemical formula but different structural arrangements can present different shapes and, therefore, different CCS. Even though isomers may have identical mass-to-charge ratios and cannot be distinguished by traditional mass spectrometry alone, their CCS values can differ due to their unique geometries44. Levels of confidence for molecular identification, as per the Metabolomics Standards Initiative, are classified into four categories45. The highest confidence level is achieved by matching an authentic standard with two orthogonal data points, such as chromatographic retention time and accurate molecular mass. The second level involves putatively identified compounds through spectrum library matching. The third level pertains to compounds putatively assigned to a compound class based on spectrum similarity. The fourth level includes unknowns. It is important to note that molecular and fragment masses alone are not sufficient for compound identification45, which explains why, in this study, we defined the class of lipid, not the compound or regioisomer. Our oxylipin analysis showed increased HETEs, HEPEs, and 9-HODE concentrations in the tBHP- treated group. This led to a preliminary interpretation of upregulated LOX enzymes' activity in response to oxidative stress. However, these increases might also be attributed to non-enzymatic 207 oxidation processes. This was particularly indicated by the uniform increase in all regioisomers of hydroxylated fatty acids from the same fatty acid rather than selective increases in our data that would typify enzymatic processes. Enzymatic oxidation, typically mediated by LOX enzymes, is known for its specificity, producing distinct regioisomers. Using tBHP, a powerful oxidizing agent, supports the likelihood of non-enzymatic lipid peroxidation contributing to the widespread increase in various regioisomers points to a generalized oxidative effect rather than solely enzymatic enzyme upregulation. This suggests that the observed lipid alterations could be a direct outcome of oxidative stress, possibly in addition to any regulated enzymatic responses. We observed a significant increase in some dihydroxy fatty acids metabolites in response to tBHP treatment. Dihydroxy fatty acids are known as proinflammatory metabolites, such as DiHETrE46, and their upregulation may represent a defensive response to mitigate the damaging effects of ROS or oxidative stress. Oxidative stress is a well-documented activator of the nuclear factor kappa- light-chain-enhancer of activated B cells (NF-κB), a pivotal transcription factor that regulates the expression of various genes involved in inflammation. Research has highlighted that epoxy fatty acids are anti-inflammatory and can modulate the NF-κB pathway. sEH is an enzyme that metabolizes epoxy metabolites into their corresponding dihydroxy metabolites, thus reducing the availability of these anti-inflammatory epoxides46. Studies involving genetically modified mice with overexpressed CYP2J2, CYP2C8, and sEH have demonstrated that elevated levels of epoxy fatty acids can effectively inhibit the activation of NF-κB and decrease the expression of pro- inflammatory cytokines. These findings suggest that by controlling the activity of sEH and consequently slowing the degradation of epoxy metabolites, it may be possible to alleviate symptoms of oxidative stress and cellular inflammation. In the context of our study on C. elegans, the observed increase in dihydroxy metabolites upon tBHP-induced oxidative stress suggests that 208 sEH may be upregulated as part of the organism's response to oxidative challenges. By modulating the balance between epoxy and dihydroxy, sEH could be crucial in controlling the inflammatory response and maintaining cellular homeostasis under oxidative stress. The unchanged concentrations of CYP metabolites like the epoxides EpETEs and EpOMEs, alongside the increase in DiHETEs, raise intriguing questions about the dynamics of epoxy and dihydroxy metabolites in the context of oxidative stress. This observation suggests the potential involvement of alternative pathways or feedback mechanisms that ensure a constant supply of these epoxies. Upregulation of CYP enzymes could lead to an increase in epoxide production to match the heightened conversion to dihydroxy, while oxidative stress might also amplify the availability of fatty acid substrates essential for epoxide formation, potentially involving the upregulation of phospholipases to release fatty acids from membrane phospholipids. The observed changes might reflect a cellular adaptive response to maintain homeostasis under stress conditions. To further investigate these dynamics, examining the expression or activity of CYP enzymes under the experimental conditions could provide deeper insights. Researchers have demonstrated that silencing the sEH enzyme, encoded by the Ephx2 gene, can mitigate hydrogen peroxide (H2O2)-induced oxidative stress and damage in intestinal epithelial cells (IEC6)47 via activating PI3K/Akt/GSK3β signaling pathway48. This potential involvement of sEH in managing oxidative stress and inflammation46 emphasizes the importance of further research into the regulatory mechanisms of this enzyme and its implications for health and disease. To address that in our animal model, a future approach would be to employ inhibitors of sEH, such as AUDA, to investigate how inhibiting the conversion of epoxy metabolites to dihydroxy metabolites affects the lipid profile under oxidative stress conditions. By blocking this conversion, we can more directly assess the role of epoxy metabolites in managing oxidative stress, evaluate the stability and function of these metabolites in the absence of sEH 209 activity, and explore their potential protective benefits against oxidative damage. This approach could provide deeper insights into the regulatory mechanisms of lipid metabolism during oxidative stress and help to clarify the specific contributions of epoxy and dihydroxy metabolites in cellular processes. Integrating our targeted oxylipin profiling with the previously observed variations in other lipid metabolites, such as prenol lipids and triacylglycerols, underscores the complexity of lipid metabolism in response to oxidative stress. As previously discussed, this chapter focused primarily on data analysis of negative ESI data, while samples were run in both modes separately. However, it's essential to consider conducting the same analysis and database search using positive ESI, as many lipids ionize more effectively in this mode. This complementary approach can enhance the coverage and depth of our lipidomic analysis, leading to a more comprehensive understanding of lipid profiles and their behaviors under oxidative stress. Our current work serves as a pilot study, laying the groundwork for further exploration into how oxidative stress influences lipid profiles. An important future direction in this research is the investigation of lipid hydroperoxides, which are critical mediators of oxidative stress. These molecules, formed when lipids are oxidized, play a significant role in cellular signaling pathways49, including those leading to ferroptosis, a form of regulated cell death driven by the catastrophic peroxidation of cellular lipids. Understanding the formation and impact of lipid hydroperoxides could provide deeper insights into the mechanisms by which oxidative stress damages cellular components and induces ferroptosis50. This exploration is particularly relevant in the context of diseases characterized by oxidative stress and ferroptosis, such as neurodegenerative disorders. By focusing on these lipid mediators, future studies could potentially unveil new therapeutic targets and strategies to mitigate the harmful effects of oxidative stress and prevent cell death 210 5. Conclusion Our study underscores the intricate relationship between lipid metabolism and the oxidative stress response. The alterations in oxylipin concentrations observed in tBHP-treated C. elegans indicate that these metabolites might serve dual roles as both indicators and regulators of oxidative stress. Notably, the specific increases in metabolites from the LOX or non-enzymatic oxidation and sEH pathways highlight their significant roles in adapting to oxidative stress. In contrast, the steady levels of other CYP metabolites suggest a complex, well-maintained biological equilibrium. Additionally, global lipidomics analyses using HR-MS revealed significant alterations in the concentrations of specific lipids from different classes, further emphasizing the broad impact of oxidative stress on lipid homeostasis. Our comprehensive analysis of lipid profiles in C. elegans offers valuable insights, enhancing our understanding of the dynamic responses of lipid metabolism to oxidative stress and laying a foundation for future research in this pivotal area of lipid biology. 211 BIBLIOGRAPHY (1) Senoner, T.; Dichtl, W. Oxidative Stress in Cardiovascular Diseases: Still a Therapeutic Target? Nutrients 2019, 11 (9), 2090. (2) Reuter, S.; Gupta, S. C.; Chaturvedi, M. M.; Aggarwal, B. B. Oxidative Stress, Inflammation, and Cancer: How Are They Linked? Free Radic. Biol. Med. 2010, 49 (11), 1603–1616. (3) Yan, Z.; Chen, Q.; Xia, Y. Oxidative Stress Contributes to Inflammatory and Cellular Damage in Systemic Lupus Erythematosus: Cellular Markers and Molecular Mechanism. J. Inflamm. Res. 2023, 16, 453–465. (4) Pereira, M. D.; Ksiazek, K.; Menezes, R. Oxidative Stress in Neurodegenerative Diseases and Ageing. Oxid. Med. Cell. Longev. 2012, 2012, 796360. (5) Gonfloni, S.; Maiani, E.; Di Bartolomeo, C.; Diederich, M.; Cesareni, G. Oxidative Stress, DNA Damage, and c-Abl Signaling: At the Crossroad in Neurodegenerative Diseases? Int. J. Cell Biol. 2012, 2012, 683097. (6) Liu, Z.; Zhou, T.; Ziegler, A. C.; Dimitrion, P.; Zuo, L. Oxidative Stress in Neurodegenerative Diseases: From Molecular Mechanisms to Clinical Applications. Oxid. Med. Cell. Longev. 2017, 2017, 1–11. (7) Frank, M. G. The Mystery of Sleep Function: Current Perspectives and Future Directions. Rev. Neurosci. 2006, 17 (4), 375–392. (8) Singh, A.; Kukreti, R.; Saso, L.; Kukreti, S. Oxidative Stress: A Key Modulator in Neurodegenerative Diseases. Molecules 2019, 24 (8), 1583. (9) Korovesis, D.; Rubio-Tomás, T.; Tavernarakis, N. Oxidative Stress in Age-Related Neurodegenerative Diseases: An Overview of Recent Tools and Findings. Antioxidants (Basel) 2023, 12 (1), 131. (10) Newsholme, P.; Cruzat, V. F.; Keane, K. N.; Carlessi, R.; de Bittencourt, P. I. H., Jr. Molecular Mechanisms of ROS Production and Oxidative Stress in Diabetes. Biochem. J. 2016, 473 (24), 4527–4550. (11) Ayala, A.; Muñoz, M. F.; Argüelles, S. Lipid Peroxidation: Production, Metabolism, and Signaling Mechanisms of Malondialdehyde and 4-Hydroxy-2-Nonenal. Oxid. Med. Cell. Longev. 2014, 2014, 360438. (12) Galadari, S.; Rahman, A.; Pallichankandy, S.; Thayyullathil, F. Reactive Oxygen Species and Cancer Paradox: To Promote or to Suppress? Free Radic. Biol. Med. 2017, 104, 144–164. (13) Introductory Chapter: Unregulated Mito- Chondrial Production of Reactive Oxygen Species in Testing the Biological Activity of Compounds. Medicinal chemistry. (14) Zaja-Milatovic, S.; Gupta, R. C. Oxidative Stress and Excitotoxicity: Antioxidants from Nutraceuticals. In Nutraceuticals; Elsevier, 2021; pp 485–497. 212 (15) Di Meo, S.; Reed, T. T.; Venditti, P.; Victor, V. M. Role of ROS and RNS Sources in Physiological and Pathological Conditions. Oxid. Med. Cell. Longev. 2016, 2016, 1245049. (16) Zhang, J.; Wang, X.; Vikash, V.; Ye, Q.; Wu, D.; Liu, Y.; Dong, W. ROS and ROS-Mediated Cellular Signaling. Oxid. Med. Cell. Longev. 2016, 2016, 4350965. (17) Angelova, P. R.; Abramov, A. Y. Functional Role of Mitochondrial Reactive Oxygen Species in Physiology. Free Radic. Biol. Med. 2016, 100, 81–85. (18) Kurutas, E. B. The Importance of Antioxidants Which Play the Role in Cellular Response against Oxidative/Nitrosative Stress: Current State. Nutr. (1). https://doi.org/10.1186/s12937-016-0186-5. 2015, 15 J. (19) Schönfeld, P.; Reiser, G. Why Does Brain Metabolism Not Favor Burning of Fatty Acids to Provide Energy? Reflections on Disadvantages of the Use of Free Fatty Acids as Fuel for Brain. J. Cereb. Blood Flow Metab. 2013, 33 (10), 1493–1499. (20) Auestad, N.; Korsak, R. A.; Morrow, J. W.; Edmond, J. Fatty Acid Oxidation and Ketogenesis by Astrocytes in Primary Culture. J. Neurochem. 1991, 56 (4), 1376–1386. (21) Esfandiari, A.; Soifiyoudine, D.; Paturneau-Jouas, M. Inhibition of Fatty Acid β-Oxidation in Rat Brain Cultured Astrocytes Exposed to the Neurotoxin 3-Nitropropionic Acid. Dev. Neurosci. 1997, 19 (4), 312–320. (22) Eraso-Pichot, A.; Brasó-Vives, M.; Golbano, A.; Menacho, C.; Claro, E.; Galea, E.; Masgrau, R. GSEA of Mouse and Human Mitochondriomes Reveals Fatty Acid Oxidation in Astrocytes. Glia 2018, 66 (8), 1724–1735. (23) Leong, X.-F. Lipid Oxidation Products on Inflammation-Mediated Hypertension and Atherosclerosis: A Mini Review. Front. Nutr. 2021, 8, 717740. (24) Sharifi-Rad, M.; Anil Kumar, N. V.; Zucca, P.; Varoni, E. M.; Dini, L.; Panzarini, E.; Rajkovic, J.; Tsouh Fokou, P. V.; Azzini, E.; Peluso, I.; Prakash Mishra, A.; Nigam, M.; El Rayess, Y.; Beyrouthy, M. E.; Polito, L.; Iriti, M.; Martins, N.; Martorell, M.; Docea, A. O.; Setzer, W. N.; Calina, D.; Cho, W. C.; Sharifi-Rad, J. Lifestyle, Oxidative Stress, and Antioxidants: Back and Forth in the Pathophysiology of Chronic Diseases. Front. Physiol. 2020, 11, 694. (25) Forman, H. J.; Zhang, H. Author Correction: Targeting Oxidative Stress in Disease: Promise and Limitations of Antioxidant Therapy. Nat. Rev. Drug Discov. 2021, 20 (8), 652. (26) Weiny, J. A.; Boeglin, W. E.; Calcutt, M. W.; Stec, D. F.; Brash, A. R. Monolayer Autoxidation of Arachidonic Acid to Epoxyeicosatrienoic Acids as a Model of Their Potential Formation in Cell Membranes. J. Lipid Res. 2022, 63 (1), 100159. (27) Liu, L.; Chen, C.; Gong, W.; Li, Y.; Edin, M. L.; Zeldin, D. C.; Wang, D. W. Epoxyeicosatrienoic Acids Attenuate Reactive Oxygen Species Level, Mitochondrial Dysfunction, Caspase Activation, and Apoptosis in Carcinoma Cells Treated with Arsenic Trioxide. J. Pharmacol. Exp. Ther. 2011, 339 (2), 451–463. 213 (28) Lakkappa, N.; Krishnamurthy, P. T.; Hammock, B. D.; Velmurugan, D.; Bharath, M. M. S. Possible Role of Epoxyeicosatrienoic Acid in Prevention of Oxidative Stress Mediated Neuroinflammation in Parkinson Disorders. Med. Hypotheses 2016, 93, 161–165. (29) Yang, H.-C.; Yu, H.; Ma, T.-H.; Tjong, W.-Y.; Stern, A.; Chiu, D. Tert-Butyl Hydroperoxide (TBHP)-Induced Lipid Peroxidation and Embryonic Defects Resemble Glucose-6-Phosphate Dehydrogenase (G6PD) Deficiency in C. Elegans. Int. J. Mol. Sci. 2020, 21 (22), 8688. (30) Kučera, O.; Endlicher, R.; Roušar, T.; Lotková, H.; Garnol, T.; Drahota, Z.; Červinková, Z. The Effect ofTert-Butyl Hydroperoxide-Induced Oxidative Stress on Lean and Steatotic Rat Hepatocytesin Vitro. Oxid. Med. Cell. Longev. 2014, 2014, 1–12. (31) Watrous, J. D.; Niiranen, T. J.; Lagerborg, K. A.; Henglin, M.; Xu, Y.-J.; Rong, J.; Sharma, S.; Vasan, R. S.; Larson, M. G.; Armando, A.; Mora, S.; Quehenberger, O.; Dennis, E. A.; Cheng, S.; Jain, M. Directed Non-Targeted Mass Spectrometry and Chemical Networking for Discovery of Eicosanoids and Related Oxylipins. Cell Chem. Biol. 2019, 26 (3), 433-442.e4. (32) Kamleh, M. A.; McLeod, O.; Checa, A.; Baldassarre, D.; Veglia, F.; Gertow, K.; Humphries, S. E.; Rauramaa, R.; de Faire, U.; Smit, A. J.; Giral, P.; Kurl, S.; Mannarino, E.; Tremoli, E.; Silveira, A.; Örvik, J.; Hamsten, A.; Wheelock, C. E. Increased Levels of Circulating Fatty Acids Are Associated with Protective Effects against Future Cardiovascular Events in Nondiabetics. J. Proteome Res. 2018, 17 (2), 870–878. (33) Sud, M.; Fahy, E.; Cotter, D.; Brown, A.; Dennis, E. A.; Glass, C. K.; Merrill, A. H., Jr; Murphy, R. C.; Raetz, C. R. H.; Russell, D. W.; Subramaniam, S. LMSD: LIPID MAPS Structure Database. Nucleic Acids Res. 2007, 35 (Database issue), D527-32. (34) Fahy, E.; Subramaniam, S.; Murphy, R. C.; Nishijima, M.; Raetz, C. R. H.; Shimizu, T.; Spener, F.; van Meer, G.; Wakelam, M. J. O.; Dennis, E. A. Update of the LIPID MAPS Comprehensive Classification System for Lipids. J. Lipid Res. 2009, 50 Suppl (Supplement), S9- 14. (35) Kind, T.; Liu, K.-H.; Lee, D. Y.; DeFelice, B.; Meissen, J. K.; Fiehn, O. LipidBlast in Silico Tandem Mass Spectrometry Database for Lipid Identification. Nat. Methods 2013, 10 (8), 755– 758. (36) LipidMass Spectrum Analysis Database. In Proceedings of the 2010 International Conference on Biomedical Engineering and Computer Science. (37) Watanabe, K.; Yasugi, E.; Oshima, M. How to Search the Glycolipid Data in “LIPIDBANK for Web” the Newly Developed Lipid Database in Japan. Trends Glycosci. Glycotechnol. 2000, 12 (65), 175–184. (38) Guijas, C.; Montenegro-Burke, J. R.; Domingo-Almenara, X.; Palermo, A.; Warth, B.; Hermann, G.; Koellensperger, G.; Huan, T.; Uritboonthai, W.; Aisporna, A. E.; Wolan, D. W.; Spilker, M. E.; Benton, H. P.; Siuzdak, G. METLIN: A Technology Platform for Identifying Knowns and Unknowns. Anal. Chem. 2018, 90 (5), 3156–3164. 214 (39) Breitkopf, S. B.; Ricoult, S. J. H.; Yuan, M.; Xu, Y.; Peake, D. A.; Manning, B. D.; Asara, J. M. A Relative Quantitative Positive/Negative Ion Switching Method for Untargeted Lipidomics via High Resolution LC-MS/MS from Any Biological Source. Metabolomics 2017, 13 (3). https://doi.org/10.1007/s11306-016-1157-8. (40) Harshfield, E. L.; Koulman, A.; Ziemek, D.; Marney, L.; Fauman, E. B.; Paul, D. S.; Stacey, D.; Rasheed, A.; Lee, J.-J.; Shah, N.; Jabeen, S.; Imran, A.; Abbas, S.; Hina, Z.; Qamar, N.; Mallick, N. H.; Yaqoob, Z.; Saghir, T.; Rizvi, S. N. H.; Memon, A.; Rasheed, S. Z.; Memon, F.- U.-R.; Qureshi, I. H.; Ishaq, M.; Frossard, P.; Danesh, J.; Saleheen, D.; Butterworth, A. S.; Wood, A. M.; Griffin, J. L. An Unbiased Lipid Phenotyping Approach to Study the Genetic Determinants of Lipids and Their Association with Coronary Heart Disease Risk Factors. J. Proteome Res. 2019, 18 (6), 2397–2410. (41) Xie, Y.; Hou, W.; Song, X.; Yu, Y.; Huang, J.; Sun, X.; Kang, R.; Tang, D. Ferroptosis: Process and Function. Cell Death Differ. 2016, 23 (3), 369–379. (42) Fuhrman, B.; Volkova, N.; Aviram, M. Postprandial Serum Triacylglycerols and Oxidative Stress in Mice after Consumption of Fish Oil, Soy Oil or Olive Oil: Possible Role for Paraoxonase- 1 Triacylglycerol Lipase-like Activity. Nutrition 2006, 22 (9), 922–930. (43) Camunas-Alberca, S. M.; Moran-Garrido, M.; Sáiz, J.; Gil-de-la-Fuente, A.; Barbas, C.; Gradillas, A. Integrating the Potential of Ion Mobility Spectrometry-Mass Spectrometry in the Separation and Structural Characterisation of Lipid Isomers. Front. Mol. Biosci. 2023, 10, 1112521. (44) Lanucara, F.; Holman, S. W.; Gray, C. J.; Eyers, C. E. The Power of Ion Mobility-Mass Spectrometry for Structural Characterization and the Study of Conformational Dynamics. Nat. Chem. 2014, 6 (4), 281–294. mner, L. W.; Amberg, A.; Barrett, D.; Beale, M. H.; Beger, R.; Daykin, C. A.; Fan, T. W.-M.; Fiehn, O.; Goodacre, R.; Griffin, J. L.; Hankemeier, T.; Hardy, N.; Harnly, J.; Higashi, R.; Kopka, J.; Lane, A. N.; Lindon, J. C.; Marriott, P.; Nicholls, A. W.; Reily, M. D.; Thaden, J. J.; Viant, M. R. Proposed Minimum Reporting Standards for Chemical Analysis. Metabolomics 2007, 3 (3), 211–221. (46) Wagner, K. M.; McReynolds, C. B.; Schmidt, W. K.; Hammock, B. D. Soluble Epoxide Hydrolase as a Therapeutic Target for Pain, Inflammatory and Neurodegenerative Diseases. Pharmacol. Ther. 2017, 180, 62–76. (47) Li, X.; Wu, X. Soluble Epoxide Hydrolase (Ephx2) Silencing Attenuates the Hydrogen Peroxide-Induced Oxidative Damage in IEC-6 Cells. Arch. Med. Sci. 2021, 17 (4), 1075–1086. (48) Li, J.; Luo, J.; Zhang, Y.; Tang, C.; Wang, J.; Chen, C. Silencing of Soluble Epoxide Hydrolase 2 Gene Reduces H2O2-Induced Oxidative Damage in Rat Intestinal Epithelial IEC-6 Cells via Activating PI3K/Akt/GSK3β Signaling Pathway. Cytotechnology 2020, 72 (1), 23–36. (49) Girotti, A. W. Lipid Hydroperoxide Generation, Turnover, and Effector Action in Biological Systems. J. Lipid Res. 1998, 39 (8), 1529–1542. 215 (50) Li, J.; Cao, F.; Yin, H.-L.; Huang, Z.-J.; Lin, Z.-T.; Mao, N.; Sun, B.; Wang, G. Ferroptosis: Past, Present and Future. Cell Death Dis. 2020, 11 (2), 88. 216 APPENDIX 1: SUPPORTING INFORMATION FOR CHAPTER 2 Experimental and methods 1.1. Chemicals We procured the chemicals for our experiments, including PUFAs and their metabolites, from three reputable suppliers. The chemicals, their sources, and relevant information are detailed below: The following compounds were all purchased from Cayman Chemical Company (Ann Arbor, MI): 8,9-Dihydroxyeicosatetraenoic acid (8,9-DiHETE), (±)7,8-dihydroxydocosa-pentaenoic acid (7,8 DiHDPE), 19,20 DiHDPE, 16,17 DiHDPE, 13,14 DiHDPE, 10,11 DiHDPE, (±)7,8-epoxy docosapentaenoic acid (7,8 EpDPE), 13,14 EpDPE, 10,11 EpDPE, (±)5,6-dihydroxy- eicosatrienoic acid (5,6 DiHETrE), (±)5,6-epoxyeicosatrienoic acid (5,6 EpETrE), 11,12- dihydroxyeicosatetraenoic acid (11,12 DiHETE), 5-oxo-eicosatetraenoic acid(5 Oxo ETE), (±)9- hydroxy-10E,12Z-octadecadienoic acid (9 HODE), 13-keto-9Z,11E- octadecadienoic acid (13 oxo ODE), 9-oxo-octadecadienoic acid (9 oxo ODE), (±)5-hydroxyeicosatetraenoic acid (5HETE), 20 HETE, 11 HETE , 9 HETE, 12 HETE; (±)8 HETE, (±)17-hydroxydocosahexaenoic acid (17 HDoHE), 5S,15S-dihydroxyeicosatetraenoic acid (5,15 DiHETE), 8S,15S- dihydroxyeicosatetraenoic acid (8,15 DiHETE), (±)-12-hydroxyeicosapentaenoic acid (12 HEPE), 15 HEPE, 5 HEPE, 8 HEPE,12-oxo-eicosatetraenoic acid (12-oxo-ETE), 13S-hydroxy- octadecatrienoic acid (13-HOTrE), 15(S)HETrE, 5S,12R-dihydroxyeicosatetraene-1,20-dioic acid (20-COOH-LTB4),5S,12R,20-trihydroxy-6Z,8E,10E,14Z-eicosatetraenoic acid (20-OH-LTB4), 9S-hydroxyoctadecatrienoic acid (9(s)HOTrE), 12,13-epoxy-9-keto-10(trans)-Octadecenoic Acid (trans-EKODE-(E)-lb), Leukotriene-B3 (LTB3), LTB5, Leukotriene-E4 (LTE4), Prostaglandin-B2 (PGB2), Prostaglandin-D1 (PGD1),PGD3, Prostaglandin-E1 (PGE1), PGE3, Prostaglandin-J2(PGJ2), Resolvin-E1, Prostaglandin-F2α (PGF2α), Thromboxane-B2-d4 (TXB2-d4), Leukotriene-B4-d4 217 (LTB4-d4), 9S-hydroxy-10E,12Z-octadecadienoic-9,10,12,13-d4 acid- d4 (9 HODE-d4), 6-oxo- 9S,11R,15S-trihydroxy-13E-prostenoic acid (6 keto PGF1α). The following compounds were received as a kind gift from the MSU Mass Spectrometry and Metabolomics Core facility: Thromboxane B2 (TXB2), lipoxin A4 (LXA4), 17,18DiHETE, 8,9 EpETrE, 17,18 EpETE, 14,15 DiHETrE, LTB4, 11,12 EpETE, 8-iso-PGF2 α, 15 oxo ETE, 10,11 EpDPE, 19,20 EpDPE, 11,12 EpETrE, PGD2, docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA), arachidonic acid (AA), 12,13-dihydroxyoctadeca-9,15-dienoic acid (12,13DiHODE), 9,10 DiHODE, 15,16 DiHODE, acid, 1-cyclohexyl-dodecanoic acid urea (CUDA), 12,13- dihydroxyoctadecenoic acid (12,13 DiHOME), 9,10 DiHOME, 9,10 EpODE, 15,16 EpODE, 14, 15EpETrE, standards, 6-keto-PGF1a-d4, 5-HETE-d8, 8,9-EET-d11, AA-d8, 15(S)-HETE-d8, PGB2- d4, 8,9-DiHETrE-d11, 9-HODE-d4, LTB4-d4, PGE2-d9. These compounds were synthesized in-house, and the synthetic procedures are described in detail, please see our previous ACS Central Science publication1:14,15-Epoxyeicosadienoic acid (14,15 EpEDE), 8,9 EpEDE, 11,12- EpEDE, 14,15-di-hydroxyeicosadienoic acids (14,15 DHED), 8,9 DHED, 11,12 DHED. In our research, we encountered compounds with alternative names, which are EpETE also known as EEQ, EpETrE is referred to as EET, and EpEDE is alternatively called EED. Including this information in our study enhances clarity and ensures that readers can readily understand the nomenclature of these compounds. Phosphate buffered saline (PBS) was purchased from Gibco, (Grand Island, NY). Ethanol, isopropanol, methanol, ethyl acetate, acetonitrile, and glacial acetic acid (HPLC Grade) were purchased from Fisher Scientific (Pittsburgh, PA, USA). Oasis HLB 60 mg SPE cartridges were purchased from Waters Co. (Milford, MA). 218 1.2 Internal standards Two different types of compounds were used as internal standards to determine the recovery of extraction and instrumental variation. The following deuterated compounds were used as Type I internal standards: 6-keto-PGF1α -d4, 5-HETE- d8, 8,9-EET- d11, AA-d8, 15(S)-HETE- d8, PGB2- d4, 8,9-DiHETrE-d11, 9-HODE-d4, LTB4-d4, PGE2-d9. The type I internal standards were added to the samples before the SPE. The similar physical and chemical properties between the Type I internal standard and targeted PUFA metabolites help calculate the extraction recovery of prostaglandins, dihydroxy- and epoxy-fatty acid, and other oxylipins. In this regard, the analytes were linked to their corresponding Type I internal standards, based on their retention time, for quantification. A type II internal standard was added at the last step before injection to LC-MS/MS, to normalize changes in volume and other instrumental variations. A synthetic compound, CUDA which is a fatty acid metabolite mimic, was used as the type II internal standard (10 nM in 75% EtOH) (Appendix- Table 1). 1.3. Standard curve preparation A 12-step dilution from the original stock of a mixture of PUFA metabolites with known concentrations generated 10 concentration calibration standards in EtOH with a constant concentration of type I (Appendix Table 1), and type II CUDA (10 nM). Another calibration curve was made to calculate the extraction recovery with different concentrations of type I and constant concentrations of type II (10 nM). These calibration standards were kept in an amber vial (Fisher Scientific), sealed under argon gas, and stored at −80°C (figure 3 A). Calibration curves were plotted using a response factor of the analyte, which is the ratio of the analyte peak area to its 219 internal standard peak area, against the concentration of the analyst with 1/x weighting factors in the regression. Regression analysis yielded R2 values of 0.998 or greater for each analyte. Table S1. The name and concentration of internal standards in the stock solution and in the calibration curve Compound Function 6-keto-PGF1α -d4 10 5-HETE-d8 8,9-EpETrE-d11 AA-d8 15(S)-HETE-d8 PGB2-d4 8,9-DiHETrE-d11 9-HODE-d4 LTB4-d4 PGE2-d9 6 5 1 4 7 2 3 8 9 MRM 373>167 327.2>116 330>268 311>267 327>226 337>179 348.3>127 299>172 339>197 360>280 concentration stock (nM) 1000 Concentration calibration curve (nM) 40 1000 1000 1000 500 400 400 400 250 100 40 40 40 20 16 16 16 10 4 220 Table S2. Linear gradient chromatographic method. The mobile phase comprised 0.1% acetic acid in water (mobile phase A) and ACN (84%) and MeOH (16%) (mobile phase B), in a total run time of 21 min Time (min) initial %Aqueous phase (A) 65 %Organic phase (B) 35 Flow rate (ml/min) 0.25 1 3 8.5 12.5 15 16 18.1 60 45 35 28 18 0 65 40 55 65 72 82 100 35 0.25 0.25 0.25 0.25 0.25 0.25 0.25 Table S3. Different concentration categories of PUFAs metabolite (A-I) are based on the sensitivity of the instrument and the biological level of oxylipin. (12 step-dilution in the calibration curve). All concentrations in nM A 0.09 0.27 0.83 2.5 5 10 20 40 80 B 0.04 0.14 0.41 1.25 2.5 5 10 20 40 C 0.02 0.07 0.21 0.625 1.25 2.5 5 10 20 D 0.018 0.05 0.17 0.5 1 2 4 8 16 E 0.011 0.034 0.104 0.3125 0.625 1.25 2.5 5 10 F 0.008 0.026 0.08 0.25 0.5 1 2 4 8 G 0.005 0.0138 0.041 0.125 0.25 0.5 1 2 4 H 0.002 0.007 0.021 0.0625 0.125 0.25 0.5 1 2 I 0.0005 0.0014 0.0041 0.0125 0.025 0.05 0.1 0.2 0.4 221 Table S4. The list of compounds that were excluded from the paper because they were COX or LOX enzymes’ metabolites. Some compounds were excluded because the accuracy and precision were not in the acceptable range Analytes MRM Cone Volt. Collision Volt. RT LLOQ LOD ULOQ (min) (nM) (nM) (nM) 20-COOH-LTB4 365.10 > 52 195.00 6-Keto-PGF1α 369.30 > 21 163.00 20-OH-LTB4 351.10> 20 195.00 8-iso-PGF2α 353.20 > 44 PGE3 TXB2 PGD3 193.00 349.10 > 20 269.00 369.20 > 51 169.00 349.20 > 28 269.00 PGF2α 353.10 > 52 PGE2 PGE1 RV D2 PGD1 PGD2 LTD4 LXA4 PGB2 LTB5 LTE4 193.00 351.10 > 44 271.00 353.20 > 28 223.00 375.20 > 28 175.00 353.10 > 28 205.00 351.20 > 20 271.00 495.30 > 20 177.00 351.20 > 55 115.00 333.10 > 44 175.00 333.10 > 44 195.00 438.10 > 36 333.00 6-trans-LTB4 335.10 > 28 LTB4 LTB3 195.00 335.20 > 21 195.00 337.20 > 60 195.00 EKODE 309.10 > 20 209.00 15-deoxy-PGJ2 315.20 > 39 271.00 22 22 16 28 16 16 16 22 16 28 22 28 16 16 16 22 16 16 16 16 16 10 10 3.08 1.25 0.416 1000 3.22 0.625 0.208 500 3.25 0.250 0.125 100 3.79 0.312 0.104 250 3.82 0.208 0.069 500 3.85 0.625 0.208 500 4.00 0.208 0.069 500 4.23 0.250 0.083 200 4.32 0.125 0.042 100 4.43 1.250 0.416 500 4.44 0.625 0.312 250 4.5 1.250 0.416 200 4.53 0.250 0.083 100 4.86 0.625 0.208 500 4.90 0.312 0.104 250 5.74 0.312 0.104 250 5.78 0.250 0.083 200 5.87 1.250 0.416 500 6.66 0.250 0.083 40 6.97 0.250 0.083 200 8.44 0.250 0.083 200 8.90 0.500 0.167 200 9.35 0.083 0.028 100 222 Table S5. The oxylipin profiles of C. elegans at day 1 adult that is either treated with AUDA or vehicle (control). All concentrations are in (pmol/gworm) Control 1 Control 2 Control 3 AUDA 1 AUDA 2 AUDA 3 Compound 20-COOH-LTB4 6-Keto-PGF1α 20-OH-LTB4 8-iso-PGF2α PGE3 TXB2 PGD3 PGF2a PGE2 PGE1 RV D2 PGD1 PGD2 LTD4 LXA4 PGB2 LTB5 LTE4 6-trans-LTB4 LTB4 LTB3 EKODE 15-deoxy-PGJ2 8,15-DiHETE 19-HETE 15-HETE 12-HETE 5-HETE 15(S)-HETrE 20-HEPE 12-HEPE 10. The number of calibration standards decreased to eight after the limit of detection was stabilized. The data points in the calibration curve with S/N <10, and deviation >20% of the actual concentration were excluded from the calibration curve. We made sure to have at least five points for each compound calibration curve. The concentrations of each analyte in the calibration standard were calculated based on the constructed calibration curves, and the calculated concentration of the calibration standard with deviations more than 20% from the actual concentrations was also excluded. A linear plot of the relation between the concentration and peak area ratio was fitted using 1/x weighting factor linear regression. The correlation coefficient was higher than 0.996 for all oxylipins in the calibration curve (n=3), showing acceptable linearity of the assay in the selected calibration range. The analytes and IS peak retention time were stable between different analytical runs, a ±0.2-minute- window was considered as an acceptable difference in retention times. 230 Figure S3. Represents calibration curve of 17,18-EpETE, with correlation coefficient: r= 0.999493, r2= 0.998987, calibration curve equation:1.97235x+1.93413, response type: internal std, area*(IS conc. /IS area), curve type: linear, origin: include, weighting:1/x, axis trans: none 2.2. Accuracy and precision Quality control (QC) samples of each analyte were used to calculate the accuracy and precision of the method. These QC samples consisted of four different concentrations of oxylipins 0.5A (concentration A is defined in Appendix Table 3) as High QC, 0.04A as Mid QC, 0.012A as Low QC, and 0.0004A as LLQC, the QC samples were all diluted in PBS. Five replicates of each QC were extracted by SPE separately and samples were analyzed together with a complete set of calibration standards in three analytical runs, in one day to calculate the intra-day accuracy. An analytical run (n=1) of all QC samples on three consecutive days was used to establish the inter- day accuracy, which was determined as the percent difference between different days. The intra- day accuracy was determined as the percent difference between the mean concentration per analytical run and the expected spiked concentration. The coefficient of variation provided the 231 measure of intra- and inter-day precision. Method recovery determines the amount of analyte spiked in the matrix that can be recovered and quantified after SPE. The PBS solutions spiked with different analytes at a series of concentrations were extracted by SPE and analyzed by LC-MS/MS to determine recovery for each analyte. As mentioned in the calibration curve section, a separate calibration curve with different concentrations of Type I internal standard Table S6. Comparing the recovery percentage under hydrolysis conditions and normal conditions, n=3 Internal standard 6 keto PGF1α -d4 PGE2-d9 PGB2-d4 LTB4-d4 8,9-DiHETrE-d11 9-HODE-d4 15-HETE-d8 5-HETE-d8 8,9-EpETrE-d11 RC % normal condition (mean ± SEM) 72.7 ± 7.3 95.3 ± 7.8 101.4 ± 6.3 81.8 ± 4.8 108.7± 3.9 86.5± 2.5 96.2± 4.4 95.3± 3.9 76.5± 12.6 RC % under hydrolysis Condition (mean ± SEM) 78.2 ± 9.3 87.3 ± 6.7 107.1 ± 9.4 78.7 ± 6.4 112.2± 5.3 80.5± 5.2 89.4± 7.4 98.5± 6.2 82.5± 7.6 of any matrix effect. Conversely, A positive value indicates enhancement, while a negative value implies suppression or interference caused by the matrix. 3. C. elegans experiments 3.1. Epoxide hydrolase inhibitor supplementation To supplement with the epoxide hydrolase inhibitor, AUDA, we prepared a 20 mM stock solution of AUDA in ethanol. This stock solution was then added to the autoclaved NGM agar solution at a temperature between 55-65°C, resulting in a final concentration of 100 μM. The inhibitor solution was poured into the Petri dishes to make the final treatment plates. The plates were left at room temperature for one day and subsequently inoculated with 250-400 μL of E. coli OP50 (2.8 × 10^8 cells/mL). 232 3.2. Hydrolysis of esterified PUFA metabolites in C. elegans In this study, we conducted hydrolysis reactions in C. elegans worms to elucidate the total fatty acid metabolites present. For each of the three samples, a mixture consisting of 10 µl of antioxidants, 10 µl of (0.4A, concentrations are in Appendix Table 1), and 160 µl of PBS was prepared. The samples were then rapidly flash-frozen in liquid nitrogen, followed by homogenization. To initiate hydrolysis, 120 µl of 1.5 M KOH in 75% methanol-water solution was added. The samples were incubated for 30 minutes at 60 °C on a beads hot plate. Subsequently, 10 µl of acetic acid in 50% water was added to each sample to neutralize the pH, followed by the addition of 1 ml of PBS. After centrifugation, the supernatant was subjected to SPE for clean-up. The eluted fractions were collected, and the solvent was evaporated using a speed vacuum. The resulting residue was re-dissolved in 100 µl of 10 nM CUDA solution, followed by vertexing for five minutes. The prepared samples were transferred to the proper insert and vial, purged with Ar gas, and kept at -80° for subsequent injection to LC-MS/MS. 3.3. Age-synchronized worm The age-synchronized population was prepared by transferring healthy and well-fed day 1 adult worms to fresh nematode growth media (NGM) plates (50-100 worms/plate) with OP50. It took about 6-10 hours for the adult worms to lay eggs, and the eggs were hatched under isolation. About 36-48 hours later, we washed off the plates with s-basal solution and transferred them to a 40 mm cell strainer on top of a centrifuge tube (50 mL). We separated large-sized L4 larvae (stick to the filter), from larva, bacteria carryover, or other potential contamination (passed through the filter) by filtration. L4 larvae were then washed with 75-100 ml of s-basal, transferred to a 1.7 ml centrifuge tube, and centrifuged at 325 x g on a table-top centrifuge for 30 s. We removed the s- basal solution by aspiration, which left behind a pellet of L4. As the last step, L4 worms were 233 resuspended in s-basal solution and transferred to the supplemented or control plates seeded with OP50. The photo before and after filtration and the recovery of filtration are available in Appendix Figure 5. Figure S5. Filtration step to provide age synchronized worms, A) before filtration, B) after filtration C) filtration recovery % is mean ± SEM (n=5). Statistical differences between worm, larva and egg were evaluated by multiple Student t tests, p-value >0.0001 234 APPENDIX 2: SUPPORTING INFORMATION FOR CHAPTER 3 Dietary PUFA supplementation and oxylipin profile Our comparative analysis of PUFA concentrations in the serum of healthy individuals and SLE individuals with SLE has been reported, emphasizing the potential effects of dietary supplementation with fish oil, flax oil, or their combination. A notable finding is that healthy individuals show the highest DHA levels when consuming fish oil or a combination of fish and flax oil, indicating that these supplements significantly elevate serum DHA. SLE individuals with flax oil supplementation have substantially lower DHA concentrations than individuals with fish oil and a combination of fish and flax oil; the same pattern was observed in healthy individuals. Furthermore, the plasma ALA concentrations significantly increased in SLE patients who supplemented with a combination of fish and flax oil. Results for EPA and LA are not shown here since there was no statistically significant change. Comparing the serum concentrations of oxylipin metabolites between healthy individuals and individuals with SLE under different supplementations illustrates that while it is not statistically significant, the subjects taking flex oil or flex oil + fish oil supplement has a lower level of 14,15- DiHETrE, suggesting that the PUFAs from these oils are competing with AA for the same metabolomic pathway. Our data also indicated that flex oil supplement significantly increases the plasma concentration of ALA epoxy-metabolites, 9,10-DiHODE. However, such changes have not been observed in SLE subjects. Interestingly, we also observed that subjects taking flex oil supplement significantly downregulated the production of 17,18-DiHETrE in both healthy and SLE subjects. Nonetheless, flex oil + fish oil increases the plasma level of 17,18-EpETrE in healthy subjects as compared to other supplement groups. Interestingly, such change was not observed in SLE subjects. 235 Figure S5. Comparative analysis of metabolite concentrations in healthy individuals and SLE patients. This figure displays the concentrations of various PUFAs (DHA, AA, ALA ) and their metabolites (14,15-DiHETE, 9,10-DiHODE, and 17,18-DiHETE) under different conditions (no supplement, supplemented with fish oil, flax oil, and fish oil + flax oil). Data were analyzed using two-way ANOVA followed by Tukey's multiple comparisons test. P-value<0.05 *, P- value<0.01 **, P-value<0.001***, P-value<0.0001****. Mechanistic study offers a detailed examination of the role sEH plays in metabolizing DHA, specifically focusing on the effect of SLE on the conversion of EpDPEs to diols DiHDPEs. This model also fits adequately. The chi-square was 3.305 (df = 7, p =0.8554). The RMSEA was 0.000 and the CFA was 1. The data shows that the conversion of 10,11-EpDPE to 10,11-DiHDPE, while positive, is not statistically significant, suggesting that sEH's role in this pathway is not robust within the sample studied. (table 4 SI). 236 Table S6. Type I IS the name and concentration (nM) in the stock solution and the calibration curve. The function is related to mass spectrometry. We defined functions based on retention time with 10-15 oxylipin in them and included one IS in each function Compound Function MRM concentration IS 6-keto-PGF1a-d4 5-HETE-d8 8,9-EpETrE-d11 AA-d8 15(S)-HETE-d8 PGB2-d4 8,9-DiHETrE-d11 9-HODE-d4 LTB4-d4 10 6 5 1 4 7 2 3 8 373>167 327.2>116 330>268 311>267 327>226 337>179 348.3>127 299>172 339>197 stock 1000 1000 1000 1000 500 400 400 400 250 Concentration in calibration curve 40 40 40 40 20 16 16 16 10 PGE2- d9 360>280 100 9 4 Table S7. Performance of the algorithms SVM GBM GLMNET Accuracy Specificity Sensitivity 0.7447 0.9265 0.1154 0.7234 0.8677 0.3462 0.7128 0.9853 0.000 Random forest 0.7660 0.9853 0.1923 NNET 0.6915 0.9265 0.1154 Overall 0.7766 0.9559 0.3077 Table S8. Importance and weight of the algorithms classification model. Overall is the important score of algorithems which shows the relative performance of algorithms based on accuracy in table 4. Weight calculated based og overall importance score SVM GBM GLMNET RF NNET Weight 0.1061 0.2622 0.000 0.5163 0.1154 Overall 20.5571 50.7899 0.0000 100.0000 22.3551 237 Table S9. Performance of regression ML models GBM KNN SVM LM R-squared 0.005 0.072 0.044 0.168 XGB 0.082 GLMNET 0.043 FR 0.087 Overall 0.092 Table S10. PUFA pathway of DHA, Est : estimate of coefficients, se: standard error of estimate P-value Coefficients Est. se 10,11-EpDPE => 10,11-DiHDPE 13,14-EpDPE => 10,11-DiHDPE 19,20-EpDPE => 10,11-DiHDPE Group => OL10_11DiHDPE 10,11EpDPE * Group => OL10_11DiHDPE 13,14-EpDPE => 13,14-DiHDPE 10,11-EpDPE => 13,14-DiHDPE 19,20-EpDPE => 13,14-DiHDPE Group => 13,14-DiHDPE 13,14-EpDPE * Group => 13,14-DiHDPE 19,20-EpDPE => 19,20-DiHDPE 10,11-EpDPE => 19,20-DiHDPE Group => 19,20-DiHDPE 19,20-EpDPE * Group => 19,20-DiHDPE 10,11-DiHDPE with 13,14-DiHDPE 10,11-DiHDPE with 19,20-DiHDPE 0.205 -0.017 0.072 -0.006 0.077 -0.038 0.324 0.095 0.008 0.020 13.839 8.447 -8.085 -22.993 0.006 1.034 0.525 0.054 0.063 0.032 0.549 0.128 0.096 0.036 0.017 0.132 23.182 28.532 3.988 25.718 0.001 0.289 0.696 0.757 0.254 0.845 0.888 0.768 0.001 0.009 0.640 0.881 0.551 0.767 0.043 0.371 0.000 0.000 238