Polyunsaturated Fatty Acids: Conversion to Lipid Mediators, Roles in Inflammatory Diseases and Dietary Sources

Polyunsaturated fatty acids (PUFAs) are important components of the diet of mammals. Their role was first established when the essential fatty acids (EFAs) linoleic acid and α-linolenic acid were discovered nearly a century ago. However, most of the biochemical and physiological actions of PUFAs rely on their conversion to 20C or 22C acids and subsequent metabolism to lipid mediators. As a generalisation, lipid mediators formed from n-6 PUFAs are pro-inflammatory while those from n-3 PUFAs are anti-inflammatory or neutral. Apart from the actions of the classic eicosanoids or docosanoids, many newly discovered compounds are described as Specialised Pro-resolving Mediators (SPMs) which have been proposed to have a role in resolving inflammatory conditions such as infections and preventing them from becoming chronic. In addition, a large group of molecules, termed isoprostanes, can be generated by free radical reactions and these too have powerful properties towards inflammation. The ultimate source of n-3 and n-6 PUFAs are photosynthetic organisms which contain Δ-12 and Δ-15 desaturases, which are almost exclusively absent from animals. Moreover, the EFAs consumed from plant food are in competition with each other for conversion to lipid mediators. Thus, the relative amounts of n-3 and n-6 PUFAs in the diet are important. Furthermore, the conversion of the EFAs to 20C and 22C PUFAs in mammals is rather poor. Thus, there has been much interest recently in the use of algae, many of which make substantial quantities of long-chain PUFAs or in manipulating oil crops to make such acids. This is especially important because fish oils, which are their main source in human diets, are becoming limited. In this review, the metabolic conversion of PUFAs into different lipid mediators is described. Then, the biological roles and molecular mechanisms of such mediators in inflammatory diseases are outlined. Finally, natural sources of PUFAs (including 20 or 22 carbon compounds) are detailed, as well as recent efforts to increase their production.

The production of lipid mediators from PUFAs can be considered in two main sections. First, there is the conversion of EFAs into classic eicosanoids. This aspect has been covered in some detail recently [ 18 , 19 , 20 ] and, in particular, the molecular mechanisms of the three types of oxidases involved have been described in detail [ 21 ]. Second, there is the production of specialised pro-resolving mediators (SPMs) which can be formed by both enzymatic and non-enzymatic PUFA oxidation [ 11 , 22 ]. 2.1. Eicosanoid Biosynthesis The biosynthesis of the classic eicosanoids (20C oxidised derivatives of PUFAs) is due to the action of three different types of oxidation. The enzymes involved are the cyclooxygenases, lipoxygenases, or the cytochrome P450 oxidase/epoxygenases [ 18 , 19 , 21 , 23 ]. All these reactions utilise a non-esterified PUFA which is usually released from a membrane phosphoglyceride by phospholipase A2 (PLA2) action. A number of PLA2 enzymes have been characterised [ 24 ]. These can be categorised into cytosolic Ca-dependent (cPLA2), cytosolic Ca-independent (iPLA2) and secreted PLA2 (sPLA2) groups. Moreover, the main substrate fatty acids (ARA, EPA, DHA) have different phospholipid sources [ 18 ] and, to an extent, different activities with the various hydrolytic enzymes [ 25 ]. Further details of the physiological interactions that influence the PLA2-mediated release of PUFAs from membrane lipids may be found in [ 18 , 19 ]. Oxidation of the 20C PUFA (usually ARA or EPA) can be through cyclooxidase action [ 26 ]. There are two major human cyclooxygenase isoforms (COX-1, COX-2) which differ in their expression and substrate selectivity [ 18 ]. COX-1 is expressed constitutively in most mammalian tissues while COX-2 has a characteristic low expression unless induced by inflammatory stimuli such as cytokines or tumour promoters [ 27 , 28 ]. The main exceptions to this are that both isoforms (COX-1, COX-2) are expressed constitutively in the brain, testes, and the macular densa of the kidney [ 29 ]. Two reactions are catalysed by both COXs. In the first, two molecules of oxygen are added to the substrate by a cyclooxygenase reaction. For ARA this leads to prostaglandin PGG2. The latter intermediate is then reduced during a peroxidase cycle to PGH2. The COX enzymes are more correctly called prostaglandin endoperoxide H synthase. Although COX-2 has a wider substrate selectivity (including n-3 PUFAs) than COX-1 [ 27 , 28 ], ARA is still the preferred substrate for both [ 30 ]. Once the endoperoxide intermediate has been produced by COX activity, it can be converted into three classes of prostanoids: prostaglandins (e.g., PGD2, PGE2, PGF2), prostacyclin (PGI2), or thromboxane TXA2—all from PGH2 ( Figure 1 ). PGI2 promotes vasodilation, inhibits platelet aggregation, and contributes to myocardial protection. In contrast, TXA2 promotes platelet aggregation [ 31 ]. Since these two prostanoids have antagonistic effects, their balance is important for vascular homeostasis and healthy cardiovascular function [ 32 ]. Notably, prostanoid synthases have isoforms with variable distributions in different tissues. Thus, the eicosanoid composition produced in individual tissues can vary significantly [ 19 ]. The isolation, purification, and characteristics of enzymes involved in the production of prostanoids via COX activity are detailed in [ 33 ]. In addition, details of the structure of COXs as well as their catalytic mechanisms and substrate selectivities are described in [ 21 ]. COX-1 and possibly COX-2 can undergo an irreversible substrate-independent self-inactivating process [ 34 ], possibly due to radical intermediates formed in either of the two partial reactions [ 33 ]. All the lipid mediators described here have local and normally transient actions in tissues. For prostanoids, they can be metabolized to other less-active metabolites ( Figure 1 ) or, in the case of 15d-prostaglandin J2, they can bind to PPAR-γ [ 35 ]. In turn, such binding can have anti-inflammatory and anti-oxidant actions e.g., via inhibition of NFκB and JAK-STAT pathways [ 36 ]. For many prostaglandins, 15-PGDH (15-hydroxyprostaglandin dehydrogenase) is a key enzyme. It causes the rapid oxidation of 15-hydroxy-prostaglandins into their less active 15-keto derivatives. It is ubiquitously expressed in mammalian tissues and is particularly important for PGD2, PGE2, and PGF2α catabolism. Lipoxygenases (LOXs) can catalyse three different types of reactions: hydroperoxide production via dioxygenation, hydroperoxidation which yields keto lipids, and a leukotriene synthase reaction to give epoxy leukotrienes [ 18 ]. They have also been proposed to be involved in the biosynthesis of SPMs. In humans, six main families of LOXs have been identified—5-LOX, 12-LOX, 12/15-LOX (15-LOX type 1), 15-LOX type 2, 12 R -LOX and epidermal LOX [ 37 ]. Traditionally, LOXs have been named according to their positional specificity for arachidonic acid so that

Recently, a multi-author publication [ 81 ] has questioned some aspects of the formation and function of SPMs. This is a controversial area which is discussed in detail at the end of Section 6.3.3 . Because the disagreements have not yet been resolved, it should be mentioned at this point that they have implications for Section 3 and Section 6 . Specialised pro-resolving mediators are so called because of their proposed function in the resolution stage of inflammation and the first representative was isolated some 20 years ago [ 82 ]. Inflammation itself is a natural protective process where the body responds to harmful stimuli such as pathogens or wounding. It functions to clear out damaged tissues and necrotic cells as well as eliminate the initial cause. Acute inflammation is actively terminated by its resolution which has been proposed to involve several mechanisms including the production of SPMs, as well as the down-regulation of pro-inflammatory substances (such as leukotrienes). If resolution does not occur, chronic inflammation will ensue with severe consequences [ 11 , 22 ]. As mentioned above, the first SPM which was isolated was a resolvin formed from EPA [ 82 ]. Four individual resolvins (E-series resolvins e.g., RvE1) can be formed from EPA and have been characterised [ 83 ]. Six D-series resolvins are formed from DHA (RvD1, RvD2 etc.) [ 84 ] while n-3 docosapentaenoic acid can form D-series resolvins (e.g., RvDn-3DPA) as well as the 13-series resolvins (RvTs) [ 11 , 22 ]. Some details of the enzymatic reactions and molecular details for the generation of E-resolvins via COX-2 and CYPs have been discussed in [ 21 ]. 18 R -HEPE can be converted by 5-LOX into a 5 S -hydroperoxy intermediate and then to a 5 S ,6 S -epoxyintermediate [ 85 ]. The hydroperoxyl intermediate can be peroxidised to resolvin E2 while the epoxy intermediate can be hydrolysed by LTA4H to resolvin E1 [ 86 ]. 18S-HEPE has also been proposed to form 18 S -analogues of resolvin E1 and E2 [ 21 ]. It is, perhaps, ironic that aspirin (which is used widely along with other NSAIDs to inhibit COX activity) [ 87 ] should have proven useful in the discovery of many SPMs. Thus, aspirin treatment can lead to the formation of 17 R -resolvins from DHA as well as the aspirin-triggered resolvins AT-RvD1 and AT-RvD2 [ 18 ]. The discovery of these reactions [ 82 ] and further details are given in [ 88 ]. As an alternative to resolvins, DHA has been proposed to be modified by 15-LOX to generate protectins or by 12-LOX to form maresins. The first protectin (NPD1) was isolated from the brain and was termed neuroprotection because of this and its protective action [ 84 ]. However, because similar docosanoids have now been reported in many tissues, the term ‘protectin’ is now preferred for their other peripheral actions [ 11 , 18 ]. Aspirin irreversibly inhibits COX-1 at its active site but only partly blocks that of COX-2 and modifies its activity. The latter can show activity similar to 15-LOX but with a difference in the configuration for oxygen insertion [ 18 ]. The aspirin-triggered epimer, 17 R -NPD1 shares the actions of NPD1 by attenuating stroke, controlling polymorphonuclear leukocytes and facilitating macrophage activity [ 11 ]. A recent review has provided more details about the biosynthesis, catabolism, and structure-functions of the protectins [ 89 ]. Particular emphasis has been placed on structure-function data for sulfido-conjugated protectins as well as the catabolism of protectin D1. It was pointed out that by 2022 some 43 SPMs possessing pro-resolution and anti-inflammatory reactions had been reported. Because mediators such as PD1 act locally, they then have to be catabolized rapidly. The role of β-oxidation in this process has been detailed [ 90 ]. Maresin synthesis has been proposed to begin with 12 S -LOX generating 14 S -hydroperoxy-DHA, followed by an epoxy-intermediate. The epoxy-intermediate can then be hydrolysed by different enzymes [ 21 ] to generate maresins 1 or 2 [ 91 , 92 ]. Recently, three series of peptide lipid-conjugated SPMs have been reported. These are conjugates with maresins, protectins, or resolvins and have been termed cys-SPMs [ 93 ]. Each series has three members, all with potent pro-regenerative and pro-repair actions [ 11 ]. Data from detailed structure-function analyses of these sulfido protectin conjugates as well as some protectin analogues have been published [ 89 ]. In contrast to the SPMs formed from n-3 PUFAs, ARA can be converted to lipoxins, which are trihydroxyeicosatetraenoic acids [ 39 , 94 ]. There are at least three routes for the biosynthesis of lipoxins. LXA4 and LXB4 can be generated by 5-LOX action on 15 S -HPETE [ 95 ]. Alternatively, a transcellular pathway using 5-LOX in one cell and 12-LOX in another has been reported [ 96 ]. A third mechanism produces epi-lipoxins (or aspirin-triggered lipoxins; ATL) using an aspirin-modified COX-2 which forms 15R-HPETE and then 5-LOX continu

A major group of compounds was discovered much later than the eicosanoids when the auto-oxidation of ARA by reactive oxygen species (ROS) through free radical reactions was characterised [ 110 ]. These compounds are termed isoprostanes, of which phytoprostanes are produced by ROS action on ALA in plants and neuroprostanes are formed by free radical-mediated peroxidation of DHA [ 111 ]. As noted above, the earliest studies on isoprostanes (IsoPs) were with ARA and, more recently, this PUFA has continued to be well studied. IsoPs can be considered to be geometric isomers of prostaglandins but where their lateral chains are in the cis configuration (rather than the trans configuration). IsoPs exert their effects through the activation of prostanoid receptors or via kinase receptors (e.g., tyrosine, Rho) [ 111 ]. In the main, IsoPs are formed by free radical reactions on PUFAs esterified to phosphoglycerides rather than using unesterified fatty acids, as for the prostanoids. The oxidation reactions produce a mixture of racemic IsoPs in equal amounts. They are released from phosphoglycerides by phospholipase A2 or platelet-activating factor hydrolase [ 112 ] after which further modifications can take place. IsoPs influence numerous diseases and are important in hypertension and ischemic health. Their metabolism and function have been covered excellently in several recent reviews to which the reader is referred [ 22 , 111 , 112 , 113 , 114 ] for further details.

6.1. Role of Eicosanoids in Inflammation and Immunity The various eicosanoids (prostaglandins, thromboxanes, prostacyclin, leukotrienes, HETEs, EETs, lipoxins) have a variety of biological actions [ 19 ]. Nevertheless, their role in inflammation and immunity is the best researched [ 119 , 120 , 121 , 122 , 123 , 124 , 125 ] compared to other health aspects. As mentioned before, there are two families of relevant PUFAs: the n-3 and n-6 groups. As a generalisation, the n-6 PUFAs give rise to inflammatory mediators while the n-3 form neutral or anti-inflammatory compounds [ 6 , 124 , 125 , 126 , 127 ]. As discussed later (see Section 7 and Section 8 ), this has led to interest in the ratio of dietary n-3/n-6 PUFAs with regard to consumer health [ 6 ]. As a result of infection, the inherent immunity response activates PLA2, COX-2, and 5-LOX so that there is an increased capacity for eicosanoid production [ 19 ]. In many cases, however, the increased production of a particular prostanoid has multiple and, to an extent, conflicting consequences. For example, Calder [ 19 ] illustrated these events for PGE2. The latter has pro-inflammatory actions in inducing pain and fever and enhancing vascular permeability to allow white blood cells to infiltrate the site of infection. Nevertheless, PGE2 is also able to reduce the production of pro-inflammatory cytokines as well as 5-LOX (which will inhibit the formation of pro-inflammatory leukotrienes) and induce 15-LOX to increase the synthesis of pro-resolution lipoxins [ 19 , 126 ]. PGE2, therefore, has both pro- and anti-inflammatory actions and is a regulator of inflammation with activity in its initiation, propagation and resolution phases [ 19 ]. Moreover, the actions of PGE2 towards different cells (dendritic cells, T lymphocytes, B lymphocytes, T-helper 1 cells, etc.) point to the complexity of responses towards individual eicosanoids. PGE2 is usually the most abundant PG in tissues [ 127 ]. It is released into the intracellular space and rapidly converted to 15-keto PGE2 which is essentially inactive. Non-steroidal anti-inflammatory drugs, including aspirin, cause acute reductions in PGE2 through the inhibition of COX activity [ 128 ]. PGE2-EP receptor-mediated effects have notable actions on acute and chronic inflammation as well as for some autoimmune diseases [ 127 ]. The mechanisms involved in PGE2-induced several acute inflammatory conditions have been reported e.g., [ 129 , 130 ]. Downstream events often involve IL-6 production and cellular Ca++ changes. PGE2-EP2/EP4 signalling is involved in T cell differentiation [ 131 ] with effects on TLR signalling via the cAMP-protein kinase A pathway [ 132 ]. PGE2 will also affect Th17 cell expansion [ 133 ] and induce IL-23 production from dendritic cells [ 127 ] and amplify this expression through canonical and non-canonical NFκB pathways [ 134 ]. Some autoimmune diseases (contact hypersensitivity, autoimmune encephalomyelitis, atopic dermatitis, and multiple sclerosis) also appear to be strongly influenced by PGE2-EP2/EP4 signalling [ 127 ]. PGD2 is an allergic and inflammatory mediator when released from mast cells [ 120 , 122 ]. It is important for the induction of asthma and can have pro-inflammatory actions on the skin to cause erythema, oedema, and leukocyte infiltration [ 135 ]. PGD2 also has multiple pro-inflammatory effects in several animal models [ 19 ]. PGD2 is one of the eicosanoids (along with PGA1, PGA2, PGD1 and PGJ2) which activates PPAR [ 136 ] and this activation is not found with many prostaglandins. Using a competition binding assay, the eicosanoids 8( S )-hydroxyeicosatetraenoic acid and 15-deoxy-PGJ2 (15d-PGJ2) bound to PPAR alpha and PPAR gamma, respectively, at physiological concentrations [ 137 ]. Further work on 15d-PGJ2 as a natural PPARgamma agonist has revealed that it can play many roles in anti-tumour and anti-inflammatory processes [ 138 ]. Simultaneously, PPARs can influence NFkappaB and JAK-STAT which regulate the expression of downstream genes and mediate inflammatory diseases such as arthritis [ 139 , 140 ] as well as cancer [ 141 ]. Moreover, the connection between PPARgamma with 15dPGJ2 activation and different cancer lines has been confirmed [ 142 ]. A similar connection has been reported for reductions in inflammatory parameters [ 143 ]. LOX products are also important in inflammation [ 19 , 120 , 122 , 144 ]. LTB4 is a potent chemoattractant and acts to recruit leukocytes to inflammatory sites. It can also act to increase the adhesion of these cells to the endothelium. Other pro-inflammatory effects include increasing the production of reactive oxygen species and pro-inflammatory cytokines [ 19 ]. The cysteinyl-LTs (LTC4, LTD4, LTE4) are produced by a number of cells (e.g., macrophages) following stimuli such as allergens. This, in turn, can have a role in asthma as well as other pro-inflammatory responses [ 119 , 145 , 146 ]. While 5-LOX produces various leukotrienes, other LO

While there are numerous clinical studies emphasising the role of dietary PUFAs in health and diseases, there are an increasing number of reports about the specific involvement in n-3 PUFA-derived SPMs. Thus, dietary supplementation with n-3 PUFA or marine oils should increase SPM synthesis to correlate with better phagocyte function [ 11 ]. In contrast, SPM biosynthesis is reduced in several diseases such as multiple sclerosis, osteoarthritis, and tuberculous meningitis [ 8 ]. In addition, it has been noted that SPM biosynthesis in COVID-19 patients is lowered (as revealed in bronchoalveolar lavages, plasma, and serum samples) [ 22 ]. The proposed routes for the formation of the SPMs (protectins, resolvins, maresins) and their peptide-conjugates have been covered well in recent reviews [ 11 , 22 ] (see Section 3 ). More details about the molecular details of their enzymatic biosynthesis will be found in [ 21 , 89 , 95 , 188 ]. In an experimental system for bacterial-triggered skin inflammation, SPMs are produced locally at the site of infection. These include several resolvins as well as lipoxin LXB4 which (when applied at physiological concentrations) reduced PMN numbers [ 189 ]. In the same model system, oral treatment with a cannabinoid anabasum increased resolvins, reduced pro-inflammatory leukotriene LTB4, and promoted resolution [ 190 ]. Moreover, in LPS-challenge experiments, the role of SPMs in inflammation resolution was confirmed [ 11 ]. Macrophages, which are involved in inflammatory reactions, can be divided into M1 and M2 types. M1-macrophages respond to bacterial infections by producing pro-inflammatory mediators such as LTB4 and PGE2. In contrast, M2-macrophages are proposed to produce SPMs. Moreover, a dual inhibitor of PGE2-synthase 1 and 5-LOX selectively reduced LT and PG production in M1-macrophages while increasing SPM (resolvins, maresins) production in M2-macrophages [ 191 ]. 6.3.1. Production of Resolvins The potential functions of the various types of SPMs have been described well in recent reviews [ 11 , 19 , 22 ]. The E-series resolvins are formed from EPA and RvE1 protects tissues from leukocyte-mediated injury as well as excessive pro-inflammatory responses [ 12 ]. This resolvin regulates vascular inflammation by protecting against stenosis including modifying oxidised LDL uptake [ 11 ]. RvE1 reduces neuroinflammation in an Alzheimer’s disease model and, in addition, shows benefits against cancerous cells [ 192 ]. Other E-series resolvins exhibit similar benefits against harmful inflammatory situations. RvE2 prevents PMN infiltration in peritonitis, RvE3 acts on airway allergic inflammation, and RvE4 has several actions in physiological hypoxia [ 11 ]. Resolvin E1 reduces injury-related vascular neointima formation by inhibiting inflammatory responses [ 193 ] and early treatment with this resolvin facilitates myocardial recovery from ischaemia in mice [ 194 ]. In fact, some studies have shown that RvE1 and other SPMs are reduced in patients with atherosclerosis [ 195 ]. In addition, RvE1 can regulate the functions of macrophages by stimulating efferocytosis in murine models [ 196 ]. Other animal models have been used to show that RvE1 can decrease atherogenesis [ 197 ] and reduce the size of atherosclerotic lesions [ 198 ]. The potential beneficial roles of resolvins and other SPMs in cardiovascular disease have been reviewed [ 8 , 107 ]. The D-series resolvins are produced from DHA either from LOX action (RvDs) or by COX-2 in the presence of aspirin (AT-RvDs/17R-RvDs) [ 11 , 18 ]. Both RvD1 and 17R-RvD1 are reported to regulate phagocytosis at very low concentrations. Thus, they may have benefits against inflammation in Parkinson’s disease, bacterial-induced lung inflammation, and tumour growth [ 11 , 199 ]. These resolvins may also benefit ischemia following tissue injury such as in ischemia/reperfusion-induced kidney injury [ 11 ] as well as regulating leukocyte functions [ 200 ]. RvD2 also suppresses tumour growth and may be a potent regulator of bacterial sepsis. It can reduce tissue necrosis after burn wounds and stimulate repair while enhancing muscle regeneration [ 11 ]. RvD3 and 17R-RvD3 are reported to show anti-cancer activity and regulate leukocyte actions, while RvD4 has been suggested to have potential for controlling thrombo-inflammatory disease. RvD5 may help control bacterial infections [ 11 ] and, interestingly, shows gender dimorphism for pain regulation [ 201 ]. While the FPR2/ALX receptor is involved with lipoxin signalling, AT-RvD1 is also an antagonist [ 107 ]. However, RvD1 and RvD2 additionally signal via DRV1/GPR32 [ 107 ]. As expected, GPR32 is expressed in vascular endothelial [ 202 ] and smooth muscle cells [ 203 ]. The beneficial effects of RvD1 on endothelial cell integrity and barrier function are blocked by antibodies against DRV1/GPR32 or FPR2/ALX, suggesting that RvD1 acts through both [ 202 ]. For RvD2, the receptor is termed DRV2/GPR18 [ 204

Apart from EPA and DHA, n-3 docosapentaenoic acid (n-3DPA) can also be converted into SPMs. These include resolvins, protectins and maresins as well as 13-series resolvins [ 11 , 22 ]. A number of studies suggest a role for DPA in inflammation [ 231 ]. The effects are due to the conversion of DPA into bioactive mediators by similar pathways to those for EPA and DHA. Diurnal regulation of RvDn-3DPA, as well as its addition to the blood of healthy volunteers, support its protective role in cardiovascular disease [ 22 ]. In addition, the benefits of RvD5n-3DPA were noted in arthritic inflammation. The same mediator can help during experimental sepsis [ 22 ]. 13-series resolvins, which are also formed from n-3 DPA, ensure neutrophil movement to infection sites, increase the uptake and killing of bacteria by phagocytes, and promote the uptake of apoptotic cells by macrophages. Moreover, the production of such mediators is upregulated by statins, which have been suggested to increase their efficacy in reducing inflammation [ 232 , 233 ]. In this section, I have concentrated on the actions of lipid mediators in inflammation per se . However, as noted before, many important diseases have chronic inflammation as a causative or pre-disposing factor. Thus, the classic eicosanoids and newly discovered SPMs, as well as elovanoids and isoprostanes, can be important in these diseases. For more details on the actions of lipid mediators in the health or diseases of the cardiovascular system, renal or gastrointestinal function, reproduction, arthritis, or cancer, the reader is referred to recent publications [ 11 , 12 , 18 , 19 , 22 , 38 , 148 ]. The role of SPMs in glial cell function, and hence in nervous system development, homeostasis, and response to injury, have been reviewed recently [ 234 ]. Receptors for all the main SPMs (including LXA4) have been noted with differences in their effects on astrocytes and neurons. As in other tissues, the binding of LXA4 to the ALX/FPR2 receptor inhibits the expression of pro-inflammatory cytokines via an NFκB-dependent mechanism [ 235 ]. SPMs strongly reduce microglial activation and their production of pro-inflammatory mediators, and increase their ability to phagocytose and remove aggregated proteins such as amyloidβ [ 236 , 237 ]. These actions have relevance to several models of neurodegenerative conditions such as Alzheimer’s disease, multiple sclerosis, and Parkinson’s disease [ 234 ]. In addition, Ponce et al. [ 238 ] reviewed the role of SPMs in reducing neuroinflammation associated with neurodegenerative disorders. Alzheimer’s and Parkinson’s diseases were particularly emphasised, along with the links between pro-inflammatory compounds and signalling pathways such as NFκB. Notably, the beneficial effects of NPD1 on beta-amyloid precursor protein metabolism were noted [ 239 ]. Such observations suggest that SPMs could be useful as add-on or standalone agents for the treatment of patients with neurodegenerative diseases [ 238 ]. Despite an impressive number of publications from different laboratories documenting the presence and biological activities of various SPMs, some key aspects have been questioned recently [ 81 ]. These authors challenge the analytical methodology when applied to leukocytes and other cells. They suggest that concentrations of SPMs in vivo are too low to have the biological effects claimed. Moreover, SPMs in tissues are too low and too unstable to be assayed with any confidence. Furthermore, SPMs did not increase with supplements of their omega-3 precursors and the evidence for their appearance during the resolution phase of inflammation was inadequate. In addition, the identity of G-protein-coupled receptors needs to be confirmed by independent means. What is not in dispute seems to be the detailed structural analysis and synthesis work to identify the various SPMs that can be formed from EPA and DHA. Moreover, the increasing body of evidence showing that SPMs have beneficial effects when applied at sufficient concentrations in various disease states is important and suggests significant future clinical applications. A meeting (since the publication of [ 81 ]) in Sweden, where the different parties were present, failed to resolve the points at issue and future clarification is needed. In the meantime, it seems prudent for journals (and their referees) to be fully cognizant of the questions raised about studies of SPMs and the difficulty of robust analysis of these unstable molecules in biological samples.

Oil crops are the main sources of PUFAs in food and feed. Currently, the main crops are oil palm, soybean, oilseed rape and sunflower, representing about 35%, 26%, 15% and 9%, respectively, of the total consumption [ 271 ]. In short, these four crops currently provide 86% of the world’s vegetable oils. Details of the production, composition, processing, and utilisation of major oil crops are given in The Lipid Handbook [ 272 ]. The dominant fatty acids of commodity vegetable oils are palmitic, oleic, and linoleic acids. While α-linolenic acid (ALA) is present in most oils, it is usually only in small amounts. Soybean (7%), oilseed rape (6%), and flax (47%) are crops with significant amounts of ALA. However, there is so much linoleic acid in soybean oil that the ratio of n-6/n-3 PUFA is about 8. Only In oilseed rape (Canola varieties) is there a ‘healthy ratio’ (about 2.3) for human consumption [ 240 , 271 ]. As noted before, because the conversion of EFAs into VLCPUFAs (which provide almost all lipid mediators) uses enzymes which work with both n-6 and n-3 PUFAs ( Figure 4 ), their dietary ratio is very important and has led to increased attention being paid to adequate dietary n-3 PUFAs. Dietary effects and recommendations are discussed in [ 240 ] while the food applications of lipids are reviewed in [ 273 ]. 8.1. Increasing Oil Crop Yields There are two main aspects to consider when discussing dietary sources of PUFAs. First, there is a continuous need for more oil year after year [ 272 ], the vast bulk of which comes from land-based crops. Second, dietary considerations dictate that the overall composition of oils and their contribution to lipid uses is important [ 271 ]. For increasing oil yields in crops, several approaches have been used [ 271 ]. These include traditional breeding and agricultural practices. The problems with climate change have increased the importance of both these aspects. For seed oils, quantitative trait loci (QTL) analysis has been used to try and identify useful genes. Moreover, a combination of transcriptome and metabolome analysis can identify regulators that can influence seed oil content [ 274 ]. Metabolic control analysis has been utilised to study individual enzymes and quantify their regulatory influence in triacylglycerol formation [ 275 , 276 ]. Such a method allows precise characterisation of how manipulating a particular enzyme can change lipid accumulation. An example was where DGAT1 was up-regulated in B. napus and increased the oil content of seeds significantly in both greenhouse [ 277 ] and field conditions [ 278 ]. Following this example, DGAT is often used in combination with other genes to increase oil accumulation. The synergistic effect of DGAT together with a transcription factor, Wrinkled 1 (WRI1), was shown to channel carbon from carbohydrate metabolism into lipid biosynthesis [ 279 ]. In fact, this combination of useful genes to increase oil accumulation has now been taken further with the ‘Push/Pull/Package/Protect’ combination [ 280 ]. In this, WRI1 is used to push carbon into lipid synthesis, DGAT pulls the carbon into triacylglycerol formation, oleosin packages oil into droplets, and lipase action is reduced to prevent catabolism. Another example of combining the expression of several genes to augment oil accumulation is given in [ 281 ]. The usefulness of flux control analysis in defining the amount of regulation caused by individual enzymes in a pathway has been utilised further. Examination of lysophosphatidate acyltransferase showed that, despite its low intrinsic flux control coefficient, it could contribute significantly to oil accumulation [ 282 ]. This agreed with the data of Liu et al. [ 281 ], also with B. napus . Instead of DGAT, triacylglycerol can also be formed by phospholipid: diacylglycerol acyltransferase (PDAT) in a non-acyl-CoA reaction. When the activity of this enzyme was increased it had a negative effect on oil accumulation and altered lipid metabolism in B. napus [ 283 ]. This contrasts with the beneficial effect of PDAT in crops producing some unusual oils [ 283 ]. Thus, flux control analysis can provide very useful information about triacylglycerol formation in oil crops. Nevertheless, it neglects the integration of lipid metabolism into a genome-scale metabolic network, which has been reviewed recently [ 284 ].

Since their discovery as essential fatty acids, linoleic and α-linolenic acids have been recognised for a number of important biochemical and physiological functions. Their actions are almost always due to their metabolic conversion to lipid mediators. Apart from the classic eicosanoids, recent research has identified many other mediators, such as the SPMs and isoprostanes. Together, the lipid mediators have very important health functions and, moreover, they impact many of the most important human disease states. With this recognition, it has become clear that dietary supplies of the necessary PUFAs, especially VLCPUFAs, are critical. Thus, efforts are being made to replace the current main source, fish oils, with photosynthetic products, especially in land crops. Success in these ventures promises to ensure these vital constituents for future diets.