CNS regulation of plasma cholesterol

Abstract

The incidence of disorders related to the control of energy homeostasis, such as hypertension, diabetes, obesity, and dyslipidemia, has dramatically increased worldwide in the last decades. The central nervous system (CNS) plays a critical role regulating the energy balance, therefore there has been increasing interest in understanding the mechanisms whereby the brain controls peripheral metabolism, in order to develop new potential therapies to treat those disorders. While the involvement of the CNS in development of hypertension, obesity, and diabetes has been thoroughly investigated, less is known about the specific role of the brain in the control of circulating lipids. Here we summarize the evidence linking CNS disorders with dyslipidemia, as well as the central mechanisms that directly influence plasma cholesterol.

Key words::

• Brain
• cholesterol
• metabolic syndrome

Key messages

• Neuronal circuits in the central nervous system control plasma cholesterol.

Introduction

Cardiovascular disease (CVD) remains the major cause of mortality and morbidity in developed societies and will become the leading cause of death in the world by 2020. Atherogenic dyslipidemia, characterized by hypertriglyceridemia, hypercholesterolemia, and low high-density lipoprotein cholesterol (HDL-C) levels, ranks among the most firmly established and best understood risk factors for atherosclerosis. As a consequence, elevated levels of atherogenic lipoproteins (low-density lipoproteins (LDL) and very-low-density lipoproteins (VLDL)) and reduced levels of anti-atherogenic lipoproteins are targets for therapeutic intervention. Because cholesterol homeostasis is of extreme importance at the whole body level, cells control cholesterol concentrations over a very narrow range with various dedicated pathways for cholesterol synthesis, uptake from LDL, and export to high-density lipoprotein (HDL) (1).

In addition to atherogenic dyslipidemia, other risk factors for CVD include metabolic disorders such as obesity and diabetes. Frequently, these disorders occur simultaneously, a situation that has been denominated the ‘metabolic syndrome’. Numerous studies have demonstrated that specific regions of the central nervous system (CNS), such as the hypothalamus and the brain-stem, play a critical role in the regulation of processes altered in the metabolic syndrome (2–4). Furthermore, emerging evidence points to the CNS as a key modulator of lipoprotein metabolism (5–8). In this review we will focus on the assessment of CNS function as a critical determinant of lipoprotein metabolism. We will then discuss current concepts regarding utilization of potential treatments that target the CNS–cholesterol metabolism axis.

Neurogenic disorders and plasma cholesterol levels

The influence of stress in the development of atherogenic dyslipidemia is well known: For example, increased cardiovascular risk factors, including atherogenic dyslipidemia, are associated with post-traumatic stress disorder in war veterans (9). Furthermore, individuals undergoing periods of mental stress, such as accountants during periods of maximum activity (10) or students during exams (11), have increased plasma cholesterol levels in comparison with periods of reduced activity. The causes of dyslipidemia associated to stress are multiple, but the activation of the sympathetic nervous system seems to play a significant role: In fact the increase of catecholamines induced by acute intense stress situations, such as car racing, is sufficient to increase plasma triglycerides (12). Under these conditions sympathetic tone and cortisol levels are also increased, both factors that contribute to the development of dyslipidemia in individuals with sleep deprivation (13). The latter has been linked to the development of hypercholesterolemia in humans, including adolescent (14) and premenopausal women (15).

Psychiatric disorders and cholesterol homeostasis

Several psychiatric disorders have been associated with impaired cholesterol homeostasis in the CNS. The critical role of cholesterol as a structural component in nervous tissue suggests that impaired cholesterol metabolism might contribute to the development of those disorders. For example, homozygous individuals carrying a deleterious allele for the enzyme of the cholesterol biosynthetic pathway 7-dehydrocholesterol reductase develop Smith–Lemli–Opitz syndrome. This is an autosomal recessive syndrome that leads to congenital malformation and mental retardation. Interestingly, individuals heterozygous for this allele are phenotypically normal but have increased suicidal behavior (16). Post-mortem analysis of brains from suicidal individuals have found a decreased amount of gray matter cholesterol content (17). Also, higher incidence of depression in an elderly population (18) and higher incidence of suicidal behavior (19–21), especially in violent attempters (22), has been associated with low plasma cholesterol levels. This may suggest a connection between brain and plasma cholesterol levels, even though it is widely accepted that the brain derives its cholesterol through local synthesis and not by plasma import (23). Furthermore, a mutation in the gene encoding apolipoprotein B that leads to hypocholesterolemia correlates with suicidal behavior in the carrier as well as in the family history (24), supporting a potential role of altered cholesterol transport in the development of psychiatric disorders. The correlation between hypocholesterolemia and psychiatric disorders fueled investigations to understand whether treatment with inhibitors of endogenous cholesterol synthesis to lower circulating LDL cholesterol levels could potentially lead to the development of psychiatric disorders. However, Stewart and colleagues reported that long-term use of pravastatin did not correlate with higher incidence of psychological alterations (25). On the other hand, antipsychotic drugs have been shown to increase glycemia and triglyceridemia, and decrease LDL and HDL cholesterol (26). To date, the mechanistic link between circulating cholesterol levels and psychiatric disorders, and which one precedes the other, is largely unknown.

Neurological diseases and plasma cholesterol levels

Given the critical role of cholesterol in the development and function of the central nervous system, alterations in cholesterol homeostasis underlie the development of several neurological diseases. In some cases, such as cerebrotendinous xanthomatosis or Niemann–Pick, the molecular basis of the disease is known and leads to intracellular accumulation of cholesterol (reviewed in (27)). One interesting characteristic of Niemann–Pick disease is that it leads to the development of neurofibrillary tangles similar to what is observed in Alzheimer's disease (AD) (28,29). This and other data have contributed to suggest a role for cholesterol metabolism in the progression of AD, although the specific mechanisms are not completely understood yet (see reviews (30,31)). Although the data are conflicting, there is interest in finding a possible link between plasma cholesterol levels and AD. On the one hand no correlation has been found between plasma cholesterol levels and risk of incidence of AD (32), but on the other hand moderately increased plasma cholesterol during the midlife has been associated with increased risk of dementia (33). In addition, others have reported that high cholesterol levels in late life are associated with decreased risk for dementia (34). Nevertheless, when changes in cholesterol after midlife are considered, the decrease in plasma cholesterol has been correlated with increased risk of AD (35–37).

A causal contribution of plasma cholesterol to the development or progression of AD is limited by the reduced ability of plasma cholesterol to cross the blood–brain barrier (BBB) (38). But it has been shown that certain metabolites of the cholesterol metabolism such as the oxysterols 24S- and 27-hydroxycholesterol (24S- and 27-OHC) actually are crossing the BBB (40). Furthermore, 24S-OHC is synthetized in the brain and eliminated by the liver, whereas 27-OHC is synthetized in peripheral tissues and passes the BBB. This bidirectional flow of oxysterols through the BBB may play a relevant role in the development of AD. Thus, the changes in expression of genes associated with increased risk of AD observed with high-fat feeding in mice has been associated with increased 27-OHC levels (41). However, the extent of the role of plasma oxysterols in the development of AD is far from being elucidated.

Several studies have associated plasma cholesterol levels with incidence of Parkinson's disease (PD), and the results are again conflicting. Low cholesterol intake has been associated with higher incidence of PD (42), whereas a high cholesterol level has been associated with decreased risk of PD (43). Other studies have reported the opposite, associating high plasma cholesterol levels to higher risk of PD (44), and others have found no significant association (45), leading to the conclusion that the interaction of PD and cholesterol homeostasis is unclear.

Neuroendocrine regulation of plasma cholesterol levels

Early evidence of central control of plasma cholesterol by the central nervous system

The sections above describe largely correlative associations between brain function and plasma lipid metabolism. However, recent studies have begun to show evidence for the direct CNS regulation of peripheral lipid metabolism. The initial indication was obtained in experiments involving lesion or electrical stimulation of specific nuclei in the hypothalamus. Lesions of the ventromedial nucleus (VMN) in the hypothalamus induce hyperphagia and obesity, whereas lesions in the lateral hypothalamus (LH) induce anorexia and body-weight reduction. Conversely, low-frequency electrical stimulation of the VMN reduces, whereas of the LH stimulates, feeding in rats (46), suggesting this intervention as a potential therapy for the treatment of obesity (47). The experiments involving lesions or electrical stimulation of specific nuclei in the hypothalamus have demonstrated that altered function of neural circuitries in the CNS results in the development of atherogenic dyslipidemia and atherosclerotic plaque development (48–51). Electrical stimulation through a 3-month period of the medio-basal hypothalamus (MBH) in rabbits fed with a high-cholesterol content diet leads to a significant increase in plasma lipids, including cholesterol, compared to non-stimulated control animals, despite a similar feeding regimen (48). In this study, the electrical stimulation-induced dyslipidemia did not alter the body-weight, as could be predicted by the experiments with lesions in nuclei of the MBH such as the VMN or the LH, which are associated with changes in body-weight. Such increases in plasma cholesterol correlated with a significant increase in atherosclerotic lesions in the aorta and coronary artery of the stimulated rabbits (48).

When lesions in the VMN are induced in rats, the animals develop hyperinsulinemia, hypertriglyceridemia, and hypercholesterolemia, even in the absence of hyperphagia (51,52). The increase in plasma cholesterol can occur relatively quickly, within the first 3 days after the lesions, without noticeable increase in the de-novo synthesis (53). However, the procedure induces an increase in VLDL secretion which can contribute to increases of both triglycerides and cholesterol levels (54). The increase in plasma cholesterol induced by the lesions persists in hypophysectomized rats (55), which suggests that mechanisms independent of the activation of the hypothalamic–hypophysial axis are involved in the increase in cholesterol. Furthermore, VMN lesions increase plasma cholesterol levels in rats rendered insulin-deficient by alloxan (55), which points to additional mechanisms beyond hyperinsulinemia that may contribute to the increase of plasma cholesterol levels induced by VMN lesions.

Regulation of plasma cholesterol levels by central afferent endocrine signals

Alteration in neural circuits that regulate appetite can lead to hyperphagia and ultimately to obesity. In addition, the CNS control of energy metabolism is not limited to regulation of food intake, and signals originated in specific neuronal circuitries of the CNS directly control glucose and triglyceride metabolism.

The most investigated model of neuroendocrine control of peripheral metabolism involves the role of leptin. Leptin is a hormone secreted by the white adipose tissue and acts on the CNS to regulate feeding and energy metabolism. Animal models lacking leptin signaling are characterized by numerous features of the metabolic syndrome including obesity, hyperglycemia, hyperinsulinemia, hypertriglyceridemia, and hypercholesterolemia. Although leptin receptors are expressed in peripheral tissues including the liver, leptin in the CNS is critical for the control of energy balance, as demonstrated by the rescue of leptin receptor expression in the CNS of db/db mice (56). Hepatic specific loss of leptin receptors leads to an increase in lipid accumulation in the liver but does not alter plasma cholesterol levels and in fact improves glucose tolerance (57). Numerous studies have investigated the mechanisms involved in the development of dyslipidemia due to the lack of leptin signaling. It has been shown that leptin-deficient mice have either higher VLDL production (58) or lack of insulin-induced suppression of VLDL secretion (59). In addition to increased hepatic triglyceride production, leptin receptor-deficient Zucker rats have delayed chylomicron clearance, which can contribute to the hypertriglyceridemia and hypercholesterolemia found in this animal model (60). Conversely, db/db mice, which lack leptin receptors, have lower hepatic levels of a receptor involved in the clearance of chylomicrons, the lipolysis-activated lipoprotein receptor (LSR), and the over-expression of LSR in ob/ob mice leads to a reduction in plasma triglycerides and cholesterol levels (61). Systemic leptin administration in rats reduces plasma VLDL cholesterol concentration and increases the proportion of cholesterol secreted in bile (62). Supporting these data, the administration of leptin increases the cholesterol content in bile of ob/ob mice fed with a lithogenic diet, when compared to saline-treated, pair-fed controls (63). This decrease is accompanied by a decrease in the accumulation of cholesterol in the liver and a decrease in the enzymatic activity of 3-hydroxy-3-methyl-glutaryl CoA reductase (HMGCR), the rate-limiting enzyme in the cholesterol synthesis pathway (63). Interestingly, icv administration of leptin reduces hepatic activity of HMGCR, which correlates with a decrease in plasma cholesterol levels without inducing changes in plasma triglycerides (64). Increased hyperinsulinemia and hepatic insulin resistance might be a major contributor of the increased VLDL production observed in leptin-deficient models. For instance, mice without insulin signaling in the liver due to tissue-specific deletion of the insulin receptor in hepatocytes exhibit dyslipidemia with increased VLDL production (65). In addition to its role in the regulation of VLDL secretion, chylomicron clearance, and hepatic cholesterol synthesis, leptin can control HDL clearance. Ob/ob mice have decreased HDL clearance, and administration of leptin (at doses that do not induce body-weight loss) reduces HDL cholesterol levels in ob/ob mice (66). This reduction in HDL cholesterol levels occurs independently of changes in food intake induced by leptin, since it is not observed in pair-fed control mice, Since hepatocytes from ob/ob mice have impaired HDL uptake, it appears that the main organ responsible for this decrease in HDL clearance is the liver (67). The exact mechanisms whereby leptin facilitates HDL cholesterol uptake are not completely understood. Higher HDL cholesterol levels correlate with lower levels of the HDL receptor SRB1 in ob/ob mice, and treatment with leptin increases the protein and mRNA levels of that receptor (68). In support of the latter, others have shown that db/db mice have decreased SRB1 expression (69). These data suggest that SRB1 can contribute to the leptin-dependent changes in HDL cholesterol levels. However, not all the reports have found significant differences in SRB1 expression in ob/ob mice (66).

The specific areas of the CNS responsible for the regulation of VLDL secretion and HDL uptake by leptin are unknown. The control of peripheral metabolism by leptin is the result of the simultaneous targeting of different neural circuits located in diverse nuclei across the CNS (70), and cell type-specific or nucleus-specific deletion of the leptin receptor produces only an intermediate phenotype (reviewed in (71)) when compared with the metabolic alteration observed in leptin receptor-deficient db/db mice. Among those many different sites, the action of leptin targeting the arcuate nucleus is perhaps the best characterized. Leptin interacts with neurons in the arcuate nucleus that produce the orexigenic peptide NPY. These neurons are a main target for leptin action, which suppresses their activity (72). The same neurons co-express the melanocortin receptor inverse agonist agouti-related protein (AgRP). Furthermore, some of the effects of leptin on body-weight and glucose metabolism are contributed by its interaction with a different set of neurons expressing pro-opiomelanocortin (POMC) (73–75), which produces the melanocortin receptor agonist alpha melanocyte-stimulating hormone (aMSH). The action of neuropeptide Y (NPY) in the CNS might contribute indirectly to the increase in plasma cholesterol levels by increasing the secretion of VLDL from the liver. Icv administration of NPY induces an increase in VLDL secretion during euglycemic-hyperinsulinemic clamp conditions (8). These data indicate that signals initiated by central NPY may contribute to the regulation of hepatic VLDL secretion, interacting with hepatic insulin action. In concordance, acute icv injection of a NPY receptor type 5 (Y5 receptor) agonist increases, whereas an Y1 receptor antagonist decreases, VLDL secretion in rats (7). In contrast to the effects of NPY receptor agonists, the CNS administration of either a melanocortin receptor agonist MT-II or inverse agonist SHU9119 (SHU) exerts only a modest effect on VLDL secretion (7). There are two melanocortin receptors expressed in the CNS, MC3- and 4-R. Both receptors are critical for the control of energy balance, and the lack of either leads to obesity in mice, especially under high-fat feeding (76). Studies performed in rats have shown that blockade of melanocortin receptors with SHU increases plasma cholesterol levels independently of food intake (6,77) or changes in insulin or corticosterone levels (77). Such increases correlate with a decrease in the HDL cholesterol clearance (6). Altogether, these data suggest that increased NPY signaling and decreased melanocortin system activity may contribute to the increased VLDL production and decreased HDL cholesterol uptake, respectively, observed in leptin-deficient mouse models. Interestingly, MC4-R-deficient mice are significantly more susceptible to the development of hypercholesterolemia than MC3-R-deficient mice (6). As in the case of leptin-deficient mice, the hypercholesterolemia observed in MC4-R may result in part from hyperinsulinemia (78) and insulin resistance associated with increased hepatic lipogenesis (78,79) that may increase VLDL production. Interestingly, pre-obese MC4-R-deficient mice have shown altered hepatic lipid metabolism in the absence of insulin resistance associated with hyperinsulinemia (79), suggesting a direct control of hepatic lipid metabolism by the CNS melanocortin system. Whether these same mechanisms contribute to the control of HDL clearance is still unknown. Mutations in the MC4-R are one of the most common causes of monogenetic obesity in humans (80), and recently a single nucleotide polymorphism linked to MC4-R has been associated with significant changes in HDL cholesterol levels (81).

In addition to leptin, other afferent signals might contribute to the control of plasma cholesterol by acting directly in the CNS. Treatment with a non-peptidic GLP-1 receptor agonist significantly reduces HDL cholesterol levels in db/db mice (82). This effect might be partially mediated by the action of GLP-1 in the CNS, since icv infusion of GLP-1 leads to a decrease in plasma cholesterol levels, independently of its food intake-suppressing effect (6). GLP-1 is expressed in multiple neurons along the CNS, including POMC neurons in the arcuate nucleus (83), which contribute to the regulation of hepatic glucose production induced by GLP-1. However, whether the central effect of GLP-1 on plasma cholesterol levels is mediated by those neurons is unknown.

Ghrelin is a hormone produced in the stomach that can increase adiposity and food intake (84) acting directly on neurons in different nuclei of the CNS, including the AGRP/NPY neurons of the arcuate nucleus, which can be stimulated by ghrelin (72). In addition, and similar to the effect of the blockade of the melanocortin receptors, the central administration of ghrelin leads to an increase in HDL cholesterol levels (6).

In addition to neuroendocrine signals, nutrient sensing, specifically glucose, plays a role in the control of lipoprotein metabolism. For example, the increase in lactate induced by the icv administration of glucose suppresses the VLDL production in rats, via a mechanism involving the reduction of hepatic SCD-1 (5).

Role of the autonomous nervous system (ANS) in the control of cholesterol homeostasis

The autonomous nervous system (ANS) plays a role in the control of plasma cholesterol levels by regulating synthesis of cholesterol in the liver and lipoprotein metabolism. Studies in rabbits fed with a high-cholesterol diet have shown that the administration of the sympathetic neurotransmitter norepinephrine increased VLDL and LDL cholesterol, as well as triglyceride levels in all lipoprotein particles including HDL (85). The destruction of the noradrenergic terminals in rabbits treated with 6-hydroxy dopamine leads to a significant decrease in cholesterol concentrations in all the different lipoprotein particles, although the triglyceride levels remained unaffected (86). In contrast, chronic administration of 6-hydroxy dopamine and the beta-adrenergic receptor blocker propanolol increased plasma cholesterol levels in rats with hypothalamic lesions and in controls (87). Indeed, a selective elimination of the sympathetic input to the liver by denervation of the hepatic portal system decreases hepatic triglyceride secretion, without affecting hepatic triglyceride and cholesterol content. However, although the denervation does not affect plasma triglyceride levels, it significantly increases the plasma cholesterol levels (88).

In addition to the role of the sympathetic innervations, the parasympathetic branch of the autonomous nervous system also plays a role in the control of plasma cholesterol levels. The suppression in VLDL secretion induced by the manipulation of glucose metabolism in the hypothalamus is prevented in rats in which the hepatic branch of the vagus nerve had been severed (5). The same procedure also prevents the increase in HDL-C levels induced by the blockade of the central melanocortin receptors (6). Despite these data, the specific mechanisms whereby parasympathetic innervation, and ultimately the CNS, regulates VLDL production and HDL cholesterol uptake remain to be uncovered.

Final remarks

Although the role of the central nervous system in the regulation of numerous physiological processes altered in the metabolic syndrome has been known for a long time, the pleiotropic nature of the factors influencing the plasma lipid transport complicated the investigation how the CNS influences plasma cholesterol metabolism. However, accumulating experimental evidence has unveiled a specific contribution of the central neural circuits in the maintenance of plasma lipid levels. In spite of that, more research is needed to characterize fully which peripheral metabolic pathways contribute to the maintenance of plasma cholesterol, and how specific neuronal circuits can regulate plasma cholesterol levels. Similarly, the potential role of peripheral cholesterol metabolism in the development of neurological diseases requires more attention due to the increasing morbidity of diseases such as Alzheimer's disease. A better understanding of the interaction between central pathways and the lipid metabolism will provide a basis for the development of new therapies to treat specific types of dyslipidemias, by using drugs that target neural pathways. Alternatively, research directed to elucidate the specific role of the CNS on lipemic control might allow to expand the indications of treatments that act targeting neuronal pathways and that had been originally designed to palliate other alterations associated to the metabolic syndrome, such as hypertension, obesity, or diabetes. Finally, future research should also address the question of how brain cholesterol metabolism may influence plasma cholesterol levels.

Declaration of interest: The authors state no conflict of interest and have received no payment in preparation of this manuscript.