Dietary linoleic acid elevates endogenous 2-AG and anandamide and induces obesity

Abstract

Suppressing hyperactive endocannabinoid tone is a critical target for reducing obesity. The backbone of both endocannabinoids 2-arachidonoylglycerol (2-AG) and anandamide (AEA) is the ω-6 fatty acid arachidonic acid (AA). Here we posited that excessive dietary intake of linoleic acid (LA), the precursor of AA, would induce endocannabinoid hyperactivity and promote obesity. LA was isolated as an independent variable to reflect the dietary increase in LA from 1 percent of energy (en%) to 8 en% occurring in the United States during the 20th century. Mice were fed diets containing 1 en% LA, 8 en% LA, and 8 en% LA + 1 en% eicosapentaenoic acid (EPA) + docosahexaenoic acid (DHA) in medium-fat diets (35 en% fat) and high-fat diets (60 en%) for 14 weeks from weaning. Increasing LA from 1 en% to 8 en% elevated AA-phospholipids (PL) in liver and erythrocytes, tripled 2-AG + 1-AG and AEA associated with increased food intake, feed efficiency, and adiposity in mice. Reducing AA-PL by adding 1 en% long-chain ω-3 fats to 8 en% LA diets resulted in metabolic patterns resembling 1 en% LA diets. Selectively reducing LA to 1 en% reversed the obesogenic properties of a 60 en% fat diet. These animal diets modeled 20th century increases of human LA consumption, changes that closely correlate with increasing prevalence rates of obesity. In summary, dietary LA increased tissue AA, and subsequently elevated 2-AG + 1-AG and AEA resulting in the development of diet-induced obesity. The adipogenic effect of LA can be prevented by consuming sufficient EPA and DHA to reduce the AA-PL pool and normalize endocannabinoid tone.

Figure 1

Selective elevation of dietary LA elevates AA precursors and endocannabinoids in mice fed medium-fat diets (open bars) and high-fat diets (coarse bars). Dietary LA at 1 en% is indicated by white bars, 8 en% LA by dark gray bars, and 8 en% LA + 1 en% EPA/DHA by light gray bars. Dietary LA (8 en%) elevates (a) liver LA in phospholipids (PL) (µg/mg) and (b) liver PL-AA (µg/mg). Compared to 8 en% LA, diets of 1 en% dietary LA allows endogenous accretion of (c) liver PL–EPA (µg/mg), levels equal to consuming 1 en% EPA/DHA directly. Increasing dietary LA from 1 en% to 8 en% elevates (d) liver 1-AG + 2-AG (ng/mg) (by over fourfold), (e) brain 1-AG + 2-AG (ng/mg), and (f) liver AEA (ng/ml). The addition of 1 en% EPA/DHA reduces the endocannabinoid precursor pool and tissue concentrations of 2-AG and AEA. ^a,b,c,dDiffering letters indicate P < 0.01 and lines by specified values by Mann–Whitney testing. Lines indicate differences at P values shown. n = 9–10. 2-AG, 2-arachidonoylglycerol; AA,arachidonic acid; AEA, anandamide; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LA,linoleic acid.

Figure 2

Dietary LA induces adiposity in mice fed medium-fat diets (open bars) and high-fat diets (coarse bars). Dietary LA of 8 en% increases (a) food intake in high-fat diets, (b) feed efficiency in medium-fat diets, (c) reduces plasma adiponectin (µg/ml/g fat tissue), and increases (d) plasma leptin (ng/ml). Compared to a diet of 1 en% LA, 8 en% LA also increases (e) body weight and (f) adiposity index. Adding 1 en% EPA/DHA reverses the effects of 8 en% diets. Feed efficiency; (body weight gain/Mcal intake), adiposity index ((subcutaneous + retroperitoneal + inguinal fat pads)/eviscerated body weight × 100). Adiponectin levels were adjusted for gram fat tissue dissected out (subcutaneous + inguinal + retroperitoneal fat pads). ^a,b,c,dDiffering letters indicate P < 0.05 by ANOVA, leptin, and adiponectin; n = 6, other parameters; n = 9–10. DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LA,linoleic acid.

Figure 3

Reducing dietary LA to 1 en% prevents adipose tissue accumulation and reverses the obesogenic effects of a high-fat (60 en%) diet. Animals fed 8 en% LA (b,e) accumulated more fat than animals fed 1 en% LA (a,d). The addition of 1 en% n-3 EPA/DHA to 8 en% LA diets (c,f) prevented the increase in adipose tissue seen in animals fed 8% energy LA in b,e. The obesogenic properties of a high-fat diet (60 en% fat) in e were reversed by selective reduction of LA from 8 en% to 1 en% and replacement by greater saturated fat in d. The animals shown are representative for the animals in each dietary treatment. Upper row; isocaloric medium-fat diets of 35 en% fat, lower row; isocaloric high-fat diets of 60 en% fat. The fatty acid composition, not total fat calories, determined the obesogenic properties of the diets. DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LA,linoleic acid.

Figure 4

Dietary essential fats and the metabolism of endocannabinoids. The endocannabinoids 2-AG and anandamide (AEA) are synthesized on demand from the essential fatty acid arachidonic acid (AA) (20:4n-6) in membrane phospholipids (AA-PL). AA in the phospholipid precursor pool can be elevated by dietary linoleic acid (LA) (18:2n-6) or diminished consumption of omega-3 fatty acids notably eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Activation of the cannabinoid receptors (CB₁ and CB₂) by the AA-PL derived endocannabinoids, AEA and 2-AG, both centrally and peripherally, favors metabolic processes that stimulate appetite, increase food intake, activates fat storage pathways, promotes adipocyte inflammation, and downregulates catabolism resulting in adipose accretion. Figure indicates the shift in available seed oil and changes in estimated tissue compositions of n-6 highly unsaturated fatty acids described in Blasbalg et al. 2010 (10). Horizontal arrows indicate the hypothesis that these shift in diet and resulting endocannabinoid levels may contribute to the increasing prevalence of obesity in the United States during the 20th century. 2-AG, 2-arachidonoylglycerol; HUFA, highly unsaturated fatty acids; NAT, N-acetyltransferase; PLA2, phospholipase A2; PLC, phospholipase C.

Figure 5

Dietary sources of linoleic acid (LA) and increasing prevalence rates of male obesity in the United States during the 20th century. Prevalence of male obesity (40–59 years, BMI >30) in each year is indicated by source: UA-Union Army Veterans, F-Framingham cohorts, N-NHANES cohorts. Scattergrams and univariate linear regression lines are indicated in each panel. Increasing prevalence rates of male obesity are positively correlated with apparent consumption of dietary sources of LA indicated in (a) poultry a percent of energy (en%), (r² = 0.94, P < 0.000000), (b) soybean oil (en%) (r² = 0.82, P < 0.00005), (c) shortening (en%) (r² = 0.86, P < 0.00002), and (d) sugars (en%) (r² = 0.37, P < 0.04) but not to (e) total calories/p/d (r² = 0.27, P < 0.08). Prevalence of obesity is negatively correlated with (f) declines in physical labor occupations, farmers, farm laborers, and laborers (r² = 0.49, P < 0.01). Animal diets of 1 en% and 8 en% LA were selected to model these changes.