REVISED UNDERSTANDING OF LIPID OXIDATION MECHANISMS

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: AFRI COMPETITIVE GRANT
• Reporting Frequency: Annual
• Accession No.: 1011866
• Grant No.: 2017-67017-26465
• Cumulative Award Amt.: $454,735.00
• Proposal No.: 2016-09071
• Project Start Date: Jun 1, 2017
• Project End Date: May 31, 2021
• Grant Year: 2017
• Program Code: [A1361]- Improving Food Quality

Recipient Organization
RUTGERS, THE STATE UNIVERSITY OF NEW JERSEY
3 RUTGERS PLZA
NEW BRUNSWICK,NJ 08901-8559

Performing Department
Food Science

Non Technical Summary
This project seeks to provide a new base of information about how lipids oxidize and cause rancidity and degradation in oils and foods. Most attention for food safety and stability currently focuses on microbial contamination and inactivation. However, once bacteria, yeasts, and molds have been eliminated, chemical oxidation of lipids becomes the major driver of degradation in foods, leading to substantial loss of sensory quality and nutritional value during storage as well as to generation of potentially toxic products and outright food destruction as oxidation progresses. Oxidation problems were largely ignored during the no/low fat food era, but new recognition of important roles of lipids in health is forcing reformulation of foods with higher contents of polyunsaturated fatty acids. These physiologically essential fatty acids are highly oxidizable in themselves and they cause extensive co-oxidation of other food molecules, particularly proteins. Lipid oxidation causes significant loss of processed, stored foods as well as emergency food aid products, particularly those targeted to regions with high ambient temperatures. Lipid oxidation is also intimately involved in many pathologies in vivo, including aging, atherosclerosis, Alzheimer's disease, and cancer. Thus, understanding lipid oxidation reactions and being able to control these processes is critically important for health, for stabilizing foods, for reducing food costs and food losses, and for maintaining food safety. Unfortunately, basic information about how lipids oxidize is outdated and incomplete so stabilization efforts in the food industry have encountered many hurdles. This program previously identified several alternate pathways that make lipid oxidation much more complex than previously thought and explain some of the problems encountered in stabilizing foods. The current project extends these studies to additional lipids in more complex systems, identifying more classes of products, some of which have toxic potential. Results will provide the first new understanding of lipid oxidation in nearly 50 years, from which more effective measures to prevent this degradation can be developed. Learning how oxidation reactions shift under different conditions will make it possible to deliberately tailor food formulations, processing methods, packaging, and distribution practices to limit degradation of food flavors, textures, and nutritional quality, as well as reduce food loss. Learning which factors have strongest directing effects or most negative impact on product quality provides invaluable guidance for redesigning industrial processing, formulations, and packaging to stabilize highly oxidizable lipids. Learning what oxidation products to look for under various conditions guides development of improved analytical methods to accurately determine extent oxidation in both foods and tissues during manufacturing, storage, and distribution of foods. These outcomes should greatly reduce food loss during storage and distribution of regular and emergency food supplies world-wide. Optimized methods for quantitating oxidation in foods integrated in a battery of tests will generate the most accurate picture of degradation yet available and provide critically-needed tools for quality control in labs and on-line. While focused on food degradation and stabilization, the chemistry elucidated is applicable also to toxicology, medicine, plant physiology and pathology, and personal products and cosmetics industries.Detailed information about how fast lipids oxidize and what products are formed will be obtained in model systems of pure lipids to simplify analyses and make it possible to identify key products. Model lipids to be studied are methyl esters of linoleic and oleic acids, the most prevalent polyunsaturated and monounsaturated fatty acids in foods, and their oil counterparts trilinolein and triolein, respectively. Lipids will be oxidized first in simple model systems of oils alone, then progress to oils dispersed on glass fiber filter paper to mimic lipids in solid food matrices, and finally oils in emulsions to determine effects of water phases and provide increasing structure that more closely represents actual food systems. Model lipids will be oxidized in two modes (open, as would occur with permeable or no packaging, and closed, as would occur in sealed impermeable packaging), at three temperatures - 25, 40, and 60 deg C (room, elevated, and accelerated storage temperatures), under different levels of oxygen and other conditions. Samples will be withdrawn periodically and analyzed for products from multiple oxidation pathways, including conjugated dienes and hydroperoxides, epoxides, hydroxylipids, volatile and non-volatile aldehydes and ketones, dimers and polymers, and volatile products. Particular attention will be given to epoxides and hydroxylipids which have toxic potential. Class analyses will focus on quantitating how much of a specific product is generated under different conditions, while chromatographic analyses with mass spectrometry detection will identify structures of products generated. A very new method that uses liquid chromatography couple with high resolution mass spectrometry will be applied to oils for the first time to identify total lipid products, including structures not previously recognized. Product patterns and kinetics will then be integrated to derive active reaction pathways and assess potential approaches for inhibition as well as consequences of modification (will blocking one pathway accelerate another pathway and cause more but different damage?). Chemometric statistical analyses will be applied to sort out relationships between reaction conditions and dominant reaction pathways and products. The same analyses are being applied to all other foods being studied in our laboratory in order to determine how accurately the model systems represent oxidation in real foods.

Animal Health Component

5%

Research Effort Categories

Basic

85%

Applied

5%

Developmental

10%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
502	5010	2000	100%

Knowledge Area
502 - New and Improved Food Products;

Subject Of Investigation
5010 - Food;

Field Of Science
2000 - Chemistry;

Keywords

analysis of lipid oxidation

epoxides

lipid oxidation

oxidation mechanisms

oxidation products

Goals / Objectives
This project continues long-term research to more fully elucidate mechanisms of lipid oxidation and apply this understanding to foods. Lipid oxidation (aka rancidity) presents great challenges for protecting quality, stability, nutritional value, and chemical safety of foods. In traditional thinking, lipid peroxyl and alkoxyl radicals merely propagate free radical chains by hydrogen abstraction, generating hydroperoxides that decompose to a wide variety of products. However, this simple action does not explain the complexities of lipid oxidation kinetics and products. Evaluation of free radical reactions published in the literature identified rearrangement, addition, scission, and elimination reactions in addition to and in competition with hydrogen abstractions of lipid radicals, and a mechanisms integrating these alternate reactions with traditional hydrogen abstraction chain reactions was proposed. Initial research with polyunsaturated methyl linoleate as a model oil verified that alternate reactions indeed occur, even dominating under common conditions, and identified epoxides as new important products. The present study extends these studies, adding trilinolein and monounsaturated methyl oleate as model lipids to increase lipid complexity and including conditions with lipids oxidized on solid phases and in emulsions as well as in oils to further characterize factors that control dominant pathways and kinetics.Specific objectives include:1. Lipids - Extend studies with methyl linoleate to gain additional details of reactions and products not previously analyzed; add parallel studies with methyl oleate and trilinolein to determine effects of lipid structure. We are particularly interested in whether epoxide formation by LOO· addition to double bonds occurs as readily as in ML when only one double bond is present and when multiple acyl chains may impede access.2. Experimental matrices - Oxidize methyl linoleate, oleate, and trilinolein as neat oils and dissolved in solvents to compare base oxidation pathways. Extend oxidations of all lipids to solid matrices (coated on glass fiber filter paper) and emulsions to track pathways under conditions that are more comparable to food structures but still simple enough to track products easily.3. Product Analyses - Add analyses of hydroxyl lipids and dimers to previous analyses of conjugated dienes, hydroperoxides, epoxides, non-volatile carbonyls, and volatile products. Move from class analyses to more detailed component analyses for all products to more accurately track pathways and determine mechanisms. Add specific analyses for hydroperoxide position and geometric isomer configuration to track positions of radical attack and starting points for subsequent decompositions, add MS/MS analyses to verify product structures and identify new products.Specific questions to be addressed in evaluating products kinetics and distributions are:1) What role do hydroxy lipids play in the alternate pathways? LOH are written in as the sole product of hydrogen abstractions by LO· in traditional free radical chain reactions, but they are almost never measured in lipid oxidation analyses. Even if hydroxy lipids have no sensory or co-oxidation consequences in foods, they have been identified as important signaling agents in vivo43-46 and they also modify proteins.47-50 Hydroxylipids are less likely than hydroperoxides, aldehydes, and epoxides to be modified during digestion, so it is important to identify their levels in oxidizing foods as a first step, and then eventually investigate their potential contributions to toxicity of oxidized lipids.2) How do oxidation pathways change with lipid structure, i.e. number of double bonds, free fatty acid/ester vs acylglycerols vs phospholipid? In particular, do epoxides form in mono-unsaturated oleic acid as readily as in linoleic acid?3) How do oxidation pathways change with food matrix? Initial studies were only in methyl linoleate as a pure or diluted oil phase which models food oils but not more complex food matrices. We need to determine how patterns exhibited in neat ML change when lipids are dispersed on solid molecular matrices and in emulsions.4) Can enough detailed product and kinetic information be obtained with current analyses to provide proof for definitive new reactions? What new or more advanced handling and analytical techniques may be needed to sort out individual products? What statistical analyses may be useful for integrating products into reactions?5) What additional alternate pathways may be active, producing other product that have either been disregarded or not detected in previous analyses?6) How do product pathways vary in the presence of various catalysts, particularly light and metals?Products from multiple pathways (conjugated dienes, hydroperoxides, epoxides, alcohols, and carbonyls) will be analyzed as classes and individual products using chemical reactions, high pressure liquid chromatography and gas chromatography with mass spectrometry detection. Clarification of dominant products generated under various processing, formulation and storage conditions will guide more accurate analyses of lipid oxidation in foods, as well as more effective strategies for protecting food sensory quality and lipid nutrition while limiting toxicity. This project addresses Priority 2: Improving food quality (A1361) - Improve our knowledge and understanding of the chemical, physical, and biological properties of foods and food ingredients.This project addresses several current priority goals in the AFRI Foundational Program, Program Area C -- Food Safety, Nutrition, and Health, Program Area Priority 2- Improving food quality. 1)"Improve our knowledge and understanding of the chemical, physical, and biological properties of foods and food ingredients"; 2) use "Knowledge ... to improve the quality, shelf-life, convenience, nutrient value and/or sensory attributes of food; 3) "develop innovative technologies and materials for ... food quality monitoring; and 4) Apply "these technologies ... to improve food security by reducing post-harvest losses and waste of foods and precisely indicating and communicating the shelf-life of foods". This project specifically also addresses the challenge of ensuring industry capabilities to produce safe, high-quality food that is nutritious, convenient, and globally competitive, as well as long-term goals of the AFRI Food Security Challenge Area (to increase availability and accessibility of safe and nutritious food), the Food Safety Challenge Area (to protect consumers from chemical contaminants and toxicants that may occur during all stages of the food chain, from production to consumption), and the continuing theme of Global Food Security and Hunger.

Project Methods
Experimental Design. Investigations are divided into three stages. Stage 1 tracks lipid oxidation reactions in oils. Initial studies of neat methyl linoleate oxidation will be extended to include additional products (LOOH positions, hydroxy products and dimers), and to identify individual products in more detail by LC and GC with MS detection. Parallel studies following all the same conditions and product analyses will then be conducted on methyl oleate and acylglycerols triolein and trilinolein to determine how lipid structure affects oxidation pathways. Class analyses of products from different pathways will be completed first, followed by more detailed analyses of individual products. To provide information perhaps more relevant to actual foods, Stage 2 will extend oxidation studies to lipids dispersed on an inert solid matrix (glass fiber filter paper) using most of the same analyses and the same incubation variables as Stage 1. In Stage 3, studies will extend further to model lipids dispersed in emulsions, retaining as many of the base incubation variables as feasible. At each stage, test lipids will be oxidized at 25, 40, 60 deg C to model room temperature and accelerated oxidation, under various conditions of oxygen, water, solvent, and lipid concentration.Reaction Systems Neat oils. To document oxidation in oils and oil phases, methyl linoleate, methyl oleate, and their corresponding triacylglyderols are incubated in both open and closed systems at 25, 40, and 60 deg C protected from light. Samples are withdrawn daily for analysis of oxidation products for 30 days. To follow secondary processes in long-term oxidations, samples will be analyzed monthly intervals up to one year. Dispersed on filter paper. To determine dispersion effects on oxidation, test lipids are dispersed directly (no solvents) on glass fiber filter papers that have been soaked in 1 M HCl to extract trace metals, then washed extensively in 18 M-Ohm water to remove residual acid. Loaded filter papers will be incubated at 25, 40, and 60 deg C in glass crimp top GC vials sealed with PTFE (teflon)/butyl caps and protected from light. To determine effects of moisture, vials with samples are incubated open in desiccators equilibrated to 11, 45, and 90% relative humidity over salt solutions (LiCl, KCO3, and 16% NaCl, respectively), then oxidized at 25, 40, and 60 deg C. Moisture contents and water activities of equilibrated samples are determined using Mettler-Toledo infrared and Decagon chilled mirror dew point analyzers, respectively. For analysis of non-volatile products, lipids will be extracted from the filter papers in chloroform and analyzed directly when possible, or dried and resuspended in appropriate analysis solvents. Volatile products in headspace and SPME baseline studies will be analyzed directly in incubation vials for oils, but filter papers must be transferred to purge tubes for collection of volatiles on thermal desorption traps. In moisture effect studies, sample vials will be sealed after incubation in desiccators then allowed to equilibrate with headspace before static headspace or SPME analyses; filter papers will be transferred directly to purge tubes for thermal desorption analyses. Triplicate samples will be removed for analyses approximately daily for the first 30 days, then periodically thereafter for tests of long-term oxidation. Emulsions. To test effects of dispersion in water specifically on lipid oxidation pathways and products, first experiments will be conducted using the corresponding free fatty acids or monoglycerides of test lipids as emulsifiers and 18 M-Ohm water as the supporting phase. Dioleoyl or dilinoleoyl phosphatidylcholine may be tested if the fatty acids stabilize inadequately. O/W emulsions from each of the test lipids (methyl linoleate and oleate, trilinolein, triolein) in a range of lipid phase-volume ratios (10, 20, 40, 85% modeling low fat salad dressings, salad dressings, heavy cream or sauces, and mayonnaise respectively) are prepared, then characterized for particle size distributions using dynamic light scattering to ensure consistency in preparation. Emulsion samples are dispensed into headspace vials, incubated at 25 and 40 deg C, both open and closed, with shaking for varying periods of time, then oxidation products are analyzed as for oils neat and dispersed on solids. Volatile products are analyzed directly with headspace or SPME sampling, or by transferring known weights to purge and trap tubes. For non-volatile products, the lipid phase must be separated by either cold high-speed centrifugation or extraction into chloroform. The standard battery of product analyses described above will then be conducted on the isolated lipid phase.Product AnalysesNon-volatile products. To quantitate classes of products, conjugated dienes, hydroperoxides will be analyzed using standard chemical assays. Hydroperoxide positions will be determined from normal phase HPLC. Epoxides, carbonyl products, hydroxy lipids, dimers, and polymers (long-term) will be determined using new methods developed in out laboratory. To determine individual products and perhaps more importantly to detect products not analyzed in current assays, total individual products in oxidized test lipids will be separated by HPLC and structures determined by triple quadrupole mass spectrometry using electrospray ionization in negative ion mode (collaboration with John Newman, USDA-ARS, UCDavis, CA). Originally developed and validated for oxidation products in human plasma, this total lipid (oxylipin) technique has nM sensitivity. This will be the first application to oxidized food lipidsVolatile products. Combinations of static headspace, solid phase micro-extraction (SPME), and purge&trap/thermal desorption methods are used to collect volatile lipid oxidation products, which are then separated by gas chromatography with flame ionization and mass spectrometry detection. Structures are determined by comparisons to standards or by mass spec analyses. Peak identities are assigned using the National Institute of Standards and Technology (NIST) Mass Spectral Search Program Version 2.0f, 2008.Analysis, assessment, and interpretation of results. Interpretation of results requires substantial comparison of the sequences and kinetics of product generation, with integration of data from all products, both volatile and non-volatile, to identify active pathways and derive plausible reaction mechanisms. The task is facilitated by reporting all products on a common basis (molar rather than % by weight) and by intensive study of the lipid oxidation literature. We have begun building a database of lipid oxidation products reported in the literature, their conditions of generation, detection methods, and any reactions proposed so that we can reconcile our results with previous reports and explain differences, when they are observed. We also are exploring chemometric statistical analyses to sort relationships between products.Application/use of results. In basic research, we will use mechanisms observed to explain oxidation patterns and antioxidant effects (or lack thereof) observed in real foods under investigation, to identify additional issues for study, and to deduce innovative methods to block specific pathways and limit products with toxicity or strong sensory effects. To assess practical applicability, the same analyses used here will be applied to all other lipid oxidation studies with a variety of food and biological materials in our laboratory, and product distributions in the various systems will be compared to model systems of pure lipids.Efforts. This study will provide a wide range of laboratory instruction and experiential learning opportunities for graduate and undergraduate students. Results will be incorporated into lipid chemistry teaching at all levels in formal courses at the university, short courses, and seminars.

Progress 06/01/17 to 05/31/21

Outputs
Target Audience:The broad Target Audience for this entire grant has been anyone interested in understanding the multiple pathways of lipid oxidation, how to measure lipid oxidation products, and which products to measure. This includes primarily academics studying any aspect of lipid oxidation or oxidative stability of foods, food industry analytical and quality control labs, and government labs having any dealings with foods. To be more specific, we have given special attention to developing more sensitive assays for lipid hydroperoxides, epoxides, and carbonyl products that can be used easily in quality control labs testing lipid oxidation in oils and foods, as well as in basic research in lipid oxidation, elucidating product patterns in oleic and linoleic oils at different temperatures and in closed vs open systems that will be essential for industry labs conducting shelf-life studies and for new and established companies designing "stable" oils, tracking conditions that enhance epoxide formation, data that will be important to protein chemists and toxicologists concerned with co-oxidations causing damage in foods and biological tissues, tracking effects of antioxidants on oxidation pathways, information that will be critical for industrial chemists and food scientists seeking to improve antioxidant effectiveness. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?This project has provided technical training in lipid oxidation pathways and multiple product analyses, plus professional development in research logic and project management to one MS student and one PhD student working directly on this project, to one MS beginning volatile product analyses on this project even after it formally ends, and one MS student working on a closely related lipid oxidation project analyzing effects of temperature on product distribution in accelerated shelf life testing. Results of this study are incorporated into three courses taught by the Principal Investigator to keep future professionals aware of the most up-to-date results and give them new insights to carry into their companies. How have the results been disseminated to communities of interest?Various aspects of results were presented to an international audience at the American Oil Chemists' Society national meeting in six keynote papers on reactions and effects of lipid oxidation and three papers on lipid oxidation analyses (oxygen consumption, DNPH analyses of lipid carbonyls, and corona discharge detector applications for lipid analyses). What we have learned about measuring multiple lipid oxidation analyses was argued in an invited paper and part of what we learned about antioxidant shifts in oxidation pathways was included in a chapter on antioxidant actions. Results of this project were applied in consulting on oil stabilization with individual companies and were incorporated into food chemistry lectures on lipid oxidation. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? This project continues long-term research to determine how lipids with different compositions oxidize under different conditions and to apply this understanding to stabilizing foods. This information is critically needed because lipid oxidation (aka rancidity) presents the greatest chemical challenges for protecting quality, stability, nutritional value, and chemical safety of foods. Progress of this grant took us in some different directions than proposed, yet we have met 2 of the 3 original goals and provided new analytical methods and important new information about reactions and products that change our understanding of lipid oxidation. Objective 3. Improved analyses of lipid oxidation products. Because lipid oxidation creates so many different products, analytical methods have been the hinge pin of this project. This laboratory has led the field in upgrading existing methods and developing new methods necessary to track products from multiple oxidation pathways from early to late stages of oxidation. In addition, we designed our analyses to be quantitative as well as qualitative, to tell us not only what products were formed but how much of each so we could better judge their importance and impact. These analytical requirements all presented significant challenge and required much more effort than we originally anticipated but have generated methods that provide much more detailed information and will be useful to anyone analyzing lipid oxidation. Focus has been on standard methods that can be performed in most quality control or academic research laboratories to quantitate classes of products and also to identify at least some individual products. Increasing the emphasis on this Objective has enabled us to 1) replace previously-used methods that failed, became inaccessible, or had low sensitivity; 2) quantitate all known classes of oxidation products accurately at micromolar levels, a critical requirement for this project; 3) provide methods that are not readily available and are being requested by other researchers; procedures to detect trace levels of hydroperoxides, epoxides, and carbonyls are being compiled for validation as standard methods by the American Oil Chemists' Society; 4) directly track individual products and their changes in distributions as oxidation progresses, without complexation or modification; and 5) distinguish released oxidation products from those remaining on the triacylglycerol backbone. The CAD detector with normal phase HPLC now allows dependable analysis of hydroxylipids, dimers, and hydroperoxyepoxides which were previously not analyzed, as well as distinction between oxidation products and unoxidized lipids in single samples. Having all these assays for multiple products available is a major advance in the field. Objective 1. Effects of lipid structure and reaction conditions on oxidation pathways. Analyses we developed provided valuable new insights into effects of lipid structure and temperature on oxidation pathways and products for oleic acid and linoleic acid methyl esters and triacylglycerols. 1) Epoxides are critically important major lipid oxidation products, especially in oleic acid. Previous dogma held that lipids with more double bonds oxidized faster but not differently than mono-unsaturated fatty acids, and monounsaturated oils composed of oleic acid were stable. This led industry to move increasingly to monounsaturated oils developed from genetic and engineering modifications. In contrast, our data shows that oxidation pathways as well as rates change with numbers of double bonds. In particular, monounsaturated oleic acid forms epoxides often at levels higher than hydroperoxides while forming low levels of carbonyls. This is a groundbreaking observation because epoxides are highly reactive with proteins and are toxic. Current analyses of only hydroperoxides and volatile aldehydes miss these critical products altogether and underestimate lipid oxidation by 50% or more. Related to this, epoxides formed via lipid oxidation are different than those generated by chemical synthesis. The dominant epoxide structure detected in lipids appears to be terminal epoxides (at the end of a chain), which is also the most reactive form. These result from radical oxidation next to the epoxide, followed by chain scission to release the epoxide. Recently, increasing industry complaints have been aired regarding off-flavor development in monounsaturated oils. Our results suggest that these off-flavors may result from epoxide decomposition products which we have detected but not yet identified. Lipid epoxides have always been considered unimportant products, so their formation and degradation were never studied. Our results argue for intensive studies of lipid epoxide generation, structures, and transformation products under different conditions. Epoxides and their products then should be analyzed in parallel with hydroperoxides and volatiles on all lipid samples. Are there any protections against epoxides? We have preliminary evidence that mixed tocopherols divert early epoxides by quenching lipid peroxyl radicals before they add to double bonds and undergo scission to epoxides. If these observations hold, we have a potential treatment to minimize toxicity. 2) Lipids do not oxidize equally at all positions as traditionally understood. Hydroperoxides form close to equally at both ends of the fatty acid double bond system, but subsequent transformations to secondary products do not. Scissions to alkanes and aldehydes that consumers smell dominate at the terminal end while rearrangements to reactive epoxides prevail towards the acid end. Aldehydes traditionally expected are not always formed. 3) Hydroxylipids (lipid alcohols) do not appear to be major products, despite their portrayal as the second-most dominant product in traditional radical chain reactions. We find only low levels in all lipids and conditions tested. 4) Heat increases initiation of oxidation but also accelerates product transformation. As temperature increases, linoleate oxidizes more rapidly; major products reach maxima earlier but at lower levels due to more rapid conversion of conjugated dienes, hydroperoxides, and epoxides to other products (some carbonyls, some unidentified). Importantly, at room temperature epoxides are dominant products in oleic acid and remain stable at higher temperatures, while hydroperoxides increase due to enhanced initiation. 5) Closed systems (e.g. packaging) may limit oxygen but also trap oxidation products. Closed systems as in sealed, impermeable packaging delay onset of oxidation, but products transform more slowly and higher levels accumulate in the oils because they condense rather than evaporate. These observations raise questions about current industrial practices for packaging and shelf-life testing. More importantly, we discovered that oxygen diffusion from headspace into and through oil is very slow, and that oxygen always dissolved in oils is sufficient to drive active oxidation. 6) TAG vs esters. Triacylglycerols in oils oxidize more slowly than their corresponding monoesters because the higher viscosity slows oxygen permeation more than the three associated fatty acids enhance oxidation by closer contact. Product distributions also varied between TAGs and methyl esters. 7) Tocopherols block initiation as well as radical transfer. We expected that tocopherols would block epoxide formation by peroxyl radical addition to double bonds, and in the process increase hydroperoxide accumulation. Instead, we found that tocopherols did cause complete loss of epoxide but also greatly reduced LOOH by reacting also with radical initiators. Overall, results show that lipids do not oxidize in traditional expected patterns and argue strongly for mandatory analysis of lipid epoxides and a broad range of products to catch all oxidation pathways in research, quality control and shelf-life testing of oils and foods.

Publications

• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Schaich, K.M. 2021. Chemical accuracy vs expediency in [accelerated] shelf-life assays, AOCS National Meeting, Virtual, May 4, https://doi.org/10.21748/am21.113.
• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Kandrac, M. and Schaich, K.M. 2021. Charged aerosol detectors facilitate detection and quantitation of lipids and lipid oxidation products without chromophores, AOCS National Meeting, Virtual, May 7, 2021, https://doi.org/10.21748/am21.106.
• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Kandrac, M. and Schaich, K.M. 2021. Revised DNPH analysis of lipid carbonyls: reaction conditions for accurate quantitation and HPLC conditions for separation of individual carbonyls, AOCS National Meeting, Virtual, https://doi.org/10.21748/am21.570.
• Type: Book Chapters Status: Awaiting Publication Year Published: 2022 Citation: Schaich, K.M. 2022. Lipid antioxidants -- more than just lipid radical quenchers. In: Lipid Oxidation in Food and Biological Systems: A Physical Chemistry Perspective, C. Bravo-Dias, Ed., Springer Nature, Cham, Switzerland, in press.
• Type: Book Chapters Status: Published Year Published: 2017 Citation: Schaich, K.M. 2017. New Perspectives in Lipid Oxidation, In: Food Lipids-- Chemistry, Nutrition, and Biotechnology, Fourth Edition, C. Akoh, Ed., CRC Press, Boca Raton, FL, pp 481-499.
• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Schaich, K.M. 2021. Are lipid oxidation products consumed in foods toxic? if so, where? AOCS National Meeting, Virtual, May 3, https://doi.org/10.21748/am21.318.
• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Schaich, K.M. 2021. Lipid Oxidation: Where have all the products gone? A wholistic look at lipid oxidation. AOCS National Meeting, Virtual, May 5, https://doi.org/10.21748/am21.349
• Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Schaich, K.M. 2021. Transition metals shift products as well as alter rates in catalysis of lipid oxidation. AOCS National Meeting, Virtual, May 6, https://doi.org/10.21748/am21.06.
• Type: Conference Papers and Presentations Status: Other Year Published: 2019 Citation: Schaich, K.M. 2019. Lipid Oxidation  More than a simple free radical chain reaction, Korean Society for Food Science and Technology National Meeting, Seoul, Korea, June (invited keynote speaker for Lipid Science session).
• Type: Conference Papers and Presentations Status: Other Year Published: 2017 Citation: Schaich, K.M. 2017. How do foods go rancid and what can we do about it? Special seminar, Dalhousie University Dept. of Food Science and Engineering, Halifax, Nova Scotia, Canada,
• Type: Conference Papers and Presentations Status: Other Year Published: 2017 Citation: Schaich, K.M. 2017. Thinking outside of the box about lipid oxidation, DSM Corp, Halifax, Canada.
• Type: Conference Papers and Presentations Status: Other Year Published: 2017 Citation: Schaich, K.M. 2017. Analyzing multiple lipid oxidation products -- required, or not?, Amer. Oil Chemists Soc. Nat. Meeting, Orlando, FL.
• Type: Conference Papers and Presentations Status: Other Year Published: 2017 Citation: Kandrac, M., Izzo, C., and Schaich, K.M. 2017. Factors affecting the DNPH reaction with carbonyl products of lipid oxidation, Amer. Oil Chemists Soc. Nat. Meeting, Orlando, FL.
• Type: Journal Articles Status: Submitted Year Published: 2021 Citation: Schaich, K.M. 2021. Why is it important to look beyond peroxide values in tracking lipid oxidation? J. Agric. Food Chem.
• Type: Conference Papers and Presentations Status: Other Year Published: 2018 Citation: Schaich, K.M. 2018. Rethinking Basic Reactions of Lipid Oxidation and Antioxidants, invited paper, Inst. Food Technol. National Meeting, Chicago, IL, July.
• Type: Conference Papers and Presentations Status: Other Year Published: 2017 Citation: Schaich, K.M. 2017. Seton Hall University, Rose Mercadante Seminar Series, Dept. of Chemistry and Biochemistry, South Orange, NJ, Aug, 29.

Progress 06/01/19 to 05/31/20

Outputs
Target Audience:Target audience during this reporting period was anyone interested in understanding the multiple pathways of lipid oxidation, how to measure lipid oxidation products, and which products to measure. This includes primarily academics studying any aspect of lipid oxidation or oxidative stability of foods, food industry analytical and quality control labs, and government labs having any dealings with foods. Changes/Problems:Studies on dry surfaces and in emulsions will be abandoned because problems with analyses have limited available time. These results will be replaced by publication of detailed procedures for analyzing products from multiple pathways in lipid oxidation. What opportunities for training and professional development has the project provided?This project has provided technical training in lipid oxidation pathways and multiple product analyses, plus professional development in research logic and project management to two MS students and one PhD students. Methods developed in this project have been shared with an academic research lab in Basque, Spain and with two industrial companies. How have the results been disseminated to communities of interest?Results for methods developments were disseminated in one student seminar. Information learned about alternate pathways of lipid oxidation was integrated into graduate and undergraduate food science courses. What do you plan to do during the next reporting period to accomplish the goals?1) Complete comparisons of oxidation products from Me oleate and linoleate and their respective triglycerides at three temps in closed and open systems with and without tocopherols to determine differences in pathway preferences forced by structure, conditions, and antioxidants. Here the methyl esters provide simple systems that facilitate product separation, while the triglycerides show how oxidation may change when fatty acids are more organized in oils. These studies are of paramount importance in providing proof of principle for several observations. First, mono-unsaturated oils have been thought to have superior oxidative stability based on hydroperoxide values alone. Consequently, high oleic acid oils are being developed by genetic modification of plants and interesterification in processing. However, there is also data to suggest that oleic acid (one double bond) forms epoxides more readily than linoleic acid (two double bonds), so early reactions of these two fatty acids may proceed by different pathways. Epoxides are much reactive in modifying proteins in foods and are also more toxic to cells so it is critical to determine whether our focus on high oleic oils is misdirected. In these studies, we will be looking particularly for different balances between primary peroxyl radical addition to double bonds and rearrangement to epoxides, secondary scission of alkoxyl radicals to aldehydes vs rearrangement to epoxides, and conditions favoring production of hydroxyl lipids to determine how unsaturation directs oxidation in fatty acids. This, in turn, will provide new information about how oxidation analyses must be tailored for different lipids. Currently, epoxides are seldom measured and hydroxylipids are almost never analyzed. Another important application of these results will be in shelf-life testing. There currently is intense controversy over use of long-term testing under ambient conditions vs accelerated testing at elevated temperatures to assess system stability and antioxidant effectiveness. Our data from the previous study shows that product patterns and relative proportions vary tremendously with temperature, i.e. reactions are not the same, so hydroperoxide analyses alone do not provide an accurate picture of oxidation extent and high temperature incubations do not accurately model what happens at room temperature. We expect our extended data to provide clarification and guidance in this controversy. In these studies, we are analyzing non-volatile methyl ester oxidation products directly without derivatization on metal-free HPLC reversed phase columns with uv-vis and mass spec detection, and comparing these results with individual class analyses of hydroperoxides, epoxides, and carbonyl products, also by HPLC. After testing several published assays for hydroxy lipids, we have found none that is specific for hydroxyl groups so will track these products by direct analyses, comparing HPLC retention times with standards and verifying structure by mass spectra. We hope that advancing to mass spec detection will also facilitate identifying products previously not accounted for. We have found that hydrolyzing triglycerides before analyses changes oxidation products. Using an alternate approach, we will thus analyze oxidized OOO and LLL triglycerides directly by normal phase HPLC that allows large unoxidized hydrophobic molecules to wash off the column rapidly while it retains and separates core oxidation products on triglycerides as well as monomer degradation products. Because the corona discharge detector works most effectively with large molecules, this detector was transferred from the metal-free reversed-phase HPLC to the new normal phase system to detect products with weak or absent chromophores. Products will be tentatively identified by comparison of retention times to those of standards, with verification by mass spectrometry using instrumentation of Dr. Thomas Hartman in our department. 2) Using the same systems and analytical strategies as in 1) above, complete tests of antioxidant effects on product pathways. These have been added because some labs testing commercial oils were seeing no epoxides and others observed delayed production of epoxides. Additionally, there are well-known observations of apparent pro-oxidant actions of antioxidants at high concentration. All of these observations can be explained by shifting of oxidation pathways by phenolic compounds -they block peroxyl radical addition and accompanying epoxide formation by rapidly adding H atoms to peroxyl radicals to preferentially increase hydroperoxides (apparent pro-oxidant activity) in early oxidation, and they block alkoxyl radical scission to aldehydes and rearrangement to epoxides in secondary stages by rapidly adding H atoms to alkoxyl radicals to preferentially increase hydroxy lipids that are invisible in current assay practices. The first effect makes antioxidants appear pro-oxidant while the second effect shows them to be antioxidant at the same time when only hydroperoxides and perhaps aldehydes such as hexanal are measured. Since commercial oils and food products always have some antioxidants added, it is important to clarify antioxidant effects on product pathways in order to know which mixture of products must be analyzed to provide an accurate picture of oxidation. 3) Last fall, a closing industrial laboratory donated a used but highly functional GC-MS to our group to replace our GC that had been damaged by other users. A new student with GC-MS experience and supported by a university fellowship has joined our group and will be using this instrumentation to analyze and identify volatile products in parallel to the non-volatile product analyses in the lipid systems described above. Volatile products usually arise from fragmentation so will provide critical adjunct data for interpreting non-volatile products and deriving pathway mechanisms. 4) Write research papers.

Impacts
What was accomplished under these goals? This year continued our focus on refinement of analyses critically needed to provide full and accurate quantitative information about multiple lipid oxidation products at micro- to nano-molar levels. This unfortunately has left us behind schedule in planned experiments but it was necessary to fill in missing pieces of the product mix, particularly hydroxy-lipids (LOH), as well as to modify or replace analyses where we encountered problems that limited their use and accuracy. In addition, one HPLC died in the first year. Request was submitted then to USDA to reorganize the budget to purchase a normal phase HPLC that would facilitate analyses of triacylglycerols, but authorization for this was not received until Fall 2019. Another factor forcing this focus was increasing request for our methods from other academic labs and from industry. Our lab has been quite rigorous in testing all assays to make sure a) lipid oxidation products were not being created, destroyed, or modified by reaction conditions, b) results were reproducible and quantitative rather than just relative (higher or lower than other samples), and c) products could be detected at micromolar or lower levels in order to see initial reaction products and distinguish them from secondary products and establish kinetics for alternate pathways. Although we had analyses in place when the project began, when we "pushed the envelope", we found that changes were required. Since apparently few, if any, other labs are working to improve and extend analyses of lipid oxidation products, we felt that perfecting methods was warranted so they could be disseminated for general use. Efforts were concentrated on critical analytical capabilities. 1). Continued refinement of the triphenyl phosphine (TPP) assay for lipid hydroperoxides. In the previous year, this assay was tested for accuracy, limits of detection, linearity, and required reactant and reaction conditions using pure cumene hydroperoxide as a standard. During this year, we applied the assay to lipid methyl esters and refined procedures for analyzing lipids with unknown peroxide concentrations. This involves using a sufficiently high concentration of TPP and monitoring the HPLC peaks for TPP, its hydroperoxide reaction product TPPO, lipid hydroperoxids, and their reduced hydroxides for complete reactions. Although some TPP was lost by as yet unidentified side reactions, we found a direct correlation between hydroperoxide reduction and appearance of TPPO and lipid hydroxide peaks. Thus, the TPPO peaks can be used for hydroperoxide quantitation and the hydroperoxide and hydroxyl lipid peaks can be used to assure reaction was complete. These procedures are now in use in lipid incubations to follow a) hydroperoxides in methyl oleate and linoleate oxidation, and b) epoxide to hydroperoxide shifts formed in these oxidations by the antioxidant tocopherol. Publications describing the method are in preparation. 2) Continued testing of chemical assays for lipid alcohols. To find a class assay for hydroxyl lipids, we tested more chemical reactions for alcohols published in the literature, including chromic acid in sulfuric acid and zinc chloride in hydrochloric acid. While all reagents tested reacted with pure alcohols, we found two complicating factors that invalidated their use with oxidized lipids where mixed products were present: reaction conditions reduced hydroperoxides giving erroneously high levels of hydroxylipids, and reactions were not specific - nearly all lipid oxidation products reacted with th reagents. GC methods detecting HO-lipids were put on hold, but preliminary runs showed that lipid oxidation products changed during derivatization procedures. Thus, for accuracy, we are now opting for direct detection of HO-lipids by loading oxidized methyl esters onto a reversed phase or normal phase HPLC and tracking appearance of HO-lipids by retention time. Authentic HO-lipids are prepared by reduction of hydroperoxides by triphenyl phosphine (see pt 1, above). Here, HO adducts to methyl linoleate and both triglycerides can be detected easily with ultraviolet and corona discharge detectors, respectively. HO adducts on methyl oleate are detectable by uv as long as the double bond is retained. We are currently testing corona discharge detection of MO-OH in normal phase HPLC. if this works, We will for the first time be able to not only detect these products, but also to determine whether different positions on the fatty acids transform preferentially to hydroxyl-lipids as opposed to other products. This information will be extremely useful in designing more accurate analytical protocols for lipid oxidation. 3) Extending HPLC analyses of lipid epoxides and carbonyls to triglycerides. HPLC assays developed for epoxides and carbonyls optimized separation conditions for monomer products, and these procedures have been used routinely with analyses of methyl esters. However, analyses of triacylglycerols (TAGs) while also separating small monomer products require complex gradients and long analysis and re-equilibration times. While we were able to achieve required separations, the run time per sample was far too long to handle the large numbers of samples in each experiment. Separating triacylglycerols from monomers by solid phase extraction both changed lipid oxidation products and lost some products remaining on the glyceride core. Thus, we opted to shift these analyses to normal phase HPLC (silica columns) where hydrophobic unoxidized lipids will not be retained and polar interactions with the -OH groups will provide both separation by chain length and resolution of both positional and cis-trans isomers of these oxidation products. The latter will be especially important in providing new capabilities for differentiating hydroperoxide isomers. A normal phase HPLC was purchased and installed, and separation conditions for derivatized epoxides and carbonyls on both methyl esters and triglycerides were being tested when covid-19 forced lab closing for nearly four months. Since the corona discharge detector is more efficient with larger molecules, it was moved to this system to provide broader detection with increased sensitivity. We expect this system to become a major analytical workhorse for lipid oxidation analyses. 4) Electrochemical detection of lipid oxidation products in HPLC. Before this project began, we had successfully detected lipid hydroperoxides with a graphite electrode and expected to be able to extend electrochemical detection to other lipid oxidation products using a combination of oxidation and reduction modes. This approach would have been extremely useful in allowing individual product classes to be distinguished and identified as they were separated without derivatization. Despite the initial promise, we have encountered considerable difficulty in detecting products other than hydroperoxides. We think several factors contribute. The electrochemical detector on our HPLC only operates at one set potential, which means the potential must be known ahead of time. After failing to find published potentials for epoxides, carbonyls, and alcohols, we tried running standards over a range of potentials with several electrolytes but found no responses. We then converted the detector software to allow cyclic voltammetry analyses on standards to determine accurate redox potentials. Working with this has shown poor product reactivity at the graphite electrode or low sensitivity of the electrode. We are thus working with the detector manufacturer to assess whether platinum, gold, or silver electrodes will offer more appropriate potential ranges, better surface reactions, and greater sensitivity for detecting lipid oxidation products. In order to complete experiments, this method will be held for future investigations, being replaced instead with mass spectrum analyses by both direct injection and HPLC pre-separation of products.

Publications

• Type: Book Chapters Status: Published Year Published: 2020 Citation: Schaich, K.M. 2020. Lipid oxidation: New perspectives on an old reaction, In Baileys Industrial Oil and Fat Products 7th Edit., Shahidi, F., ed., John Wiley, New York, 1-88 (on-line version).
• Type: Book Chapters Status: Published Year Published: 2020 Citation: Schaich, K.M. 2020. Toxicity of dietary oxidized lipids, In Baileys Industrial Oil and Fat Products 7th Edit., Shahidi, F., ed., John Wiley, New York, 1-72 (on-line version)

Progress 06/01/18 to 05/31/19

Outputs
Target Audience: Nothing Reported Changes/Problems:As described in Results reported above, we have encountered major problems in specificity of hydroxylipid assays, in time required to run assays and separate products by HPLC, and in developing complex gradients to separate monomer and TAG products in a single run. We have explored many other options and are convinced that we need to move the HPLC analyses to normal phase systems so that unoxidized lipid components will not be retained, leaving the polar products behind to separate. Dr. Thomas Hartman in our department has recently been adding LC-MS to his GC-MS facility and we have been exploring with him how to identify and quantitate lipid oxidation products by MS, either by direct injection/electron ionization, or by pre-separation by LC and MS analysis of individual fractions as they elute. We petitioned USDA last year for authorization to reorganize the budget to purchase an HPLC system but received no responses to phone, email, or written requests. We are at a point that we can generate class information about product shifts in oxidation of fatty acids or esters, but this requires four separate HPLC systems to handle the major assays (LOOH, LOH, epoxides, carbonyls) simultaneously. We have only one HPLC for these analyses so are limited in the numbers of samples that can be handled per experiment. Beyond this, we need detailed information about individual products to track pathways. Hence, obtaining a normal phase HPLC to connect to corona discharge and MS detectors has become a critical requirement for definitive mechanistic studies. Normal phase is needed for removing unoxidized lipids and, in addition, the hydrophobic solvents (e.g. hexane) used provide cleaner backgrounds for both corona discharge and MS detection. LC-MS analysis will allow us to separate lipid samples directly and analyze products without derivatization, saving much time and providing more accurate product identification. Overall, we feel this change will greatly improve efficiency and trhoughput for the experiments. We are discussing possibilities for an instrumentation grant from ThermoFisher (Dionex HPLC systems) and will be submitting a formal request for project modification and budget reorganization to NIFA soon. What opportunities for training and professional development has the project provided?The project has provided technical training in lipid oxidation pathways and multiple analyses, plus professional development in research logic and project management to two MS students, two PhD students, and one international exchange student from Spain. In addition, it provided training in HPLC methods (in the contet of TPP assay development) to two undergraduate students, and it provided to one high school honors student a broad introduction to laboratory methods used to study lipid oxidation. How have the results been disseminated to communities of interest?Results for methods developments were disseminated in three seminars, public thesis/dissertation defenses, and published theses (2) and PhD dissertation (1). Information about alternate pathways of lipid oxidation was disseminated in graduate and undergraduate food science courses. What do you plan to do during the next reporting period to accomplish the goals?Complete comparisons of Me oleate and linoleate at three temps to determine differences in pathway preferences forced by structure. In particular will be looking for different balances between rearrangement to epoxides and scission to aldehydes. Complete tests of antioxidant ability to block peroxyl radical addition and accompanying epoxide formation, causing a preferential increase in hydroperoxides. Initiate tests of Fe3+ and Fe2+ on product distributions. Literature reports suggest that metal-LOOH interactions enhance epoxide formation. Test pathway shifts when methyl esters are oxidized on an inert solid matrix instead of in fluid form. Initiate studies with TAG oxidation. Write research papers.

Impacts
What was accomplished under these goals? This was a year of retrenchment to focus on development of critically needed analyses rather than outreach. Answering mechanistic questions in this project is totally dependent on being able to analyze multiple products - at least hydroperoxides (LOOH), epoxides, soluble carbonyls (RCHO), and alcohols (hydroxylipids, LOH) at micromolar or lower concentrations. We encountered three obstacles with these analyses - our oldest HPLC died completely, reagents for the commercial LOOH assay kits became undependable, and hydroxylipid assays were not ready. Hence, efforts were concentrated in four areas to provide critical analytical capabilities. Evaluation and development of hydroxylipids (LOH) assay. The literature on hydroxylipids assays was extensively and critically reviewed for sensitivity of detection, accuracy of analyses, likelihood of not changing lipid oxidation products during reaction, and ease in handling multiple samples. Analytical areas included titration, optical (test tube reaction), LC-MS, GC-MS, NMR, and IR. Two critical limitations of nearly all approaches were apparent. First, the -OH functional group has no optical properties and it hydrogen bonds to itself and other molecules so is not volatile, Hence, all analyses except NMR require -OH derivatization for detection. This approach is straightforward in simple reaction systems. However, triacylglycerols (TAGs) require hydrolysis to release fatty acids then hydrogenation to block double bonds before derivatization, and this process alters existing lipid oxidation products. In addition, tests with TAG hydrolyses before derivatization required times too long for handling multiple samples in shelf-life studies. Finally, we have yet to find a reagent that reacts specifically with LOH and does not cross-react with other lipid oxidation products, so once again the product levels reported cannot be attributed conclusively to LOH. This was particularly troublesome in tests of chlorosulfonic acid (CSA) complexation with LOH. Sulfated adducts of saturated lipid alcohols absorbed with optical maxima at 295 nm, but additional peaks above 400 nm were present in unsaturated LOH. To determine whether the CSA reacted with double bonds, we attempted to separate adducts by reversed-phase HPLC. We could find no columns or conditions to resolve the multiple products so analyzed the reactions directly by ion trap MS. We found that the 2% ethanol stabilizer in the chloroform solvent was swamping the reaction. MS analyses repeated after removal of the ethanol revealed that, contrary to literature claims, the CSA reacted with nearly every class of lipid oxidation product we tested. Thus, reaction with chlorosulfonic acid may work as a general assay for total oxidation, but not as a specific assay for LOH. To avoid side reactions and the need for separate LC systems for multiple product assays, we are currently focusing on separating products in oxidized methyl esters directly by HPLC in a metal-free system using retention times from authentic standards, reactions (e.g. reduction of LOOH to LOH with triphenylphosphine), and electrochemical plus optical detection to identify product classes. The corona discharge detector which we expected to detect and quantify products without specific absorption turns out to only function with large molecules. Hence, its use is postponed until studies with triacylglycerols. Full development of triphenylphosphine (TPP) reaction to replace xylenol orange assay and quantitate LOOH at nanomolar levels. For many years we have successfully used the PeroxySafe test marketed by MPBiomedical to quantitate LOOH at micromolar or lower levels. Recently, some key regents became difficult to obtain and we found the kits undependable. This forced us to give high priority to development of the TPP assay to track low levels of LOOH. The TPP reaction has long been used qualitatively to detect hydroperoxides in solvents. Tests of the reaction in solution for rapid assays showed spectral overlaps of the reagent and product too great for quantitation. We thus separated reaction components by HPLC and tested the assay for specificity, detection ranges, reproducibility, and accuracy using cumene hydroperoxide (99%) as a standard. We evaluated several types of HPLC columns for ability to completely separate reactants and products in less than 20 minutes, and we optimized reaction conditions and elution solvents for the fluorophenyl column, then applied these conditions to determine requirements for complete reaction and to test assays of LOOH in methyl linoleate. Current conditions appear to be accurate and robust and are in the process of repeat validations before using the assay in mechanisms studies. Transferring the reaction to HPLC provided an important advantage of being able to watch the disappearance of all reactants along with generation of final products, and hence to assure that the reaction is complete. Adventitiously, in working with methyl linoleate hydroperoxides, we were able to identify elution times for unoxidized ML, MLOOH, and the reduction product MLOH, and are using this information in direct HPLC analyses of underivatized products, particularly LOH. Refinement of dinitrophenylhydrazine (DNPH)/HPLC assay for lipid carbonyls. In the first year of this project we adapted the chemistry of the DNPH assay to make it fully quantitative as well as to separate and identify individual carbonyls for mechanism studies. While the assay was useful for class analysis, we needed to resolve the critical pairs to track individual oxidation pathways so this year we worked to resolved critical pairs (unsaturated aldehyde plus saturated aldehyde 2 carbons shorter) in the HPLC elution. We compared hydrazone separations on several different columns, with addition of formic acid and other solvent modifications, and with complexing agents. We can now distinguish separate components enough to quantitate. 4. Extension of solvent gradients to include detection of products remaining on triacylglycerol cores. In assays developed for epoxides and carbonyls, separations were optimized for monomer products. Such procedures work fine for model systems based on methyl esters. However, we need to move the experiments to triacylglycerols (TAGs) in this project and also want our methods to be useful for general analyses in research and quality control labs. Thus, we must adapt procedures to mixtures of TAGs and small products. This requires complex gradients going from rather polar solvents (elute oxidation products) to very nonpolar solvents (elute unoxidized TAGs), then adding time for column re-equilibration to original conditions. This makes for very long run times per sample and severely limits the number of analyses that can be handled daily. Under these conditions, we found it difficult to achieve adequate resolutions of TAGs. An alternate approach has been to separate monomers from TAGs by solid phase extraction, then analyze monomer and TAG fractions separately on two different columns. However, this process modifies products during handling, so we choose to avoid it. The best option we see to improve removal of unoxidized lipids (the largest component of oils and extract, by far) and then separate oxidized products based on their polarity is to convert the separations to normal phase HPLC (silica columns). Hydrophobic unoxidized lipids will not be retained on these columns and polar interactions with the -OH groups will improve resolution of the various classes of polar products beyond chain length. We have been looking for normal phase columns with 90-120 Å pore size, as well as a normal phase HPLC system (new or used). Gaining this analytical capability will be critical for extending studies with methyl esters to more complex lipids, so it will become a key focus in the coming year.

Publications

• Type: Theses/Dissertations Status: Published Year Published: 2019 Citation: Izzo, C. 2019. Chris Izzo. Near Infrared spectroscopic investigation of lipid oxidation in model solid food systems, Ph.D. dissertation, Rutgers University, Dept of Food Science, New Brunswick, NJ.
• Type: Theses/Dissertations Status: Published Year Published: 2019 Citation: Palaniswamy, Indumathi K. 2019. Hydroxylipids: Significance and analytical methods, M.S. thesis, Rutgers University, Dept. of Food Science, New Brunswick, NJ.
• Type: Theses/Dissertations Status: Published Year Published: 2019 Citation: Yeager, Z. 2019. Evaluation of the triphenylphosphine assay for quantitating hydroperoxides in oxidized lipids. M.S. thesis, Rutgers University, Dept. of Food Science, New Brunswick, NJ.

Progress 06/01/17 to 05/31/18

Outputs
Target Audience:Target audiences wereprimarily academic food and lipid research communities as well as the food industry, but results are also relevant and applicable to personal products industries and physiological toxicology. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Laboratory training in analyses of multiple lipid oxidation products was provided to three Masters graduate students. One of these students will continue to a PhD in this project, applying revised analyses to track and compare detailed oxidation pathways in methyl linoleate and oleate and their triglyceride forms. Laboratory training inoptical analyses, reaction chemistry, HPLC analyses, and research ethodswas provided to one undergraduate student who worked on revisions of the diethyldithiocarbamate complexation of epoxides. Professional development was provided throughresearch results incorporated into a graduate food chemistry course, lipid oxidation seminars presentedto two universities, and workshops presented to two international companies. How have the results been disseminated to communities of interest?Results of this project have been disseminated by students in department seminars and by Project Director in industrial seminars, seminars at other universities, and papers presented at American Oil Chemists' Society and Institute of Food Technologists national meetings. What do you plan to do during the next reporting period to accomplish the goals?Experimentation will follow the general plan outlined in the grant proposal with afew modifications: 1) Compare oxidation pathways in methyl linoleate (two double bonds) and methyl oleate (one double bond) in lipids incubated as pure lipids and coated in films on glass fiber solid supports. Look particularly for shifts in epoxide production (less in oleic acid, more in films). In these tests, use separate analyses of each lipid product. Include both volatiles and non-volatile products. 2) Test approach of determining total lipid products simultaneously by HPLC with multiple detectors. Use standards for preliminary identification; verify structures by external LC-MS analyses. 3) Test ability of phenolic antioxidants (strong H atom donors) to shift oxidation pathways, e.g. to prevent epoxide formation but enhance hydroperoxide and hydroxy lipid formation using the same procedures as above.

Impacts
What was accomplished under these goals? Lipid oxidation, or rancidity,is the major chemical reaction limiting quality and shelf lifeof foods after processing and during storage. The reaction was shown to be a free radical chain reaction over 70 years ago and this scheme has been used to develop analyses, interpret oxidation data, and design antioxidant strategies. However, too often the simple radical reactions do not explain observed reaction kinetics or products, and antioxidants appear to be pro-oxidant. Free radical chemistry has identified additional reactions of free radicals that compete with traditional hydrogen abstraction and alter sequences and rates of lipid oxidation. These alternate reactions have been extensively documented in model systems but have seenonly limitedapplication tolipids, especially food lipids. This project has integrated known alternate reactions into what is hoped will be a useful total reaction scheme that can be used practically to guide analyses and product interpretations to more accuratelydocument the extent and direction of lipid oxidation. Results of the previous grant on this topic began documenting the actions of alternate pathways and stimulated particular interest in measuring epoxides in both academic studies and in industrial analyses. Recommendations have now been made to include epoxide and hydroxyl values along with peroxides in routine lipid analyses. This is clearly a step forward, but general analyses of different classes do not go far enough. Information about individual products is needed in basic research to determine reaction sequences. There are several importantconsequences of having alternate reactions active: 1) the hydroperoxides and volatile products commonly measured can miss a large proportion of products and greatly underestimate lipid oxidation, 2)products from all pathways need to be measured, but sensitive assays are not always available and running multiple individual analyses is very time-consuming,3) reactions of antioxidants with different products shift active pathways and products, so can distort the picture of lipid oxidation presented, either under- or over-estimating lipid oxidation, and 4) productsfrom different pathways (e.g. epoxides, hydroperoxides, alcohols, aldehydes) have different potential for co-oxidation of other molecules in foods, broadcasting damage to other molecules and amplifying quality degradation, and for toxicity. These consequences and the need to elucidate and document specific pathways underlie the goals of this project -- to develop required assays of multiple lipid oxidation productsand apply them to document which alternate pathways are most active and understand how the balance between oxidation pathways changes under different reaction conditions and with different lipids or catalysts. Our previous project made important discoveries regarding production of epoxides but also encountered limitations with analytical methods. Thus, major efforts this year focused on refining and improving detection methods for individual products to facilitate planned comparison of oxidation in methyl linoleate and oleate. Particular attention was given to carbonyl and epoxide assays in light of reports indicating that oleic acid undergoes cyclization to epoxides less readily but scission to carbonyls more rapidly than linoleic acid. Also, removal of our previous HPLC reagent from the market necessitated evaluation of different methods that would be both accurate and sensitive at the submicromolar level.A new triphenylphosphine method resulted from these efforts. Later return of the xylenol orange hydroperoxide reagent to the market, modified in undisclosed ways, then necessitated a re-evaluation of the new kit. We have altered supplier-recommended conditions to obtain reproducible results with accurate stoichiometry, and replaced unstable calibratorswith atomic absorption standard Fe(III) to generate standard curvesand accuratelydetect side reactions. Both of these assays are ready to apply to pending lipidoxidation studies. The DNPH assay we previously used was found to underestimate lipid carbonyls. The high acid conditions normally used in the assay protonated DNPH and catalyzed carbonyl condensation, both of which diminished levels of hydrazones detected. Effects of acidity, DNPH concentrations, andreaction times were tested and optimum pH that gave complete reaction of both saturated and unsaturated aldehydes with minimal isomerization were identified. HPLC columns were changed and elutions conditions were modified to obtain full resolution of all aldehydes. We had hoped to adapt the diethyldithiocarbamate (DETC) assay previously developed for HPLC to a solution assay that could easily quantitate all epoxides as a class in oils or extracts. Tests in solution, however, showed that the there was too much overlap in optical spectra of initial DETC and product complexes and adding phosphoric acid did not decompose the DETC as claimed in the literature. However, these studies revealed an intermediate in the acid degradation that could be used to follow thefull reaction -- loss of starting DETC and epoxide plus appearance of intermediates and product complexes -- by HPLC. This provided a means of refining previous incubation times and temperatures to ensure complete and equal reaction of all epoxides regardless of structure. A new metal-free HPLC system was purchased and installed. Metal-free was necessary to avoid decomposition of lipid hydroperoxides as they passed through the column and lines. The system is equipped with three detectors -- uv-vis, electrochemical, and corona discharge -- to maximize characterization of lipid oxidation products without mass spectrometry detection. Standards for hydroperoxides, aldehydes, alcohols, and epoxides of various structures and chain lengths are being tested to determine response characteristics (potentials, wave forms, and intensities) with the electrochemical detector. Responses of all standard series are also being recorded with the uv and corona discharge detectors (detects all products,quantitates bymass)to build a library from which peaks can be tentatively identified. This is being done with three columns -- standard C18 reverse phase with solid particles, reverse phase with coated particles, and HILIC -- to determine conditions that will provide maximum resolution of all products classes from fatty acids. The intention is to inject oxidizing methyl linoleate or oleate directly and separate all products in a single run to obtain tentative product distributions.

Publications

• Type: Book Chapters Status: Awaiting Publication Year Published: 2019 Citation: Lipid Oxidation: New Perspectives on an Old Reaction, in Baileys Industrial Oil and Fat Products (7th Edition), Wiley Publishers, Hoboken, NJ.
• Type: Theses/Dissertations Status: Submitted Year Published: 2018 Citation: Factors affecting the 2,4-dinitrophenyl hydrazine reaction with lipid carbonyls, Morgan Kandrac, M.S. Thesis, Dept. of Food Science, Rutgers University, New Brunswick, NJ.