Major shifts in human lifestyle and dietary habits toward sedentary behavior and refined food intake triggered steep increase in the incidence of metabolic disorders including obesity and Type 2 diabetes. Patients with metabolic disease are at a high risk of cardiovascular complications ranging from microvascular dysfunction to cardiometabolic syndromes including heart failure. Despite significant advances in the standards of care for obese and diabetic patients, current therapeutic approaches are not always successful in averting the accompanying cardiovascular deterioration. There is a strong relationship between adipose inflammation seen in metabolic disorders and detrimental changes in cardiovascular structure and function. The particular importance of epicardial and perivascular adipose pools emerged as main modulators of the physiology or pathology of heart and blood vessels. Here, we review the peculiarities of these two fat depots in terms of their origin, function, and pathological changes during metabolic deterioration. We highlight the rationale for pharmacological targeting of the perivascular and epicardial adipose tissue or associated signaling pathways as potential disease modifying approaches in cardiometabolic syndromes.
Metabolic disorders including Type 2 diabetes (T2DM) and obesity are steadily increasing global health threats associated with significant morbidity and mortality. T2DM is a major risk factor for coronary artery disease (CAD), stroke, and peripheral vascular diseases , as well as microvascular dysfunction leading to renal failure and blindness . As well, obesity predicts incidence of heart failure [3–5], most notably in women toward heart failure with preserved ejection fraction (HFpEF) [6,7]. Therefore, increased adiposity is a major driving force for the development of cardiovascular complications in metabolic disorders. Recently, the idea of adipose tissue as a neutral storage organ for triglycerides has been challenged. In fact, adipose tissue as an endocrine organ is increasingly recognized, especially as it pertains to glucose and lipid homeostasis . In this regard, studies have shown that adipose tissue inflammation contributes to insulin resistance and poor glycemic control [9–11]. Prolonged caloric excess triggers chronic inflammatory changes in adipose tissue and several types of immune cells and inflammatory cytokines have been implicated . However, the mechanistic and temporal association between adipose tissue inflammation and cardiovascular dysfunction triggered by metabolic disease has only been recently recognized. The heterogeneity of adipose tissue depots of the body and the breadth of signaling pathways involved are underscored by each unique location, biochemical, developmental, and structural properties [13,14]. This offers a valuable therapeutic opportunity to intervene with cardiovascular complications of metabolic disease by harnessing adipose depot-specific targets or modulating the signaling pathways of adipose tissue dysfunction. In this review, we will summarize the physiological and pathological features of the major fat depots of the heart and blood vessels, the epicardial (EAT) and perivascular adipose tissues (PVAT). We will provide an overview of possible therapeutic interventions specifically modifying pathological processes in these adipose depots. Whenever possible, we will also extrapolate known actions in other adipose depots with perceived similarities to PVAT/EAT as potential targets.
Perivascular adipose tissue
Structural and functional roles
Perivascular adipose tissue (PVAT) is an adipose tissue depot that is contiguous with the tunica externa of most arteries and veins with an internal diameter greater than 100 µm, including small vessels like subcutaneous small arteries and coronary arteries, as well as large systemic vessels like the aorta, with the exception of the cerebral and micro-vasculature [15,16]. The physical proximity of this adipose layer to blood vessels and its intertwinement with fibrous adventitia provides a cushioning effect that both protects vessels and insulates them from their surroundings . Initially, the role of PVAT has been limited to this structural function, but in 1991 Soltis et al. demonstrated that aortas with intact periaortic fat displayed a decrease in norepinephrine-induced contractility compared with those in which the PVAT has been removed . This anti-contractile effect remained when vessels without adipose tissue were treated with PVAT-conditioned media . The name adipocyte derived relaxing factor (ADRF) was coined, and since then the ability of PVAT to reduce the contractility of vascular smooth muscle cells (VSMCs) has been well documented [20–24]. This suggests a mechanistic pathway by which PVAT secretes a complex array of factors to modulate vascular tone at the level of the tunica externa, thus regulating cells of the tunica media and intima, and ultimately vascular health and disease.
Like other adipose tissue depots, PVAT plays a fundamental role as an endocrine organ, secreting a myriad of cytokines, chemokines, and hormones that are collectively referred to as adipokines . The location of PVAT confers the unique ability to communicate with VSMCs and endothelial cells in a paracrine manner. This localized interaction provides a direct mechanism through which diseases involving adipose dysfunction, such as T2DM and metabolic syndrome, may impair vascular function.
PVAT as compared to other adipose depots
Healthy PVAT is composed of adipocytes, stromal cells, microvasculature, nerves, stem cells, and immune cells. Phenotype and gene expression studies showed that PVAT could demonstrate white and brown adipose properties in different vascular beds. Early investigation showed that periaortic and perirenal PVAT expressed brown adipose-specific genes but at a much lower level compared with bona fide brown adipose pools . Similar findings were reported for peri-coronary adipose tissue showing mixed gene expression patterns [27,28]. Yet, other studies showed that variations in PVAT cell composition and gene expression profiles exist across vascular beds and even at different sites within the same vessel. For instance, while peri-aortic PVAT tended to bear closer resemblance to brown adipose tissue, mesenteric PVAT showed the characteristics of white adipose pools [29,30]. Moreover, PVAT transitions from having mostly brown-like features to resembling white adipose tissue as it traverses the thoracic aorta to the abdominal aorta [31,32]. The ability to “beige”, or undergo transition from white to brown-like features in response to given environmental stimuli, is observed in white adipocytes forming beige adipocytes existing in predominantly white adipose beds . However, PVAT cells have a developmental origin distinct from white and brown adipocytes, likely derived from vascular smooth muscle precursor cells . Whether this distinct origin allow PVAT adipocytes more plasticity in switching between white and brown phenotypes and the consequent impact on vascular function are unclear. Evidence indicates that PVAT appeared to be in a proinflammatory state [35,36] that contributed to vascular dysfunction in various disorders. Although uncoupling protein-1 (UCP1) expression was confirmed in PVAT from various vascular beds in mice, rats, and humans [29,30,37], variations in its expression and function have not been systematically examined in relation to the observed proinflammatory phenotype.
PVAT/vascular dysfunction in metabolic disorders
Several epidemiological studies examined the relationship between PVAT dysfunction and several vascular disorders. Not only was thoracic peri-aortic fat correlated with a number of markers of metabolic risk including glucose intolerance, dyslipidemia, and diabetes, but also appeared tightly associated with vascular diseases such as atherosclerosis, hypertension, peripheral artery disease, thoracic and abdominal aortic calcification, and CAD [38,39]. In the context of CAD, extensive research showed that changes in coronary artery PVAT is tightly associated with pathological changes leading to myocardial infarction [40,41]. Indeed, multiple changes in the adipokine secretory profile and inflammatory cytokine expression is seen in the coronary PVAT associated with CAD or coronary arterial dysfunction . These data highlight the role of PVAT in transducing metabolic dysfunction to vascular pathologies.
Structural and functional alterations
While a large body of literature investigates phenotypic and genetic changes in visceral adipose depots in various stages of metabolic dysfunction, a dearth of evidence exists for similar changes in PVAT. Animal studies of diet-induced obesity examining PVAT describe an increase in adipocyte diameter that is usually accompanied by increased inflammatory cell infiltration and loss of the anticontractile effect . Long-term high-fat feeding in mice resulted in distorted PVAT lipid droplet architecture, with droplet enlargement and coalescence, together with signs of mitochondrial damage and reduced PPARγ expression . The previous studies and others examined the impact of metabolic deterioration on PVAT among a larger context of systemic adipose tissue dysfunction. Interestingly, recent studies from independent groups examined the vascular consequences of isolated PVAT dysfunction. One study reported structural PVAT changes including adipocyte hypertrophy together with reduced adipocyte-specific gene expression including PPARγ as the earliest signs of adipose dysfunction observed after a mild increase in calorie intake from fat . Interestingly, this was accompanied by signs of PVAT hypoxia, mitochondrial dysfunction, inflammatory cell infiltration, and increased inflammatory cytokine release, with no indication of systemic inflammation. Significantly, while these changes were associated with signs of increased vascular contractility in multiple vascular beds and an augmented in vivo pressor response, they were not observed in several visceral adipose depots. A follow-up study demonstrated a compromised endothelial function due to a deficiency in K+ channel activity under the same circumstances . The second study demonstrated, through PVAT transplantation from high-fat fed to control mice, that localized PVAT inflammation could induce remote vascular effects that are independent of the status of other adipose depots .
PVAT inflammatory changes
Hyperinsulinemia and increased caloric intake characteristic to metabolic syndrome are the main drivers of adipose tissue inflammation. The commonly accepted model is that hyperinsulinemia contributes to adipose inflammation by triggering adipocyte hypertrophy . Increased adipocyte diameter beyond the diffusion distance of oxygen, without a commensurate increase in vascularization promotes a hypoxic environment in the expanding adipose tissue . This hypoxic response is thought to lead to immune cell chemotaxis and inflammation . Indeed, increased levels of hypoxia-inducible factor-1α (HIF-1α) expression were observed in vitro in response to obesity-related factors such as insulin and adipogenesis  and in adipose tissue of several models obese animals . While other transcription factors, including nuclear factor-κB (NF-κB), are involved in the hypoxic adipose tissue response , there appears to be a significant cross-talk between their function and that of HIF-1α. Whereas NF-κB was shown to regulate HIF-1α transcription [52,53], previous studies showed that hypoxia-upregulated HIF1-α expression triggered NF-κB-mediated cytokine production, including IL-1β [54,55]. Further studies showed that this pathway involved the increased activation the κB inhibitor kinase, increased κB inhibitor (IκBα) phosphorylation, and increased NF-κB expression and activation [56,57]. Genetic activation or ablation of HIF-1α increased or decreased κB inhibitor kinase activity, κB inhibitor phosphorylation and NF-κB expression and activation, respectively. The ensuing inflammatory reaction in PVAT involves the recruitment of multiple inflammatory cell types and the increased production of large profile (∼30) of inflammatory cytokines that were reviewed elsewhere [12,58]. The release of these mediators in a paracrine manner affects the structure and function of the vasculature in the immediate vicinity. Examination of localized PVAT inflammation in a rat model of early metabolic dysfunction revealed increased IL-1β and TGF-β1 production, which was associated with reduced AMP-activated protein kinase (AMPK) activation, increased vascular Erk1/2 phsophorylation, medial hypertrophy, oxidative stress, increased rho-associated kinase (ROCK)-mediated calcium sensitization and a hypercontractile response [44,59]. This isolated PVAT inflammation model was also associated with impaired endothelial relaxing function due to reduced expression/function of inward rectifier K+ channels . Contractile and endothelial manifestations were reversed by treatment of the local PVAT inflammation.
While this inflammatory framework is no different in PVAT compared with other adipose depots susceptible to inflammation in metabolic syndrome, diabetes, and obesity, the mechanistic context underlying the previously described “proinflammatory” phenotype of PVAT  and its increased vulnerability to inflammation in early stages of metabolic dysfunction  remains uncertain. A likely explanation pertains to the unique nature of PVAT as an adipose depot with both white and brown characteristics, including UCP1 expression [60,61]. While the role of UCP1 is to increase mitochondrial energy dissipation, its expression level is associated with increased oxygen consumption [62–64]. Previous results showed that selective insulin resistance in brown adipose tissue, evoked by local insulin receptor knockdown in mice, led to increased UCP1 expression . As well, while multiple prior studies showed up-regulation of UCP1 expression in brown adipose tissue in response to an increased caloric intake challenge , only a recent study showed increased PVAT UCP1 expression in response to high-fat feeding . The latter study also reported an increased PVAT adipocyte size in line with the changes seen in white adipose tissue during adipose expansion triggered by high-fat feeding. These observations that were associated with PVAT hypoxia and inflammation were not observed in white adipose tissue depots from the same rat. Particularly, diet-induced UCP1 up-regulation is thought to be driven by elevated sympathetic stimulation . Indeed, increased sympathetic activity is a common observation in metabolic disorders, even in early stages , and thought to be triggered by elevated levels of circulating insulin . In fact, the very same rat model of early metabolic challenge, isolated PVAT inflammation, and increased PVAT UCP1 expression showed an altered autonomic activity balance towards increased sympathetic activity . With dense sympathetic innervation , it could be plausible that diet-induced increase in UCP1 expression in PVAT with the associated oxygen consumption causes it to be more sensitive to hypoxia consequent to early adipocyte hypertrophy compared to white adipose depots. In parallel, an additional mechanism could be proposed for the observed disparity in sensitivity to inflammation between PVAT and visceral white adipose depots. Differences in microenvironment among various adipose depots lead to unequal expansion in response to metabolic stress [71,72]. In these studies, different adipose depots showed altered rates of adipocyte precursor recruitment, differentiation, adipogenesis, and patterns of expansion in both an anatomic and gender-specific manner. Although none of these mechanisms were explored specifically in the PVAT, a model for PVAT inflammation induced vascular dysfunction has been proposed (Figure 1).
The proposed model for induction of localized PVAT inflammation and subsequent vascular dysfunction in early stages of metabolic deterioration
Of note, a component of adipose tissue inflammation with relevance to vascular function is often overlooked in the context of metabolic disorders. Specifically, inflamed adipose tissue, common to these conditions, produces an increased amount of tissue factor, which in turn activates the clotting cascade leading to elevated levels of activated thrombin and factor X (Xa) [73–75]. Both thrombin and Xa are endogenous ligands of protease-activated receptors (PARs). Dysfunctional PAR signaling downstream of either factor was implicated in an impaired vascular phenotype including altered endothelial cell motility and increased permeability, increased EC production of adhesion molecules and inflammatory cytokines, and altered endothelium-dependent vasomotor tone [76,77], as well as VSMC phenotypic alteration associated with vascular remodeling and atherosclerosis [78–81]. Interestingly, both the progression of adipose tissue inflammation and vascular dysfunction downstream of PAR signaling appear to occur via common targets including HIF1-α, NF-κB, ROCK, and cyclooxygenase-2 [44,81–84]. Significantly, tissue factor/PAR signaling in adipocytes and immune cells could possibly provide feedback augmentation of AT inflammation [74,85]. Yet again, the role of these mediators was not evaluated in PVAT inflammation and the ensuing vascular dysfunction.
Alteration of PVAT adipokine secretasome
Multiple adipokines are involved in the control of vascular contractility, either by acting directly on VSMCs or by modulating the function of endothelial cells. Several studies have shown that the major anticontractile effect of PVAT-derived adipokines under normal conditions is lost in certain disease states like diabetes. For example, Azul et al. have shown that the anticontractile effect of PVAT in response to phenylephrine stimulation was lost in Goto-Kakizaki diabetic rats accompanied by a shift in PVAT phenotype into a proinflammatory state . In the below sections, changes in the production of individual adipokine by PVAT in metabolic disorders and the consequent vascular abnormalities will be discussed.
Adiponectin is a 30-kD adipokine that is known for its insulin-sensitizing effects and its protective role against cardiovascular dysfunction . Obesity is associated with a decrease in adiponectin levels, despite the increase in adipose tissue mass. Low plasma adiponectin level is a hallmark of insulin resistance associated with obesity and T2DM . On a systemic level, decreased adiponectin production indirectly affects the vasculature by promoting a hyperglycemic environment that brings about endothelial dysfunction, oxidative stress, and inflammation [89–91]. In addition, adiponectin plays a more direct paracrine role in mediating vascular contractility by stimulating AMPK activation, which in turn promotes vascular relaxation by two mechanisms: The first is by increasing phosphorylation of endothelial nitric oxide synthase (eNOS), leading to an increase in nitric oxide (NO). In addition, AMPK has a direct role in phosphorylating calcium-activated potassium channels, bypassing the endothelium in mediating vascular relaxation . Adiponectin is also a potent inhibitor of VSMC proliferation, and its marked decrease in obesity contributes to the proliferative phenotype acquired by VSMCs during this state . As expected, reduced PVAT adiponectin production in isolated PVAT inflammation was associated with increased atherosclerotic lesions and impaired endothelial function in mice subjected to a high-fat diet .
Leptin is a major adipokine that mediates a direct vasodilatory effect on blood vessels by enhancing endothelium-dependent vasodilation . This, in turn, increases adipocyte lipid metabolism, reminiscent of the acute vasodilatory effect of insulin [95,96]. However, similar to insulin, chronic elevations in leptin levels may mediate an unfavorable clinical picture locally by increasing endothelial oxidative stress, and thus reducing NO production or activity. Moreover, studies elucidating the vasodilatory effects of leptin were carried out in vitro or by infusing leptin to ob/ob mice that lack leptin, a setup that is quite different from the hyperleptinemic state that characterizes obesity . In fact, the hyperleptinemia of obesity leads to leptin resistance, especially to its hypothalamic function of mediating satiety . Yet, hyperleptinemia that accompanies the metabolic syndrome is associated with reduced vascular distensibility and hypertension, possibly due to sympathetic overactivation . Locally, the increase in leptin secreted by PVAT from obese models has been shown to stimulate VSMC proliferation to a greater degree than from control animals. However, this effect was ameliorated in obese Zucker rats deficient in leptin receptors, indicating that leptin resistance in obesity may put the effect of this adipokine in terms of vascular remodeling into question . Surprisingly, a study investigating hyperleptinemia versus leptin resistance as a cause of the cardiovascular consequences of obesity showed that intact endothelial leptin signaling limits neointimal formation, and that obesity represents a state of endothelial leptin resistance . Interestingly, studies on rat aortic VSMCs show that leptin plays a proliferative role via up-regulation of cyclin D1, ERK 1/2, and NF-kB, indicating that leptin has seemingly opposing effects on these two cell types [101,102], and leptin resistance at the level of the endothelium and smooth muscle may ameliorate both its beneficial and harmful effects . This is exemplified by the finding that human carotid VSMCs from patients with atherosclerosis decreased in proliferation and downregulated their leptin receptor upon exposure to leptin .
Like leptin, chemerin and resistin are two adipokines that have been associated with endothelial cell proliferation, angiogenesis, oxidative stress, and the expression of adhesion molecules [104,105]. Furthermore, chemerin has been shown to have a dose-dependent pro-contractile effect on the vasculature that is enhanced in arteries from obese humans, indicating that resistance to the hypertensive effect of chemerin does not occur in obesity, despite the elevation in its levels . Shortly after the discovery of ADRF, an adipokine known as visfatin was identified as the first known growth factor released by PVAT. Visfatin has a pro-survival effect on VSMCs, endothelial cells, and macrophages, and is implicated in the development of atherosclerosis by inducing the proliferation of cells that are required for the formation of a stable atheromatous plaque . Its proliferative effect on VSMCs is mediated by activation of the ERK1/2 and p38 MAPK pathway . Interestingly, visfatin levels were shown to be higher in PVAT than in subcutaneous and visceral adipose tissue .
Fatty acid binding protein 4 (FABP4) is a transport protein released by PVAT whose natural ligands include oleic, retinoic, and arachidonic acid, and functions to deliver these fatty acids to PPARγ . Recently, local FABP4 production by PVAT has been shown to increase macrophage, VSMC, and endothelial inflammatory marker expression in a dose-dependent manner, and has been correlated with coronary artery stenosis . Furthermore, the difference in FABP4 level between the aortic root and coronary sinus of patients with CAD was more strongly correlated with coronary artery stenosis than its independent levels in either arterial or venous circulation, indicating that FABP4 mediates its effect in a paracrine manner facilitated by the local production of FAB4 by coronary PVAT .
Alteration of the renin/angiotensin-II and angiotensin 1–7/Mas receptor functions
The renin–angiotensin–aldosterone system (RAAS) is the main humoral signaling pathway responsible for long-term regulation of arterial tone, fluid/electrolyte balance, and blood pressure through effects on autonomic nervous system, heart, kidneys, and vascular tissue . Disturbances of RAAS activation and function are implicated in many cardiovascular disorders including hypertension, myocardial ischemia, atherosclerosis, aortic aneurysms, cardiac arrhythmias, and heart failure . Classically, this pathway is initiated by renal renin secretion; in response to sympathetic stimulation, hypotension, hyponatremia, and renal hypoperfusion, which in turn acts on angiotensinogen producing angiotensin-I. The latter is converted by ACE into angiotensin-II, which activates angiotensin-II receptors, AT1 and AT2 . Alternatively, angiotensin-I can be processed by ACE2 producing angiotensin 1–7, which shows an opposite, i.e. more favorable effect by activation of Mas receptors . Whereas exaggerated angiotensin-II activity led to hypertension due to vasoconstriction and sodium and water retention, impairment of baroreflex, in addition to vascular and cardiac remodeling, angiotensin 1–7 reduced blood pressure in hypertensive animals, promoted baroreflex sensitivity, produced vasodilation, and improved endothelial function and vascular and cardiac remodeling . Interestingly, recent research showed that adipose tissue expressed all components of RAAS pathway and actively secreted angiotensin-II contributing ∼30% of the circulating levels in rodents . Research has shown that angiotensin-II production in adipose tissue is up-regulated consequent to increases in caloric intake and is associated with inflammatory changes that are observed in obesity [115,116]. Moreover, early studies showed that AT1 receptor blockade reduced subcutaneous and non-vascular visceral adipose tissue oxidative stress and inflammation in addition to decreased atherosclerosis in mouse models of obesity [117,118]. Despite the reported expression of all RAAS component proteins in PVAT , there is a gap of knowledge regarding changes in the RAAS activity localized to PVAT, as well as alterations in status of ACE2 activity, angiotensin 1–7 production, and Mas receptor function in metabolic disorders. Few studies examined the vascular consequences of PVAT-specific modulation of RAAS during inflammation. PVAT-produced angiotensin-II was shown to increase nerve-mediated mesenteric arteriole contraction . Interestingly, in vitro simulation of PVAT hypoxia in small arterioles led to a loss of its anticontractile effect that recovered after exposure to AT1 receptor blockers . Furthermore, deletion of AT1a receptors in peri-aortic PVAT in mice reduced macrophage infiltration and aneurysm formation . Significantly, a recent study demonstrated that PVAT-derived angiotensin 1–7 contributed to its anticontractile effect under normal conditions through the Mas receptor-mediated activation of PI3K/Akt pathway and subsequent increase in NO activity . However, whether this was altered in conditions of PVAT inflammation was not examined.
On the other hand, a reasonably large body of evidence describes the role of aldosterone and mineralocorticoid receptor in PVAT inflammation and vascular dysfunction in metabolic disease. Systemic aldosterone levels are elevated in models of diabetes and obesity associated with an increased mineralocorticoid receptor signaling-mediated vascular dysfunction . In a more local context, increased PVAT-derived aldosterone in obesity is associated with augmented PVAT and vascular oxidative stress, impaired endothelial function, vascular contractile dysfunction, and increased macrophage infiltration, inflammatory cytokine production, and vascular stiffness [124,125]. Specifically, a recent study showed that aldosterone receptor blockade restored PVAT anticontractile function in a mouse model of obesity via normalization of ROCK-mediated calcium sensitization .
Senescence refers to the stress and damage response that arrests cell growth and function in an attempt to prevent the replication and translation of potentially damaged genetic material . This is mediated by inducing a global change in cell structure and function, including DNA methylation, heterochromatin accumulation, chemical modification of both structural proteins and enzymes, mitochondrial dysfunction, and production of a senescence-associated secretory phenotype . As well, senescence is associated with increased production of matrix metalloproteases and interleukins, which promote a proinflammatory environment that is at odds with the purpose of the senescent response . Metabolically impaired states like obesity and the metabolic syndrome accelerates this phenomenon by providing continuous cellular insults and induce premature senescence . In addition, senescent adipocytes have directly been implicated in the development of insulin resistance , indicating that this process begins early on in the pathogenesis of diabetes. For example, increased expression of senescence marker p53 has been observed in mice fed a hypercaloric diet, and its inhibition was shown to improve insulin resistance . In fact, one of the mechanisms by which metformin may mediate its therapeutic benefits is by inhibiting the senescence-associated secretory phenotype through interfering with proinflammatory NFkB signaling . To date, few studies have investigated PVAT senescence in the context of metabolic disease. LeFranc et al. showed that obesity induced premature aging in PVAT by uncoupling the mitochondrial electron transport chain and thus increasing reactive oxygen species production. This in turn was associated with loss of PVAT anticontractile effect on VSMCs. Interestingly, mitochondrial damage was reversed, in part, by the administration of mineralocorticoid receptor antagonists, making them a potential therapeutic option for targeting PVAT dysfunction in obesity . Moreover, senescent PVAT stromal cells display a transcriptional profile that favors a decrease in PGC1α expression, resulting in a reduced potential to differentiate into brown adipocytes. Furthermore, aged PVAT stromal cell-derived adipocytes were shown to promote VSMC proliferation and contribute to vascular remodeling during aging . Of note, these findings were observed in aged cells, but it has been shown that diabetes and aging impart similar pro-senescence changes .
Cardiac adipose tissue
Epicardial and pericardial depots
Adipose tissue of the heart is distinguished based on anatomic location into two layers: (1) the epicardial adipose tissue (EAT) and (2) the pericardial adipose tissue (PAT). Inconsistencies in the anatomic location of the EAT and PAT are prevalent in the literature, therefore findings from different studies should be examined thoroughly to understand the specific depot assessed [133,134]. The EAT is a visceral fat depot localized between the visceral pericardium and myocardium, whereas the PAT resides within the pericardial cavity on the outer surface of the parietal pericardium (Figure 2A) [135,136]. Therefore, the PAT does not share physical proximity with the myocardium. In contrast, the EAT surrounds the adventitia of the coronary arteries and lies adjacent to the myocardium without structural boundary, therefore these two layers share a common microcirculation [135,137–139] allowing local modulation of both the myocardium and coronary vessels [140,141]. In contrast, the PAT is supplied by a branch of the internal thoracic artery termed the pericardiacophrenic artery .
Local inflammatory changes in EAT in obesity
Anatomically, the EAT contributes to approximately 15% of cardiac mass in physiological conditions . It is primarily located in the atrioventricular and interventricular grooves; however as the amount of epicardial fat increases, it progressively expands to cover the ventricles and the epicardial surface in its entirety . Interestingly, ventricular mass is positively correlated to EAT mass and no such correlation exists with overall adiposity (BMI). Therefore, myocardial hypertrophy is a greater predictor of EAT mass than both overweight and obesity . However, studies also demonstrate the clinical association between EAT thickness/volume with metabolic syndrome and atherosclerosis [133,143]. Indeed, clinical imaging studies with echocardiography indicate that obesity predisposes the accumulation of epicardial fat; however they concluded that fat distribution, specifically abdominal fat accumulation (determined by waist circumference), was a stronger correlate for the amount of epicardial fat . Taken together, these studies indicate that fat distribution is the essential parameter, and increased waist circumference is a positive correlate of EAT. The EAT and PAT also differ by embryologic origin and biochemical properties. The EAT originates from the splanchonopleuric mesoderm of the gut, therefore is of the same embryologic origin as mesenteric and omental adipocytes . In contrast, the pericardial fat is derived from the primitive thoracic mesenchyme, which divides to form the parietal pericardium and the external thoracic wall . The EAT is a heterogeneous tissue, consisting of adipocytes, nervous and nodal tissue and resident immune cells [136,145].
The EAT is distinct from other fat depots with a smaller adipocyte size, high protein content, and higher rates of fatty acid incorporation and synthesis . Physiologically, the EAT maintains fatty acid homeostasis; acting as a local source of free fatty acids (FA) and mediating their sequestration when levels escalate. As FA oxidation meets 50–70% of the metabolic demand of the heart, the EAT may maintain the energy demand and incorporate excess FA to circumvent lipotoxicity .
The EAT is a source of bioactive molecules, including adiponectin, leptin, apelin, and cytokines [14,137,148] (Table 1); therefore, it plays a role in determining the inflammatory status of the surrounding myocardium in physiology and disease. As discussed previously, adiponectin inhibits platelet aggregation and macrophage activation and is responsible for insulin sensitivity, glucose uptake and lipid catabolism through its activation of AMPK [88,149,150]. On the other hand, leptin signaling in the cardiovascular system is associated with pro-angiogenic activity, and promotion of atherosclerotic lesions [14,149]. Apelin, a factor secreted by adipocytes, vascular stromal cells and the myocardium, confers cardioprotection, vasodilation, is anti-atherosclerotic, and improves cardiac metabolism in obesity-related HF [14,149]. Physiologically, a balance is sustained between harmful and protective factors; however, pathology denotes an alteration that favours inflammation.
|Category .||Factor .||Alterations .||Role .||Expression .||References .|
|MCP-1||Increased||Proinflammatory (chemoattractant)||protein, mRNA|||
|IL-6||Increased||Proinflammatory (angiogenesis)||protein, mRNA, secretion||[145,155,156]|
|Category .||Factor .||Alterations .||Role .||Expression .||References .|
|MCP-1||Increased||Proinflammatory (chemoattractant)||protein, mRNA|||
|IL-6||Increased||Proinflammatory (angiogenesis)||protein, mRNA, secretion||[145,155,156]|
EAT facilitates and potentiates local inflammation
Under physiological conditions, the EAT mediates positive effects, such as regulating FA homeostasis to circumvent lipotoxicity, serving as a fuel source for the myocardium, and secreting anti-inflammatory and anti-atherogenic adipokines [136,143]. However, as the metabolic profile shifts in obesity, the EAT secretasome is altered and the EAT becomes pathogenic, shifting the adipokine balance toward an up-regulated release of FA and proinflammatory cytokines [136,147] (Table 1). Indeed, resident immune cells of the adipose tissue demonstrate significant plasticity, allowing the transition from a protective to a detrimental phenotype. The context is essential as mast cells, macrophages and neutrophils can both potentiate inflammation or promote its resolution depending on the presence of extrinsic pathogens or danger signals [151,152]. As the EAT expands, extensive immune cells infiltrate and generate a proinflammatory microenvironment, a discovery that prompted the investigation of specific immune cells and cytokine profile of adipose tissue as key to metabolic inflammation [136,152,153]. Key alterations in cytokine and adipokine profile as a result of EAT inflammation are summarized in Table 1 [14,145,154–159].
EAT harbor mast cells , macrophages , dendritic cells  and lymphocytes  (Figure 2B). In fact, Mazurek et al. characterized the presence of inflammatory cells in the EAT compared with subcutaneous fat. They identified the presence of T lymphocytes (CD3+), macrophages (CD68+) and mast cells (tryptase+) unique to EAT compared with subcutaneous adipose tissue. No cellular retention was observed in subcutaneous fat in these studies . Further, the specific cytokine and chemokine profile was assessed from plasma and conditioned media of cultured adipose tissue from both subcutaneous and epicardial adipose tissue of patients with CAD. Protein and mRNA levels of interleukin-6 (IL-6), interleukin-1-beta (IL-1β), monocyte chemotactic protein-1 (MCP-1), and tumor necrosis factor alpha (TNF-α) were significantly elevated in EAT . However, this study was limited to patients with CAD undergoing coronary bypass surgery, therefore the physiological immune cells of healthy patients was not assessed . In a more recent study, the impact of macrophages in EAT inflammation has been investigated in CAD and non-CAD patients undergoing elective cardiac surgery. Macrophage polarization was examined by immunohistochemical staining of CD11c and CD206: two macrophage markers indicative of proinflammatory M1 phenotype and anti-inflammatory M2 phenotype, respectively . CAD patients demonstrated an increased ratio of M1/M2 phenotype compared with non-CAD. Further, this ratio was positively correlated with the severity of CAD . A causal relationship was later demonstrated in murine experiments whereby the EAT M1/M2 macrophage ratio was increased in high-fat-fed mice compared to controls, further indicating a relationship with obesity and adipose tissue inflammation .
EAT/cardiac dysfunction in metabolic disease
Obesity, diabetes, and EAT
The EAT is considered a transducer of local and systemic inflammation in obesity, and this is exemplified as adiposity exacerbates conditions of chronic inflammation including rheumatoid arthritis, psoriasis and multiple sclerosis [164–166]. This connection is bidirectional, as obesity, rheumatoid arthritis, psoriasis, and multiple sclerosis are associated with accelerated CAD and/or an increased risk of myocardial infarction [167–169]. Interestingly, EAT thickness and volume is also increased in these patients, as well as in patients with other chronic inflammatory disorders, including inflammatory bowel disease [170–172]. The pathological shift in obesity and chronic inflammation elucidates the intimacy between these two phenotypes.
Patients with T2DM are likely to have concomitant obesity and enhanced visceral adiposity with a marked increase of EAT thickness, and are predisposed to develop HFpEF [173,174]. Although the association with enhanced EAT and systemic inflammation is likely the driving force, another risk factor may be the pharmacological treatment of T2DM. The side effect of weight gain accompanied with some anti-hyperglycemic drugs have been associated with increased HF in randomized clinical trials . Insulin, for example, increases the amount of epicardial fat and is associated with an enhanced risk of HF . In contrast, newer glucose-lowering medication such as dipeptidyl-peptidase-4 (DPP-4) inhibitors are also associated with increased risk of HFpEF despite a neutral effect to adiposity . Interestingly, DPP-4 inhibitors were reported to reduce EAT volume ; however they may potentiate EAT inflammation [174,177]. Another class of incretin-based antidiabetics includes the glucagon-like peptide-1 (GLP-1) receptor antagonists. Although these drugs reduce EAT volume rapidly and substantially  and promote weight loss, they did not ameliorate EAT dysfunction  thus no improvements were observed for the rate of HF-related adverse events with GLP-1 receptor antagonists [174,180,181].
HFpEF and EAT
Heart failure with preserved ejection fraction (HFpEF) is characterized by LV diastolic dysfunction with retained systolic function and represents greater than 50% of heart failure patients [182,183]. HFpEF is associated with considerable morbidity and mortality, and is increasing in prevalence despite recent advances in HF treatment . In fact, clinical trials for patients with HFpEF have failed to meet their primary endpoints, limiting the availability of efficacious therapeutic approaches for HFpEF . Impaired LV filling is a result of physical restriction from limited pericardial space or a result of fibrosis, which enhances LV stiffness thus limiting the capacity to relax and fill . Clinically, patients with HFpEF present signs and symptoms of HF with preserved LV ejection fractions (≥50%), reduced cardiac output and impaired longitudinal strain [187,188]. HFpEF is frequently associated with comorbidities, including obesity and T2DM [184,189]. Multiple lines of evidence suggest proinflammatory adipocytokines released from EAT may contribute to the comorbidities observed with HFpEF, especially with obesity and T2DM. In obese patients, the accumulation of EAT is accompanied by diastolic filling abnormalities, enhanced vascular stiffness and impaired coronary microcirculation [190–193]. Further, patients with concomitant obesity and HFpEF demonstrate an increase in EAT thickness with a shift in cytokine profile to a proinflammatory state. Indeed, this inflammatory phenotype and diastolic filling irregularities precedes clinical diagnosis of HFpEF . Therefore, as EAT expansion facilitates inflammation and fibrosis of the adjacent myocardium, this provides a mechanistic link between HFpEF related to obesity and T2DM particularly given the close connection with weight gain and accumulation of EAT in these conditions [14,147,174].
As discussed previously, PVAT houses several pathways that control vascular function and structure. The large body of evidence describing vascular contractile and endothelial dysfunction as well as medial and intimal remodeling and atherosclerosis secondary to PVAT inflammation in metabolic disorders makes these pathways a lucrative target for therapeutic interventions. Similarly, as EAT expansion and inflammation contributes to HF and metabolic disorders, targeting EAT volume and inflammation is a potential therapeutic strategy. Nevertheless, none of the available therapeutic tools thus far were developed to specifically target either PVAT or EAT. Drugs with an ameliorative effect on adipose inflammation in general work by a “pleiotropic” or an “off-target” effect by indirectly modulating these key fat depots. Subsequent sections will provide an overview of possible therapeutic interventions with a reported adipose anti-inflammatory effect in metabolic disorders, some of which with a consequent normalization of cardiovascular function. A few hypothetical targets are discussed based on recent studies of EAT, PVAT, as well as analogous pathways in other adipose depot inflammation and signaling.
Dietary and lifestyle interventions
Non-pharmacological measures involving changes of diet and lifestyle are emphasized in all guidelines, not only for management of diabetic and cardiovascular patients, but also for prevention of conversion from prediabetes to diabetes . Such interventions aimed at reducing caloric load and improvements of hyperinsulinemia and insulin resistance are expected to reverse the pathological process leading to adipose inflammation. Indeed, caloric restriction was associated with reduced macrophage infiltration in high-fat fed mice . As well, exercise reduced white  and brown  adipose tissue inflammation in the context of obesity. Increased PPARγ signaling leading to a reduced inflammatory macrophage polarization in adipose tissue was proposed as the underlying mechanism . Specifically, dietary intervention and exercise were shown to reduce PVAT inflammation [44,70,197] in obese and prediabetic animal models; an effect associated with improved vascular contractile and cardiac autonomic function. Interestingly, only dietary switching to a low-fat chow was effective, while a calorie-restricted intake of high-fat diet to a caloric value equivalent to the control chow had no ameliorative effect . This raises the question of whether it is the high calorie or high-fat content per se versus a loss of micronutrients upon diet refining that underlies the exacerbation of the PVAT inflammatory phenotype.
A meta-analysis of weight loss strategies including bariatric surgery, diet and physical exercise in adults identified EAT as a modifiable risk factor. Further, comparisons demonstrate that intervention type and resulting weight loss achievement predicts EAT reduction . Aerobic exercise training in obese men also demonstrated a reduction in visceral adiposity and EAT thickness . Although the inflammatory status of EAT upon weight loss has yet to be investigated, lifestyle modification-induced weight loss reduced EAT thickness and was associated with reduction in soluble CD40 ligand, a systemic inflammatory mediator . Further, weight loss by invasive interventions reduced the incidence of HF in obese patients compared with intensive lifestyle modifications alone . Although the evidence for lifestyle modifications appears promising, compliance is a concern when physicians prescribe diet and exercise as a treatment plan.
Interest in the vasculoprotective effects of anti-diabetic drugs peaked following the findings from recent cardiovascular outcome trials on newer drug classes. Both the EMPA-REG  and the LEADER  outcome provided the first clear evidence that certain anti-diabetic drug classes could provide a cardioprotective effect by a mechanism independent of blood glucose level reduction. Findings extended beyond the newer sodium-glucose co-transporter 2 (SGLT2) inhibitors and glucagon-like peptide 1 (GLP1) agonists to older drug classes. The IRIS  and the MET-REMODEL  trials showed clear cardioprotective benefits for pioglitazone and metformin respectively, in non-hyperglycemic patients with metabolic disorders. A vast expanse of evidence proposes an anti-inflammatory effect for metformin [207,208] and pioglitazone [209,210] on adipose inflammation in diabetes and obesity. Importantly, studies have shown that treatment with either drug ameliorated PVAT inflammation followed by improvement of atherosclerosis , vascular contractile function , and cardiac autonomic function  without altering blood glucose levels in metabolically unhealthy, yet non-hyperglycemic patients or animal models. Further investigation is required to uncover the exact mechanism of this anti-inflammatory effect. Indeed, different studies report varying molecular pathways underlying the anti-inflammatory effect of either drug. For instance, metformin is known to switch the cellular metabolism into an anaerobic state , and thus would drastically lower the cellular demand for oxygen in adipocytes  and alleviate hypoxia more effectively than pioglitazone, which in turn, has an insulin sensitizing effect reducing circulating insulin and adipocyte size [44,209]. While a similar extent of investigation for GLP1 agonists and SGLT2 inhibitors is not available as of yet, an adipose anti-inflammatory effect was suggested for both drug classes [214,215]. A study suggested that GLP1-based therapy reduced PVAT inflammation and improved atherosclerosis and endothelial function in diabetic mice . Similar results were also recently reported for SGLT2 inhibitors . More detailed studies will be required to guide the use of these drugs in diabetic patients with suspected adipose inflammation, in prediabetic or other non-hyperglycemic patients to curb vascular dysfunction, and at non-hypoglycemic doses to investigate adipose tissue status and vascular function. In the context of targeting adipose tissue inflammation in cardiometabolic syndromes, the pharmacological mechanism of action of anti-diabetics is essential to illustrate their ameliorative effects, of which we will highlight two classes.
Metformin, an insulin sensitizer, is the most prescribed monotherapy for the treatment of T2DM due to its safety profile, low cost, and cardiovascular benefits . The putative mechanism of action of metformin is through inhibition of complex I of the mitochondrial respiratory chain . The downstream effects include phosphorylation and activation of AMPK, thus increasing lipolysis, mobilizing triglycerides, and enhancing fatty acid oxidation. Further, metformin enhances insulin sensitivity by increasing GLUT-4 mRNA expression and protein content in the membrane, thus facilitating enhanced glucose uptake . Mitochondrial complex I has been reported to contribute to reactive oxygen species production , and metformin limits oxidative stress by inhibiting this complex. Therefore, the protective anti-inflammatory properties of metformin and its ability to reduce EAT quantity support its use in patients with HFpEF, even independent of T2DM [174,222]. In fact, metformin treatment both prevents the development of HF in spontaneously hypertensive, insulin resistant rats  as well as reduced HF events in randomized controlled trials [174,224].
Sodium–glucose cotransproters-2 (SGLT2) inhibitors
SGLT2 inhibitors are a unique class of antidiabetic drugs that preclude the reabsorption of glucose in the proximal tubule to enhance its urinary excretion and improve glycemic control. Due to the effect of SGLT2 inhibitors on reducing overall adiposity, they were suspected to also reduce EAT volume. In fact, EAT volume was significantly reduced with SGLT-2 inhibitor treatment in patients with T2DM both with and without concomitant obesity, and the reduction was significantly correlated with BMI changes. Further, these drugs improved the inflammatory status of the EAT [225,226]. SGLT2 inhibitors prevented cardiomyopathy in experimental murine models of T2DM as well as reduced cardiovascular events when SGLT-2 inhibitors were used as an adjunct to the standard of care in T2DM patients [227,228].
Statins are one of leading prescribed medications with a reported 2.8 million patient population aged 35–79 between 2007 and 2011 in Canada alone . Statins effectively manage dyslipidemia and reduce risk of subsequent cardiac events by inhibiting 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, the rate-limiting enzyme of cholesterol synthesis [230,231]. In context, statin therapy is efficacious to reduce EAT quantity and improve its proinflammatory status [232–236], which has been suggested to play a role in improving systemic inflammation . The effect on EAT ameliorates diastolic dysfunction and cardiac fibrosis [174,237–239]. Accordingly, statins are associated with a reduced risk of HF or a reduced risk of morbidity and mortality in patients with established HF [174,240]. Statin therapy in patients with HFpEF is associated with reduced mortality [241,242]; however, no benefit on clinical outcome is observed in patients with heart failure with reduced ejection fraction (HFrEF) .
Mineralocorticoid receptor antagonists (MRA)
Classically, MRAs are defined as potassium-sparing diuretic drugs which act as competitive antagonist of the mineralocorticoid (type 1 glucocorticoid) receptor, blocking the actions of aldosterone to upregulate potassium exchange pump synthesis in the distal tubule by altering gene expression . The non-epithelial effects of aldosterone result from the discovery of mineralocorticoid receptors in non-epithelial locations, such as the heart and adipocytes [245,246]. Aldosterone has been implicated to exacerbate cardiovascular inflammation and injury; therefore MRAs are indicated to treat hypertension and HF . In addition, aldosterone promotes adiopose tissue inflammation and the accumulation of EAT . In clinical trials, eplerenone improved outcomes in HF patients both with and without visceral obesity, although greater improvements were seen in patients with abdominal fat accumulation . Aldosterone inhibitors were effective to reduce risk of adverse cardiovascular endpoints in patients with HFpEF and mildly elevated natriuretic peptides , potentially because this is a clinical feature of obesity-related HFpEF with increased visceral adiposity and EAT .
The RAAS is notorious for promoting detrimental effects in HF and cardiovascular dysfunction. RAAS overactivation is characteristic of HF and angiotensin-II elevation is pathogenic as it promotes aberrant myocardial remodeling. Diabetic cardiovascular disease and obesity are also associated with RAAS activation either directly or indirectly by increased risk for the development of HF . Genetic ablation of ACE2 in mice induced insulin resistance, glucose intolerance, EAT inflammation, and macrophage polarization to a M1 phenotype in response to high-fat diet; a phenotype that was rescued with exogenous administration of angiotensin 1–7 . Administration of recombinant ACE2 to increase local angiotensin 1–7 in the myocardium also ameliorates the effect of angiotensin-II signaling mediated fibrosis . Therefore, potential therapeutic approaches targeting the RAAS system include administration of recombinant human ACE2, potentiating the activity of ACE2, or exogenous administration of angiotensin 1–7. In all, the ACE2/angiotensin 1–7/Mas receptor counter-regulatory axis is a potential therapeutic avenue for the treatment of HF.
As discussed previously, EAT resident macrophages adopt a proinflammatory state characterized by phenotypic conversion from M2 to M1: a phenotype that correlates with severity of cardiovascular dysfunction . In accordance, we speculate that targeting the proinflammatory status of macrophages may be beneficial in metabolic disease and HF.
Macrophage polarization as a therapeutic target in myocardial infarction has been investigated. Initially, 5-azacytidine was found to elicit cardioprotection by modulating macrophage phenotype and inhibiting fibrosis in a rat model of myocardial infarction . The mechanism of action of 5-azacytidine is to inhibit the inducible isoform of nitric oxide synthesis (iNOS): a potent activator of macrophages to the M1 proinflammatory phenotype [251,252]. Interferon regulatory factor-1 (IRF-1) was subsequently elucidated as the driving factor for 5-azacytidine-mediated cardioprotection and the phenotypic switch from M1 to M2 . Di Filippo et al. demonstrated that pretreatment with telmisartan reduced myocardial infarct size and improved left ventricular ejection fraction in Zucker diabetic fatty rats with metabolic syndrome. Telmisartan reprogrammed macrophage polarization in the heart by favoring an M2 cytokine and chemokine profile [252,253]. Although these studies focused on macrophage infiltration in myocardial infarction, the principle of macrophage phenotypic switch is applicable to other disease states in which macrophage polarization is implicated, such as EAT inflammation in cardiometabolic syndromes.
Attenuation of thromboinflammation
As discussed previously, metabolic disorders associated with adipose inflammation are intimately associated with an upregulation of components of the clotting cascade to the point that caloric restriction not only reduced markers of adipose inflammation in obese mice, but also decreased clotting factor production . As well, the potential for PAR activation to influence vascular function makes the interference with clotting factor activation and/or PAR signaling a possible therapeutic intervention to ameliorate the detrimental vascular outcome of the combined thrombo- and adipose inflammation. Different classes of direct oral anti-coagulant drugs are used to treat patients suffering from or at risk of thromboembolic disorders. These drugs prevent blood clotting through direct inhibition of thrombin activity (dabigatran) or generation by interfering with factor Xa (rivaroxaban). Thus, it is plausible to assume that these drugs would reduce endogenous PAR stimulation, and hence potentially preclude PAR-mediated deterioration in EC/VSMC phenotype and vascular function. To our knowledge, this potential effect has not been systematically examined. Apart from very few studies reporting an improved endothelium-dependent relaxation following Xa inhibitor treatment and describing a portal venous pressure lowering effect of rivaroxaban [255–257], the direct effect of these drugs on vascular structure and function and the underlying mechanism remain largely unknown.
Direct modulation of adipokine receptors
Based on findings of alteration of adipokine secretion from inflamed PVAT, a direct intervention with adipokine receptors might constitute a possible route for the amelioration of consequent vascular changes. Despite its relevant roles as an endocrine and paracrine agent in the regulation of vascular disease, adiponectin as a therapeutic agent has been understudied, in part due to its short half-life of 75 min . However, this may be bypassed by targeting adiponectin receptors AdipoR1 and AdipoR2. This was demonstrated to be effective in mice, where an orally administered adiponectin receptor agonist has recently been shown to improve insulin sensitivity and glucose tolerance . Several classes of adiponectin receptor agonists exist, including the peptidomimetic ADP355, a decapeptide containing non-natural amino acid residues that are able to bind to the active site of the adiponectin receptor . It has been shown to increase the phosphorylation of eNOS and AMPK and decrease AKT phosphorylation when administered intraperitoneally every other day for 14 days ; however, ADP355 was not directly investigated in terms of its impact on vascular function. However, adiporon, a non-peptide adiponectin receptor agonist, has been shown to inhibit VSMC proliferation, DNA synthesis, and cyclin D1 expression at an IC50 of 25-50µM in vitro. Although adiporon was shown to activate AMPK in these cells, its anti-proliferative effects were sustained after siRNA downregulation of AMPK and were shown to be mediated by decreasing mTOR/p70S6K signaling downstream of platelet-derived growth factor, a potent mitogen released at the site of arterial injury. Furthermore, in vivo work on mice showed that oral administration of adiporon also reduced arterial injury-induced neointima formation by 57% . Natural flavonol-type compounds that exhibit adiponectin-like activities have been investigated and were found to improve glucose uptake and fatty acid oxidation in vitro at more favorable concentrations than adiporon, but their effect on vascular function has not been investigated . These effects encourage further examination in the context of vascular impairment associated with PVAT dysfunction.
As expected, administering leptin to obese patients has been proven ineffective in decreasing body weight , and despite its favorable glucose- and lipid-lowering profile compared with insulin in models of diabetes [265,266], its routine clinical use has been limited by its potential hypertensive effects . Thus emerged the potential of the leptin receptor as a therapeutic target, and leptin-sensitizing molecules, leptin receptor agonists, and molecules that increase leptin receptor density were proposed as possible interventions, many of which have shown benefits in animal models in terms of weight loss and inhibition of food intake . However, these therapies are still in their early stages of development, and their effects on vascular health, which is likely to be more obscure due to the multiple opposing effects of leptin on the vasculature, have not been investigated. On the other hand, the consistently detrimental effects of chemerin on the vasculature makes it a more appealing therapeutic target than leptin, and the chemerin antagonist CCX832 has been shown to inhibit phenylephrine-induced contraction in rat mesenteric arteries . However, there is still much to be done before it can be tested in clinical trials .
Owing to the vascular detrimental effect of PVAT senescence, anti-senescent therapy might constitute a viable option for intervention with vascular complications of metabolic disorders. Senescent cell targeting therapy has been discussed for its potential use in preventing diabetes and its complications. This may be accomplished by preventing or reversing the senescent phenotype, inhibiting its proinflammatory secretasome, or by eliminating senescent cells . In fact, treatment with senolytics has been shown to improve insulin sensitivity and decrease adipose tissue macrophage burden , leading to the possibility of introducing them for use in obese, prediabetic, or diabetic patients. However, their therapeutic benefit has not been investigated specifically on PVAT inflammation and the associated vascular dysfunction.
Possible targeting of UCP1
Induction of “adipose tissue browning” via increased UCP1 expression is generally viewed as a tool to up-regulate energy expenditure and offer an additional route of energy assimilation that might be of value in diabetes and obesity . However, many of the tools shown to increase adipocyte glucose consumption and increased UCP1 expression, in vitro, failed to produce any effect when used in vivo, and even resulted in an opposite effect of decreased UCP1 expression, particularly PPARγ activators [271,272]. Moreover, all of the studies and trials examining the effect of UCP1 modulation, so far, did so in a more global whole-body framework and at an advanced stage of metabolic dysfunction. Other than our recent work , the role of UCP1 as a possible instigator of PVAT dysfunction has neither been examined nor questioned. In this regard, the increased expression of UCP1 in PVAT in response to metabolic challenge, with the subsequent increase in oxygen consumption and augmented hypoxia compared to other adipose pools, is the main trigger for its early selective involvement. However, the use of UCP1 inhibitors as potential therapies, targeted to PVAT in early stages of metabolic deterioration, is challenged by the fact that UCP1 expression is not restricted to PVAT. Furthermore, the recently available inhibitor possesses a fairly high IC50 value (∼20 μM)  precluding systemic administration unless considerably high doses were used, potentially associated with significant systemic exposure. As such, future investigation of innovative approaches for a PVAT-selective or otherwise measured intervention with UCP1 activity is warranted.
The anatomic proximity of the EAT and PVAT allow endocrine and paracrine modulation of the adjacent myocardial and vascular tissues. The unique properties of either depot provide a valuable opportunity for therapeutic intervention in different stages of metabolic deterioration. Potential therapeutic approaches targeting EAT and PVAT may have disease-modifying value in HF and vascular complications associated with obesity and T2DM. Extensive future research is needed to explore the impact of EAT/PVAT-selective intervention with a number of pathways including thromboinflammation, adipokine signaling, and cellular senescence.
The authors declare that there are no competing interests associated with the manuscript.
This work was supported by an AUB-FM MPP [grant number 320148 (to A.F.E.)]; and CIHR and HSF funding (to G.Y.O.).
These authors contributed equally to this work.