Metabolic actions of angiotensin receptor antagonists: PPAR-γ agonist actions or a class effect?

Abstract

Accumulating basic and clinical data support the hypothesis that angiotensin receptor blockers have beneficial effects on glucose and lipid metabolism that are not shared by other classes of antihypertensive agents. These metabolic actions might only partially be shared by angiotensin-converting enzyme inhibitors. Specific benefits beyond those of other angiotensin receptor blockers have been claimed for telemesartan and, to a lesser extent, irbesartan based on a partial agonist action on PPAR-γ receptors. Although the evidence is strong in vitro, specific actions not shared by other angiotensin receptor blockers have not yet been convincingly demonstrated in vivo or in clinical trials. In many cases, a full range of doses has not been compared, and the apparent superiority of telmesartan could be an artifact of its higher receptor binding affinity, greater tissue penetration owing to lipophilicity, and longer half life.

Introduction: importance of the metabolic actions of antihypertensive agents

Many antihypertensive agents are efficacious and lack serious side effects; however, hypertension rarely occurs in isolation, and there is increasing interest in the impact of antihypertensive agents on common accompanying conditions. Insulin resistance and hyperlipidemia commonly occur along with hypertension, a cluster of conditions known as metabolic syndrome or prediabetes that leads to increased cardiovascular disease independent of the development of type 2 diabetes [1]. Although there is controversy over whether the separate risk factors comprising this syndrome have multiplicative or additive effects, there is agreement that they commonly occur together[2,3].

Metabolic syndrome is commonly treated with multiple agents targeting separate abnormalities, with multiple agents being needed for the tight control of each risk factor. Antihypertensives with beneficial metabolic effects could improve control of other risk factors, notably plasma glucose and lipids. Generally, thiazide diuretics and β-adrenergic receptor antagonists have slight adverse effects, whereas α₁-adrenergic receptor antagonists and inhibitors of the renin-angiotensin system (RAS) elicit significant benefits [4–7].

Clinical trials comparing different classes of antihypertensive agents have produced conflicting results. Limitations of clinical studies include marked heterogeneity between subjects regarding medical history, diet, exercise levels and levels of risk factors other than blood pressure, which include tobacco use, psychiatric conditions such as depression, and educational and socioeconomic background. These factors also influence compliance with the prescribed therapeutic plan. Few studies have focused on hypertensive patients with metabolic syndrome, who are the most likely to benefit from antihypertensive agents with additional pharmacological actions. Preclinical trials in animal models overcome almost all of these limitations. Together with mechanistic studies at the cellular and molecular level, these laboratory studies provide the clearest insight into distinct actions of drugs. Previous laboratory studies of the metabolic effects of antihypertensives are few in number and many have significant problems. Most studies compared one or two antihypertensive agents, and failed to characterize dose–response relationships that can lead to misleading results. Furthermore, most studies used hypertensive models or metabolically disturbed animals, but seldom studied animals that were both hypertensive and metabolically abnormal, a combination of abnormalities closer to the typical clinical picture [8].

Metabolic effects of inhibiting the RAS

Angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) are increasingly being seen as the treatments of choice for hypertensive patients with metabolic syndrome. ACE inhibitors and ARBs have been shown to slightly improve insulin resistance without affecting circulating lipids or body weight [9]. Both ACE inhibitors and ARBs reduce the incidence of new cases of type 2 diabetes [10,11]. Possible mechanisms for this apparent antidiabetic effect include hemodynamic changes improving substrate delivery, cross-talk between angiotensin and insulin receptor signaling pathways, and prevention of the adverse pancreatic actions of angiotensin [12•].

A major difference between ACE inhibitors and ARBs is that ACE inhibitors have the additional property of increasing levels of the vasodilator peptide bradykinin. Treatment with a bradykinin receptor antagonist blocked the beneficial effects of the ACE inhibitor ramapril on insulin resistance in the fructose-fed rat model, suggesting that bradykinin was responsible for the beneficial effect [13]. Bradykinin receptor antagonist treatment did not attenuate the antihypertensive effect, suggesting a separation between the hemodynamic and metabolic actions of bradykinin. If bradykinin mediates the actions of ACE inhibitors, then ARBs should not affect glucose and lipid metabolism. By contrast, some investigators have reported that ACE inhibitors and ARBs have equal effects on metabolism, and that blockade of bradykinin receptors has no influence [14]. Consistent with the latter result, angiotensin has been proposed to contribute to insulin resistance and diabetes [10,12•]. Thus, the majority of evidence favors angiotensin inhibition as the most significant mechanism in the improvement in glucose and lipid metabolism. Surprisingly, the metabolic effects of renin inhibitors are currently unknown.

Metabolic actions of AT₁ receptor antagonists

Angiotensin II affects glucose and lipid metabolism through multiple direct and indirect mechanisms, as shown in Figure 1 and discussed in detail below. Unfortunately, studies into the interactions between the RAS and glucose metabolism have produced an array of contradictory results. Angiotensin II appears to have opposing immediate or long-term effects. The major angiotensin receptor subtypes, AT₁ and AT₂, usually mediate opposite actions, such as AT₁-mediated vasoconstricton and AT₂-mediated vasodilation [15]. Blockade of AT₁ receptors leads to compensatory increases in angiotensin II levels and the subsequent increased activation of AT₂ receptors. Thus, some of the actions of AT₁ antagonists might reflect increased AT₂ receptor stimulation.

Figure 1.

Metabolic effects of AT₁ receptors. The schematic shows the influence of angiotensin II on the adipose tissue, liver, skeletal muscle, pancreatic β-cells, and the SNS. Adipocytes are a major source of circulating angiotensinogen, along with the liver. In adipose tissue, AT₁ receptors inhibit differentiation of new adipocytes, thereby reducing the ability of adipose tissue to take up glucose and lipid. In the liver, AT₁ receptors increase glycogenolysis, which favors hyperglycemia, but have an opposing action of gluconeogenesis. AT₁ receptors also promote secretion of triglycerides into the circulation. Hemodynamic effects link AT₁ receptor-mediated vasoconstriction to reduced delivery of glucose and insulin to skeletal muscle. In the pancreas, AT₁ receptors promote insulin secretion in the short-term, but chronic stimulation leads to apoptosis and a loss of function. Activation of the RAS ties into SNS overactivity through actions in the central nervous system and presynaptic effects to promote norepinephrine release. Increased catecholamines promote impairments in glucose and lipid metabolism through multiple mechanisms. Modified with permission from [7]; figure created using Smart-Draw 6 (Smart-Draw, San Diego, CA).

RAS in the endocrine pancreas

Acute application of angiotensin II can increase insulin secretion under both low glucose and high glucose [16,17]. By contrast, other reports on isolated rat and human pancreatic islets show that angiotensin II impairs the first phase of glucose-stimulated insulin secretion [18–20]. Angiotensin II also impairs insulin biosynthesis and promotes β cell apoptosis [21]. Thus, despite its acute action to stimulate release of insulin, the long-term effects of angiotensin II on the pancreas promote the development of diabetes. We speculate that, in vivo, vasoconstriction induced by angiotensin II acutely reduces the supply of glucose to the islets, thereby increasing insulin secretion, whereas, in the long-term, restricted blood flow promotes β cell apoptosis and limits β cell proliferation.

Consistent with the damaging long-term effects of angiotensin II on pancreatic function, inhibition of the RAS by either AT₁ receptor antagonists or ACE inhibitors preserves insulin secretion and prevents the development of islet fibrosis and diabetes in the ZDF rat model of type 2 diabetes [12•]. Possible mechanisms include a reduction in oxidative damage and damaging cytokines, such as transforming growth factor-β. ACE inhibitors might yield additional effects by promoting insulin secretion through accumulation of bradykinin [22].

RAS in the liver

Acutely, angiotensin II reduces both gluconeogenesis in hepatocytes and hepatic glucose output [23]. Conversely, blockade of AT₁ receptors in insulin-resistant fructose-fed rats increased hepatic glucose output [24]. Thus, the actions of angiotensin II oppose those of glucagon [25]. By contrast, angiotensin II promotes glycogenolysis in perfused liver [26], which favors hyperglycemia. It is not clear which of these hepatic actions predominate in intact systems. It appears likely that the main effects of angiotensin II on glucose metabolism are mediated outside the liver. Angiotensin II has unequivocally adverse effects on lipid metabolism: infusions increase plasma triglycerides by promoting triglyceride synthesis [27]. Conversely, AT₁ receptor antagonists ameliorate hypertriglyceridemia and fatty liver [27].

Systemic actions of the RAS

The RAS also demonstrates important interactions with the sympathetic nervous system (SNS) that might be relevant to metabolism [28]. Angiotensin II acts within the brain to increase SNS activity along with other unfavorable cardiovascular changes, and this mechanism is a primary factor in experimental models of hypertension [29]. Many ARBs are hydrophilic and do not cross the blood–brain barrier. However, telmisartan is sufficiently lipophilic to penetrate the brain and block the sympathoexcitatory effects of angiotensin II in the central nervous system. ARBs can also act directly on catecholaminergic nerve terminals to reduce the release of norepinephrine and on the adrenal medulla to reduce secretion of epinephrine [30]. Even a modest component of sympathoinhibition can contribute to favorable changes in glucose and lipid metabolism [31].

The hemodynamic effects of ARBs also contribute to their effects on glucose and lipid metabolism, because vasodilatation can facilitate disposal of glucose and lipids in peripheral tissues [32]. For example, telmisartan could have enhanced vascular effects owing to its lipophilicity. Relative to an equivalent dose of the hydrophilic ARB losartan, telmisartan improved nitric oxide release from the endothelium and reduced vascular remodeling in stroke-prone spontaneously hypertensive rats, a model of severe hypertension [33].

RAS and adipose tissue

Adipocytes are a major source of angiotensinogen in the circulation [34]. Angiotensinogen secretion increases with rising adipose mass, and thus the adipose-derived RAS might contribute to hypertension in obesity. Angiotensin II inhibits adipocyte differentiation through AT₁ receptors while promoting differentiation via AT₂ receptors [35]. Mice with a genetic deletion of AT₂ receptors have decreased adiposity [36•]. Conversely, according to at least one report, blockade of AT₁ receptors prevents differentiation of new adipocytes in obese rats [37].

A novel action of several AT₁ antagonists is partial agonist activity at the peroxisome proliferator-activated receptor γ (PPAR-γ) — a nuclear receptor expressed primarily in adipose tissue [38•,39,40]. PPAR-γ subtypes are activated by fatty acids or prostaglandins in vivo, which allows them to act as transcription factors that regulate the expression of key genes by binding to peroxisome proliferator response elements within upstream DNA sequences in promoter regions. PPAR-γ regulates the gene expression of proteins involved in the differentiation of adipocytes from precursor fibroblast-like cells and the storage of fatty acids. All PPARs dimerize with the retinoid X receptor and bind to specific response elements within DNA to activate target genes.

The most potent PPAR-γ activity is displayed by telmisartan and, to a lesser extent, irbesartan [39,41]. In a mouse embryonic fibroblast cell line (3T3-L1), telmisartan was shown to activate two target genes of PPAR-γ receptors: phosphoenolpyruvate carboxykinase-1 (PEPCK-C) and acetyl coenzyme A carboxylase 2 (ACC2). These genes are markers of the differentiation of fibroblasts into adipocytes, which is promoted by PPAR-γ agonists. The metabolic actions of telmisartan in vivo are dependent upon the presence of an intact CD36 protein, a fatty acid transporter active in adipocytes and regulated by PPAR-γ [42]. This suggests that the PPAR-γ-regulated gene critically activated by telmisartan is CD36, resulting in enhanced clearance of fatty acids from the bloodstream into adipocytes. Presumably, reduced availability of free fatty acids will reduce ectopic storage of fat in non-adipose tissues, which could be a major contributor to insulin resistance [35].

Another study showed that telmisartan reduced triglyceride accumulation in muscle and liver, and reduced visceral adiposity, in dietary obese mice [43]. This latter effect was attributed to an increase in metabolic rate; however, there is no evidence for an anti-obesity action of telmisartan in humans.

Adiponectin is a major secretory product of adipocytes and acts to improve insulin sensitivity and lessen inflammation. Angiotensin II acting through AT₁ receptors decreases circulating adiponectin [44], which contributes to insulin resistance and adipose tissue inflammation. Conversely, the AT₁ receptor antagonist candesartan and the ACE inhibitor ramapril both increase adiponectin in humans, which might contribute to the metabolic impact of this class of drug [45,46]. Telmisartan has the same effect as other AT₁ receptor antagonists [47], and so there is no reason to link the activity of telmisartan at PPAR-γ to its effects on adiponectin. However, the greater lipophilicity of telmisartan might allow it to penetrate more completely into adipose tissue, giving it improved access.

Conclusions: does a PPAR-γ agonist action of telmisartan account for its beneficial cardiometabolic effects?

The concept of an agent with dual PPAR-γ agonist and AT₁ receptor antagonist actions is indeed promising [48], particularly from the standpoint of synergistic metabolic actions. However, there are several limitations that should be noted. Firstly, the major action of the glitazones and other antidiabetic PPAR-γ agonists in vivo is the differentiation of new adipocytes, particularly in subcutaneous regions, and a reduction in the size of visceral adipocytes.

Telmisartan appears to have the opposite action, reducing the formation of new fat cells [35] and leading, in some studies, to reduced subcutaneous fat [43]. Secondly, PPAR-γ effects are elicited only by micromolar concentrations of telmisartan, whereas low nanomolar concentrations are sufficient to block AT₁ receptors [49]. Thus, telmisartan is >1000-fold more potent as an AT₁ receptor antagonist than as a PPAR-γ agonist. Furthermore, the action of telmisartan to promote adipocyte differentiation is shared by other AT₁ receptor antagonists. Thus, most of the effect of telmisartan on metabolism is likely to be mediated by blockade of AT₁ receptors, together with compensatory overstimulation of AT₂ receptors as angiotensin II levels rise during chronic AT₁ blockade. Nonetheless, some contribution of a synergistic PPAR-γ effect seems likely.

However, another limitation at the present time is that the PPAR-γ-activating action of telmisartan has yet to be demonstrated in human cells [50]. This could be significant, as the role of the RAS might differ between rodent and human adipose tissues [50]. By contrast, the metabolic effects of angiotensin II and ARBs in the pancreas, liver and autonomic nervous system have been shown to operate in humans as well as in rodents. The effectiveness of telmisartan might be adequately explained by its greater potency, longer half-life and greater lipophilicity. However, the synergy between PPAR-γ activation and angiotensin receptor blockade is clear from preclinical data. Therefore, combined trials of thiazolidinediones and ARBs for the prevention or treatment of diabetes in hypertension might be warranted.

Update

Several important new articles have appeared in the past few months. Telmisartan increased leptin levels in diabetic hypertensives, which could have contributed to improved insulin sensitivity in the absence of changes in body weight [51]. Additional clinical trials confirmed the ability of telmisartan to reduce insulin resistance and fasting insulin levels [52,53], and improve pancreatic β cell function [53].

Telmisartan has now been shown for the first time to induce differentiation of human adipocytes [54••]. Furthermore, in human hepatoma cells, telmisartan has been shown to reduce levels of C-reactive protein expression through a specific PPAR-γ agonist action [55]. This action was not shared by candesartan, an ARB that lacks PPAR-γ effects, but was blocked by a selective PPAR-γ antagonist. Importantly, these are the first clear demonstrations of PPAR-γ actions of telmisartan in human cells. The data in hepatoma cells imply that telmisartan may have anti-inflammatory effects that reduce vascular injury, independent of blood pressure lowering. In addition, telmisartan might have a direct suppressive action on lymphocyte expression of pro-inflammatory surface proteins [56].

Telmisartan has now been shown to retard the progression of atherosclerosis in the apolipoprotein-E-deficient mouse model through an antioxidant action [57]. Conversely, other investigators have questioned the relevance of PPAR-γ agonist actions in the effects of telmisartan on insulin sensitivity, noting that effects are much greater in cell-free systems than in intact cells or in vivo [58].