Endocannabinoids and the Control Of Behavior and Cardiovascular Function

我们之前已经证明，内源性大麻素通过CB1受体以年龄依赖的方式促进自愿饮酒，使用的是两瓶/自由选择范式的小鼠模型（PNAS 100:1393, 2003）。本研究中使用的脑渗透CB1逆激动剂利莫那班随后被引入治疗肥胖，但由于神经精神方面的副作用，包括焦虑、抑郁和自杀意念，不得不在2008年退出制药市场。我的实验室支持的另一种方法是引入外周限制性CB1逆激动剂，它保留了利莫那班的代谢功效，但在代谢综合征的啮齿动物模型中缺乏其神经行为作用。矛盾的是，这些化合物还被发现可以减少食物摄入，假设这是一种中枢调节效应。这一悖论得到了解决，研究表明，通过抑制脂肪组织中瘦素的产生和增加肾脏中瘦素的清除率，外周CB1阻断在饮食诱导的肥胖小鼠中通过逆转其高瘦素血症，迅速逆转其瘦素抵抗。这让我们想知道C57Bl6小鼠的高酒精偏好，CB1激活促进的另一中枢功能，是否也可能通过外周CB1的阻断间接影响。一种可能的机制与胃饥饿素有关，胃饥饿素是一种胃肽，通过大脑中的胃饥饿素受体促进食欲。正如Leggio博士和他的团队最近在我们研究所所证明的那样，胃饥饿素也会促进对酒精的渴望和自我管理（Biol Psychiatry, 2014; Mol Psychiatry, 2017）。

We have earlier demonstrated that endocannabinoids acting via CB1 receptors promote voluntary alcohol drinking in an age-dependent manner, using a mouse model of two bottle/free choice paradigm (PNAS 100:1393, 2003). The brain-penetrant CB1 inverse agonist rimonabant used in this study was subsequently introduced as a treatment of obesity, but had to be wihdrawn from the pharmaceutical market in 2008, due to neuropsychiatric side effects, including anxiety, depression and suicidal ideation. An alternative approach, championed by my laboratory, was the introduction of peripherally restricted CB1 inverse agonists that retained the metabolic efficacy of rimonabant but were devoid of its neurobehavioral effects in rodent models of the metabolic syndrome. Paradoxically, such compounds were also found to reduce food intake, hypothesized to be a centrally mediated effect. This paradox was resolved by the demonstration that peripheral CB1 blockade in diet-induced obese mice rapidly reversed their leptin resistance by reversing their hyperleptinemia, via inhibiting leptin production in adipose tissue and increasing leptin clearance in the kidney. This made us wonder whether the high alcohol preference of C57Bl6 mice, another central function promoted by CB1 activation, may also be affected indirectly through blockade of CB1 in the periphery. A possible mechanism involves ghrelin, a gastric peptide that promotes appetite via ghrelin receptors in the brain. Ghrelin also promotes alcohol craving and self-administration, as demonstrated recently in our Institute by Dr. Leggio and his group (Biol Psychiatry, 2014; Mol Psychiatry, 2017). The preliminary findings reported last year have been extended and the findings clearly indicate that rimonabant and the non brain-penetrant CB1 inverse agonists, JD5037, were equi-effective in markedly reducing total alcohol intake as well as ethanol preference in wild-type but not in CB1R-/- C57BL6 mice, using a 'two bottle, free choice' paradigm as well as a 'drinking in the dark' paradigm. Furthermore, peripheral CB1 blockade significantly reduced plasma levels of the biologically active acetylated ghrelin, with no change in the level of its precursor, desacyl ghrelin, suggesting that the ghrelin acylation process, involving the enzyme GOAT and its substrates ghrelin and octanoate, maybe CB1R target(s). In additional experiments we found that both alcohol preference and absolute intake are lower in ghrelin knockout and ghrelin receptor1 (GHSR-1)knockout mice, with no additional reduction caused by peripheral CB1 receptor blockade. The effects on plasma ghrelin and the finding in the knockout strains are compatible with acylghrelin involvement in the effects of JD5037 on alcohol drinking. We also found that afferent vagal (sensory) denervation of the stomach as well as subdiaphragmatic surgical denervation of the stomach increased alcohol drinking and drinking preference and abolished the efficacy of peripheral CB1 blockade as well as GHSR-1 blockade to reduce drinking. These findings are compatible with a scheme whereby alcohol drinking increases ghrelin acylation and/or acylghrelin secretion through CB1 receptor activation, and the released acylghrelin signals via ghrelin receptors (GHSR1) on vagal sensory terminals to promote ethanol drinking behavior. We next explored the cellular mechanism involved in the selective decrease in acyl-ghrelin production by peripheral CB1 blockade. Ghrelin-producing MGN3-1 stomach cells generate octanoic acid exclusively from long-chain fatty acid precursors, and also generate endocannabinoids and express CB1 receptors. JD5037 suppressed octanoyl-ghrelin but not desacyl-ghrelin production in MGN3-1 cells by reducing the levels of octanoyl-carnitine generated from palmityl-carnitine due to an increase in fatty acid oxidation (FAO), which was further confirmed by measuring FAO directly using a fluorescent FAO assay kit. Furthermore, treating MGN3-1 cells with selective inhibitors of 3 different enzymes in the FAO pathway selectively reduced acyl-ghrelin production and occluded the similar effect of JD5037. These findings have now been published in Cell Metabolism.