Dietary supplements: PUFAs
Polyunsaturated fatty acids (PUFAs) play fundamental roles in many cellular and multicellular processes, including inflammation, immunity, and neurotransmission. They must be obtained through diet, and a proper balance between omega-6 (ω-6) PUFAs and ω-3 PUFAs is essential. The typical Western diet contains a surfeit of ω-6s and a deficiency of ω-3s [130].
Arachidonic acid (AA) is the archetypical ω-6, with 20 carbons and four double bonds (20:4ω-6). Some of its metabolites cause chronic diseases seen in Western populations: prostaglandins cause pain and swelling, and leukotrienes cause bronchoconstriction and asthma. The inflammatory metabolites of AA are countered by dietary ω-3s. The two best-known ω-3s are eicosapentaenoic acid (EPA, 20:5ω-3) and docosahexaenoic acid (DHA, 22:6ω-3).
eCBs are derived from AA (see Figure 2). Several preclinical studies showed that dietary supplementation with AA increased serum levels of AEA and 2-AG, summarized in Table 1. Although we clearly need AA to biosynthesize eCBs, excessive levels of AA, administered chronically, may lead to excessive levels of eCBs. This in turn may lead to desensitized and downregulated CB1 and CB2 receptors. Linoleic acid, an 18:2ω-6 PUFA, is converted into AA, and it elevated 2-AG and AEA levels and induces obesity in mice [131].
10.1371/journal.pone.0089566.t001 Table 1  Effects of PUFA supplementation upon eCB levels.     1  ↑, increase; ↓, decrease; ≈, no change; Dietary supplementation with ω-3s predictably increased the concentration of EPA and/or DHA in tissues, cells, and plasma, and decreased the relative concentration of AA in tissues, cells, and plasma [132], [133]. ω-3 supplementation also decreased AEA and 2-AG in tissues, cells, and plasma (Table 1). However, the effects of ω-3 supplementation are nuanced and complex:
Piscitelli et al. [134] fed mice a high-fat diet (cholesterol and saturated fatty acids) with little AA. This diet caused a decrease in AEA and 2-AG in the liver. Supplementing that diet with DHA and EPA increased AEA and 2-AG in the liver. In contrast, the high-fat diet increased AEA and 2-AG in muscle tissue, and supplementation with krill oil decreased AEA and 2-AG. Similar trends were seen in heart, kidneys and white adipose tissue.
Adequate levels of dietary ω-3s are required for proper eCB signaling. Mice supplemented with ω-3s, compared to mice on a control diet, expressed greater levels of CB1 and CB2 mRNA. Mice supplemented with ω-3s also expressed greater levels of eCB synthetic enzymes—NAPE-PLD, DAGLα, and DAGβ [132]. Supplementation with ω-3s also modulated the concentrations of “entourage compounds” such as PEA and OEA [133], [134].
In apparent contrast with the above findings, Lafourcade et al. [135] showed that ω-3 deficiency abolished eCB-mediated neuronal functions. They reasoned that lifelong ω-3 deficiency causes chronically elevated eCB levels within brain synapses, which leads to CB1 desensitization. They tested a rodent model of depression-like behavior (the forced-swim test), and ω-3-deficient mice performed like CB1 −/− knockout mice. The administration of WIN55212-2 did not change their behavior, whereas in ω-3-rich mice, WIN55212-2 imparted typical cannabimimetic effects. Larrieu et al. [136] demonstrated depressive-like symptoms in ω-3-deficient mice compared to mice fed an ω-3 enriched diet. They used the forced-swim test as well as the more valid open-field and social-investigation tests. Mice deficient in ω-3 showed impairment in the CB1 signaling pathway—ERK1/2 phosphorylation in the hippocampus was reduced after treatment with WIN55212-2, and the antianxiety effects of WIN55212-2 were absent in ω-3-deficient mice.
ω-3 PUFAs may impact the eCB system via a second mechanism: eCB biosynthetic enzymes readily accept ω-3s as substrates. An ω-3-rich diet markedly elevated the N-acyl-ethanolamide metabolite of DHA, called DHEA, the N-acyl-ethanolamide metabolite of EPA, called EPEA, and the sn-2-glycerol-ester metabolite of EPA, called 2-EPG [133], [137]. FAAH catabolized DHEA [138], [139]. DHEA and EPEA act as eCBs: DHEA and EPEA showed high binding affinity for CB1 (K i = 124 and 55 nM respectively) and acted as partial agonists [139]. Their affinity nearly equals that of AEA—a meta-analysis of affinity studies using the same binding assay (mouse brain, [3H]CP55940 displacement, presence of PMSF) produced a modal K i value of 61 nM for AEA [22]. DHEA, aka synaptamide, stimulates neurite growth and synaptogenesis in developing hippocampal neurons [140].
In natural fish oil, DHA and EPA are esterified in triacylglycerides (TAG), whereas in many fish oil capsules, DHA and EPA are esterified in EE (ethyl-ester) or TAG (rTAG). Krill oil contains DHA and EPA esterified in phospholipids, primarily phosphatidylcholines, which may improve their bioavailability; furthermore krill oil contains less AA than fish oil [141]. Batetta et al. [142] supplemented the diet of obese Zucker rats with fish oil or krill oil, which contained nearly identical amounts of EPA and DHA. The visceral adipose tissue of krill oil-supplemented rats contained less AEA and 2-AG than fish oil-supplemented rats. In the liver only AEA levels were significantly less. The effects of these dietary sources of DHA and EPA on brain eCB levels were much less pronounced, with krill oil producing only a small decrease of 2-AG levels [143]. The same research group reported similar results in an obese cohort mostly composed by women: krill oil but not fish oil significantly decreased serum 2-AG levels; no significant changes were seen in normo-weight subjects [144]. In a yet unpublished study, one of us observed that in obese men, dietary krill oil reduced plasma AEA levels and concomitantly counteracted hypertriglyceridemia (Di Marzo, unpublished data).
In summary, dietary ω-3s seem to act as homeostatic regulators of the eCB system. In obese rodents fed a high-AA diet, ω-3s significantly decrease eCBs, especially 2-AG, particularly in tissues that become dysregulated, such as adipose and liver tissues. Plasma eCB levels are reduced by krill oil also in obese humans. Little change in eCB levels are seen in normo-weight individuals not fed a high ω-6 diet, and dietary ω-3s are required for proper eCB signaling.