- Split View
-
Views
-
Cite
Cite
Gemma Vilahur, Soumaya Ben-Aicha, Lina Badimon, New insights into the role of adipose tissue in thrombosis, Cardiovascular Research, Volume 113, Issue 9, July 2017, Pages 1046–1054, https://doi.org/10.1093/cvr/cvx086
- Share Icon Share
Abstract
Central obesity is independently associated with an elevated risk of cardiovascular disease, particularly thrombotic complications. Increasing data supports a link between excess body weight and the risk to suffer acute myocardial infarction, stent thrombosis after percutaneous interventions, ischemic stroke and vein thrombosis. Experimental and in vitro data have provided insights as to the mechanisms currently presumed to increase the thrombotic risk in obese subjects. Obesity is characterized by a chronic low grade inflammation and systemic oxidative stress that eventually damage the endothelium losing its antithrombotic properties. Obesity also stimulates the expression of leptin and attenuates adiponectin release, a protective adipokine. Although the contribution of adipokines to thrombosis has been questioned, recent work has suggested that they enhance platelet activation and, although to a lesser extent, induce the coagulation cascade through tissue factor (TF) expression. Increased body weight also impairs platelet sensitivity to insulin signaling and enhances the production of bioactive isoprostanes further promoting platelet reactivity. Finally, obese subjects have shown elevated circulating levels of von Willebrand factor, TF, factor VII and VIII, and fibrinogen, favoring a mild-to-moderate hypercoagulable state, and, on the other hand, increased secretion of plasminogen activator inhibitor (PAI)-1 and thrombin activatable fibrinolysis inhibitor (TAFI) contributing to impair the fibrinolytic system. In the present review, we provide an overview of the impact of excess body weight on thrombosis. We will focus on the link between dysfunctional adipose tissue and endothelial damage, platelet reactivity, enhanced coagulation and impaired fibrinolysis; mechanisms currently recognized to increase arterial thrombotic risk in obese subjects.
1. Adipose tissue and the risk of thrombosis
Abdominal obesity, one of the most prevalent metabolic disorders of the 21st century, increases the likelihood of various diseases such as cardiovascular disease (CVD), type 2 diabetes mellitus, musculoskeletal disorders and cancer; remaining CVD the main cause of obesity-related morbidity and mortality.1 Multiple clinical studies have demonstrated that obesity enhances the risk to suffer thrombotic complications including acute myocardial infarction (AMI), stroke, vein thrombosis and, although not that clear, peripheral artery disease. The increased tendency to form coronary thrombi is reflected by a 1.5–2.5-fold elevated risk for arterial thrombosis (suffer AMI) in obesity as compared to lean subjects.2 In addition, obese patients have a higher risk for long-term cardiovascular thrombotic events following percutaneous coronary intervention (PCI) with placement of drug-eluting stents than patients with normal weight.3
Body mass index (BMI; Table 1) is the most widely used measure to diagnose overweight and obesity although does not accurately assess fat tissue. Epidemiological studies have found a clear association between BMI > 29 and the risk to suffer stroke in both men and women. The mechanisms by which obesity may cause stroke are very similar to coronary heart disease (CHD) and include the prevalence of atherosclerosis, increased thrombotic risk, and high blood pressure. Multiple studies have shown an association between obesity and the risk of venous thrombosis (deep vein thrombosis and pulmonary embolism).4 The fact that obese patients have chronically raised intra-abdominal pressure and decreased blood velocity in the common femoral vein may further contribute to increase the risk by limiting the venous return.5 Finally, only few and contradictory data exists concerning the impact of obesity on peripheral artery disease.6
Weigh category . | Body mass index . | |
---|---|---|
Children . | Adults (kg/m2) . | |
Underweight | Below 5th percentile | Below 18.5 |
Healthy weight | 5th percentile to less than 85th percentile | 18.5–24.9 |
Overweigh | 85th percentile to less than 85th percentile | 25–29.9 |
Obese | 95th percentile or above | 30 or above |
Weigh category . | Body mass index . | |
---|---|---|
Children . | Adults (kg/m2) . | |
Underweight | Below 5th percentile | Below 18.5 |
Healthy weight | 5th percentile to less than 85th percentile | 18.5–24.9 |
Overweigh | 85th percentile to less than 85th percentile | 25–29.9 |
Obese | 95th percentile or above | 30 or above |
Weigh category . | Body mass index . | |
---|---|---|
Children . | Adults (kg/m2) . | |
Underweight | Below 5th percentile | Below 18.5 |
Healthy weight | 5th percentile to less than 85th percentile | 18.5–24.9 |
Overweigh | 85th percentile to less than 85th percentile | 25–29.9 |
Obese | 95th percentile or above | 30 or above |
Weigh category . | Body mass index . | |
---|---|---|
Children . | Adults (kg/m2) . | |
Underweight | Below 5th percentile | Below 18.5 |
Healthy weight | 5th percentile to less than 85th percentile | 18.5–24.9 |
Overweigh | 85th percentile to less than 85th percentile | 25–29.9 |
Obese | 95th percentile or above | 30 or above |
. | Endothelial dysfunction . | Platelet activation . |
---|---|---|
Adiponectin | ||
Adipsin | ||
Angiotensinogen | ||
Apelin | ||
Calprotectin | ||
Cathepsins S, L, K | ||
CCL 3, 5, 7, 8, 11 | ||
Clusterin | ||
CRP | ||
Fetuin A | ||
Ghrelin | ||
IGF-I | ||
IL-1β | ||
IL-6 | ||
IL-8 | ||
IL-10 | ||
Leptin | ||
Lipocalin-2 | ||
MCP-1 | ||
Omenti | ||
Osteopontin | ||
PAI-1 | ||
PGI2 | ||
Resistin | ||
SAA | ||
Tissue factor | ||
TNFα | ||
VEGF | ||
Visfatin |
. | Endothelial dysfunction . | Platelet activation . |
---|---|---|
Adiponectin | ||
Adipsin | ||
Angiotensinogen | ||
Apelin | ||
Calprotectin | ||
Cathepsins S, L, K | ||
CCL 3, 5, 7, 8, 11 | ||
Clusterin | ||
CRP | ||
Fetuin A | ||
Ghrelin | ||
IGF-I | ||
IL-1β | ||
IL-6 | ||
IL-8 | ||
IL-10 | ||
Leptin | ||
Lipocalin-2 | ||
MCP-1 | ||
Omenti | ||
Osteopontin | ||
PAI-1 | ||
PGI2 | ||
Resistin | ||
SAA | ||
Tissue factor | ||
TNFα | ||
VEGF | ||
Visfatin |
: induces endothelial dysfunction or platelet activation; , prevents endothelial dysfunction and/or platelet activation. CCL, chemokines; CRP, C reactive protein; IL, interleukin; MCP-1, monocyte chemoattractant protein-1; PAI, plasminogen activator inhibitor; PGI2, prostacyclin; SAA, serum amyloid A; TNF, tumor necrosis factor; VEGF: vascular endothelial growth factor.
. | Endothelial dysfunction . | Platelet activation . |
---|---|---|
Adiponectin | ||
Adipsin | ||
Angiotensinogen | ||
Apelin | ||
Calprotectin | ||
Cathepsins S, L, K | ||
CCL 3, 5, 7, 8, 11 | ||
Clusterin | ||
CRP | ||
Fetuin A | ||
Ghrelin | ||
IGF-I | ||
IL-1β | ||
IL-6 | ||
IL-8 | ||
IL-10 | ||
Leptin | ||
Lipocalin-2 | ||
MCP-1 | ||
Omenti | ||
Osteopontin | ||
PAI-1 | ||
PGI2 | ||
Resistin | ||
SAA | ||
Tissue factor | ||
TNFα | ||
VEGF | ||
Visfatin |
. | Endothelial dysfunction . | Platelet activation . |
---|---|---|
Adiponectin | ||
Adipsin | ||
Angiotensinogen | ||
Apelin | ||
Calprotectin | ||
Cathepsins S, L, K | ||
CCL 3, 5, 7, 8, 11 | ||
Clusterin | ||
CRP | ||
Fetuin A | ||
Ghrelin | ||
IGF-I | ||
IL-1β | ||
IL-6 | ||
IL-8 | ||
IL-10 | ||
Leptin | ||
Lipocalin-2 | ||
MCP-1 | ||
Omenti | ||
Osteopontin | ||
PAI-1 | ||
PGI2 | ||
Resistin | ||
SAA | ||
Tissue factor | ||
TNFα | ||
VEGF | ||
Visfatin |
: induces endothelial dysfunction or platelet activation; , prevents endothelial dysfunction and/or platelet activation. CCL, chemokines; CRP, C reactive protein; IL, interleukin; MCP-1, monocyte chemoattractant protein-1; PAI, plasminogen activator inhibitor; PGI2, prostacyclin; SAA, serum amyloid A; TNF, tumor necrosis factor; VEGF: vascular endothelial growth factor.
Obesity is strongly associated with the traditional cardiovascular risk factors clustered in the metabolic syndrome (abnormalities in glucose metabolism, atherogenic dyslipidemia, and arterial hypertension).7 Although the metabolic syndrome is characterized by a highly prothrombotic state, obesity per se has also shown to increase thrombotic risk. Indeed, there is an important link between obesity (which produces a low-grade chronic inflammatory state) and an enhanced platelet response, mild-to-moderate hypercoagulability,” and reduced fibrinolysis (Figure 1).8
In the present review, we provide an overview of the impact of excess body weight on arterial thrombotic risk. We will focus on how adipose tissue damages the endothelial layer, increases platelet reactivity, enhances coagulation and impairs fibrinolysis; mechanisms currently presumed to increase the thrombotic risk in obese subjects.
The strategy followed to write this review has based on the Pubmed search and combination of the following terms: adipose tissue, obesity, inflammation and inflammatory/immune cells, thrombosis, platelet, coagulation, fibrinolysis, adipokines, endotoxemia, endothelium, and oxidative stress.
2. Adipose tissue and the vasculature
2.1 Impact of obesity on the endothelial function
The healthy endothelium synthesizes and releases molecules that modulate the coagulation cascade, platelet aggregation, and fibrinolysis. The anticoagulant properties of the endothelium are conferred by the expression of two tansmembrane proteins, thrombomodulin and heparin sulphate proteglycan, and the release of tissue factor pathway inhibitor (TFPi). Thrombomodulin serves as a binding site for thrombin to activate protein C eventually disabling FVIII and FV activation (thrombin formation); heparan-like molecules act as cofactors for antithrombin III; and TFPi inhibits both factor Xa and tissue-factor-factor VIIa complex. Endothelial cells also produce tissue-type plasminogen activator (tPA) which activates the fibrinolytic system. As per the antiplatelet properties of the endothelium, these are mainly regulated by the surface exposure of ecto-adenosine diphosphate (ADP)-ases CD39 and CD73 and the synthesis and release of prostacyclin (PGI2) and, most importantly, nitric oxide (NO). NO is a soluble gas continuously synthesized in endothelial cells from the amino acid L-arginine by the constitutively expressed endothelial nitric oxide synthase (eNOS). NO not only inhibits platelet aggregation but promotes arterial vasodilation in conjunction with endothelium-derived hyperpolarizing factor (EDHF).9 Kruppel-like factor 2, an endothelial transcriptional factor, has also shown to contribute to maintain the antithrombotic endothelial surface by strongly inducing thrombomodulin and eNOS expression.10 Yet, the continuous exposure to cardiovascular risk factors (hyperlipidemia, oxidative stress, and inflammation) activates the endothelial layer, enhances endothelin-1 production (vasoconstrictor and pro-inflammatory peptide) and, conversely, impairs NO- and EDHF- related vasodilation leading to endothelial dysfunction.11,12 Alterations in endothelial function have been shown to precede the development of atherosclerotic lesions and to have a prognostic value for future cardiovascular events in ACS patients.13 Obese patients display endothelial dysfunction14 likely induced by the continuous exposure to inflammatory stimuli and oxidative damage.15,16
On the one hand, the dysfunctional adipocyte triggers for inflammatory cell recruitment and secretion of inflammatory cytokines and adipokines largely contributing to the inflammatory state that characterizes central obesity. However, recent studies have also suggested a contributory role of microbes or microbial endotoxins such as lipopolysaccharide (LPS; metabolic endotoxemia) in initiating and perpetuating such chronic low grade inflammatory response.17 Indeed, LPS interaction with its receptor Toll-Like Receptor 4 (TLR4) may initiate a signaling cascade that elicit many pro-inflammatory pathways.18 On the other hand, there is an enhanced production of reactive oxygen species (ROS) by uncoupled eNOS, nicotinamide adenine dinucleotide phosphate (NADPH) oxidases, xanthine oxidase, lipoxygenase, cyclooxygenase, microsomal P-450 enzymes, and pro-oxidant heme molecules.19 Moreover, several components of the rennin–angiotensin–aldosterone system (RAAS), particularly of angiotensinogen, are produced by the dysfunctional adipocyte leading to an over-activation of the RAAS system with the consequent increase in ROS production.20 Concomitantly, endothelial dysfunction is accompanied by a decrease in the anti-oxidant defense mechanisms, particularly the nuclear factor-E2-related factor 2 (Nrf2)/Kelch-like ECH-associated protein 1 (Keap1)-antioxidant response element (ARE) signaling pathway.21 Under physiological conditions, Nrf2 is sequestered in the cytoplasm by Keap1 which facilitates its proteasomal degradation. Oxidative stress causes Nrf2 to dissociate from Keap1 and to subsequently translocate into the nucleus where it binds to ARE resulting in the transcription of antioxidant genes involved in vascular protection. However, Nrf2 has been found down-regulated in metabolic syndrome patients diminishing the anti-oxidative response.22
Despite all the adverse events derived from gaining weight at the vascular level, endothelial dysfunction is found to be reversed upon weight loss leading to lower adipose tissue-related inflammation and increasing NO availability.23
2.2 Cross-talk between the perivascular adipose tissue and the vascular wall
Adipose tissue is categorized into subcutaneous (SAT), visceral (VAT) and perivascular adipose tissue (PVAT). White PVAT accumulates around the blood vessels and in addition to provide structural support and vaso-protection also influences the homeostasis and contractile state of the vessel wall by releasing numerous cytokines and adipokines (i.e. adiponectin). However, whereas there is a protective cross-talk between the PVAT and the vascular wall under healthy conditions (PVAT dampens oxidative stress and restores vascular redox balance), a disproportionate increase in PVAT mass, such as occurs in obesity, induces PVAT dysfunction increasing the release of inflammatory adipokines while reduces adiponectin contributing to vascular inflammation and dysfunction.24 Supporting this hypothesis, enhanced PVAT mass has been associated with increased presence of atherosclerotic plaques.25 Of note, whereas obesity expands white PVAT, conversely reduces the expression of beige and brown PVAT which, besides exerting thermogenic and metabolic functions, have been shown to induce vascular protective effects.26
3. Adipose tissue and platelet function
Platelets are anuclear cell fragments released from megakaryocytes that play an essential role in primary hemostasis and in the pathophysiology of atherothrombosis (for in-depth review please refer to Badimon and Vilahur).27
3.1 Platelet hyperactivity
Obese subjects have ‘angrier’ platelets and higher rates of ischemic events. Increased platelet activation has been found in obese women free of cardiovascular risk factors.28 Particularly, android obesity has been associated with a four-fold higher rate of urinary excretion of 11-dehydro-TxB2 (the major metabolite of TXA2) as compared to non-obese women.28 Such increase is comparable to that associated with cigarette smoking,29 hypercholesterolemia,30 and type 2 diabetes mellitus.31
It has been postulated that the chronic low-grade inflammatory state and oxidative stress associated with abdominal obesity enhances lipid peroxidation with the consequent increase in 8-iso-prostaglandin F2α (iso-PGF2α) levels, bioactive isoprostanes produced from arachidonic acid, through a process of non-enzymatic free radical-catalyzed lipid peroxidation (Figure 2).32 8-iso-PGF2α not only is a marker of oxidative stress but also affects platelet function. As such, 8-iso-PGF2α has been suggested to amplify the response of human platelets to low concentrations of other agonists through a mechanism mainly involving the TX receptor (TP). Interestingly reduction of weight loss by around 10% is accompanied with more than a 50% reduction in TX biosynthesis and a decline in 8-iso-PGF2α.
Epidemiological evidence has shown that platelet activation markers such as CD40L and P-selectin are found increased in obesity likely reflecting persistent in vivo platelet and endothelial activation (Figure 2).33 CD40L is a trimeric protein structurally related to the tumor necrosis factor alpha (TNF-α) superfamily and is stored in the α-granules on resting platelets. Upon platelet activation, CD40L is rapidly exposed in the platelet surface where it is cleaved to form a soluble fragment, sCD40L. sCD40L has shown, through autocrine, paracrine, and endocrine actions, to enhance platelet recruitment and to induce a vascular inflammatory response further participating in the atherothrombotic process.34 P-selectin is also stored in α-granules and also rapidly translocates to the platelet surface upon activation where it facilitates monocyte recruitment to the vessel wall.
The size and volume of circulating platelets, two parameters related with higher susceptibility of platelets to activation,35 have also been found increased in obese subjects and to correlated with a higher rate of MI.36 Large platelets contain denser granules and are, therefore, more active and have a higher prothrombotic response to aggregating agents than small platelets. On the other hand, larger mean platelet volume (MPV) has been described to predict a poor outcome following MI.37 Platelet volume is mainly determined during megakaryocyte fragmentation; hence, increased MPV may reflect higher circulating reticulated cells (young platelets) likely suggestive of an increased platelet turnover. Megakaryocyte maturation and release has been suggested to include the involvement of inflammatory mediators (e.g. interleukin (IL) -1, 3, 6, 8, 11, and 18), NO and thrombopoietin.38 Administration of IL-6 to humans results in an increased platelet count and IL-6 increase as observed in obesity has been shown to alter the morphology and reactivity of platelets released from the bone marrow.39 Hence, one could postulate that the reduced NO production and increased cytokine release found in obesity might influence megakaryopoiesis and consequent platelet volume. In addition, higher MPV has been associated with a lower effectiveness of prostacyclin (PGI2) to inhibit platelet activation.40
Platelet count, a parameter associated with adverse clinical outcomes on STEMI presentation, has also been found to be increased in obese women.41 We have demonstrated that obese non-diabetic rats show increased platelet count and an increased MPV which are associated with an increased thrombotic risk assessed in vivo by real time intravital microscopy.42 In fact, thrombotic risk was comparable to that observed in obese-diabetic rats.43 We have also demonstrated that platelet number, MPV, and thrombotic risk were directly correlated with weight and that a reduction of peripheral insulin resistance may contribute to reduce thrombotic risk in obesity.42 At the clinical arena, weight loss after bariatric surgery has been found to be associated with changes in the MPV,44 and a recent study has also demonstrated a parallel reduction in platelet count.45
3.2 Decreased response to insulin, PGI2 and NO: the role of insulin resistance
Functional insulin receptors are present in the platelet membrane with a density comparable to that of other insulin-sensitive cells, around 500 receptors per cell. Upon activation by insulin binding, the β-subunit is auto-phosphorylated thereby triggering the classic intracellular pathway of insulin signaling (without the classical response to glucose uptake). In healthy subjects, insulin has shown to exert antiaggregatory effect upon platelet challenge with multiple agonists and reduce platelet deposition over collagen-coated surfaces under flow conditions.46 Insulin has also shown to reduce the intracellular concentrations of calcium.47 The mechanism by which insulin reduces cytoplasmic calcium levels and platelet deposition involves the activation of eNOS and the enhanced production of PGI2.48 eNOS activation leads to NO synthesis which in turn induces the stimulation of guanylate cylcase (GC) and an increase in cyclic GC phosphate (cGMP). cGMP increases the concentration of cyclic adenosine mono-phosphate (cAMP) by inhibiting cAMP phosphodiesterase. cAMP is, in turn, activated by PGI2. Both cGMP and cAMP phosphorylate a series of protein kinases ultimately leading to a reduction in intracellular calcium levels preventing granule release and further platelet activation/aggregation (Figure 2).49
Platelets from obese individuals have shown to display a reduced sensitivity or impaired platelet response to insulin and consequently to PGI2 and NO, thereby increasing platelet susceptibility to activation (Figure 2).50 In line with these findings, obese individuals show an impaired ability to induce platelet cGMP and cAMP in response to NO donors or a synthetic PGI2 analogue (iloprost), respectively.50 Conversely, either the administration of the insulin sensitizer pioglitazone or weight reduction have been concomitantly associated with amelioration in insulin sensitivity and reduced platelet activation.49
It is though that the presence of insulin resistance (IR) largely developed in obese subjects blunts the physiological antiplatelet actions exerted by insulin leading to higher platelet reactivity. In fact, obesity is a major correlate of IR being certain patterns of fat tissue distribution, particularly central fat, more strongly related to impaired insulin signaling.51 It is thought that the increased production of pro-inflammatory mediators and increased release of free fatty acids by the dysfunctional adipose tissue induces IR through an impairment in insulin receptors and intracellular signaling (Figure 2).52
Insulin has also been proposed to reduce platelet sensitivity to multiple platelet agonists including ADP, collagen, thrombin, and epinephrine (Figure 2).53 Moreover, besides modulating platelet function, insulin has shown to regulate transcription factors involved in tissue factor (TF) and plasminogen activator inhibitor (PAI)-1 expression suggesting an involvement of insulin in thrombin formation and fibrinolysis in obese individuals.54
3.3 Increased CRP levels
The chronic low grade-inflammatory state that characterizes obese subjects is reflected by high levels of TNF-α and C-reactive protein (CRP). Visceral adipose tissue is infiltrated by macrophages which secrete inflammatory mediators, such as TNF-α and IL-6 (Figure 1). IL-6, in turn, drives the upregulation of CRP synthesis in the liver although inflamed adipose tissue has also reported to produce CRP.55 Accordingly, CRP not only is found increased in obese subjects but positively correlates with BMI.55 High-sensitivity CRP assessment has been robustly associated with an increased MI risk and a recent study has demonstrated that this association is also present in obese patients.56 Accumulating evidence support a pro-thrombotic activity for CRP. As such, CRP has been shown to enhance the expression of endothelial adhesion molecules, stimulate macrophages to synthesize cytokines, and induce TF expression in monocytes.57 Moreover, CRP has been reported to modify the fibrinolytic balance of endothelial cells and thus promote fibrin formation, to enhance the expression of PAI-1 in human endothelial cells via up-regulation of Rho kinase and inhibits tPA activity. We and others have demonstrated that CRP also contributes to thrombus progression and growth.58,59 Particularly, we have revealed that the classically studied serum CRP (natCRP; a pentamer formed by five non-covalently bound globular subunits) undergoes subunit dissociation in the platelet surface into an individual monomeric unit with pro-thrombotic potential.58 These monomeric subunits contribute to platelet activation, enhance platelet deposition, and increase thrombus growth under arterial flow conditions.58,60 So far, extremely obese patients undergoing bariatric surgery have shown a significant reduction in CRP levels upon weight loss likely reducing both the pro-inflammatory and pro-thrombotic risk.61
3.4 Increased platelet-derived microparticles
Platelet-derived microparticles (pMP) are small membrane vesicles with <0.1µm of diameter released from platelets either activated by shear stress or by agonists. Hence, pMP have been considered an index of platelet activation. PMP have shown to exert both pro-inflammatory and pro-thrombotic effects likely contributing to the progression of the athero-thrombotic response.62 In fact, about 25% of the procoagulant activity of stimulated platelet suspensions is associated with pMP released upon platelet activation and their surface may be ∼50–100-fold more procoagulant than the surface of activated platelets per se. We have demonstrated that pMPs enhance thrombus formation under flow conditions and also stimulate platelet aggregation playing a functional role in thrombus progression and growth.63 Recent studies have reported that circulating levels of pMPs are elevated in obese subjects in comparison with age-matched non-obese subjects, with a positive correlation with BMI, waist circumference and fat tissue mass, likely reflecting a state of platelet activation.64 As other markers of platelet activation, pMP levels have been found to be reduced, in part, by a successful weight loss.64
3.5 The influence of adipokines
Adipokines are a group of bioactive molecules synthesized by the adipose tissue that act as signaling molecules.65,Table 2 includes adipokines that may induce thrombosis either by contributing to inflammation/endothelial dysfunction or by enhancing platelet activation.66–68 Among them, leptin and adiponectin have particularly shown to interfere with platelet function as discussed below.
3.5.1 Leptin
Leptin is a non-glycosilated protein hormone predominantly produced and secreted by the adipose tissue. Leptin interacts with the leptin receptor in the hypothalamus to suppress food intake and increase energy expenditure. The fact that most obese people have high circulating concentrations of leptin suggests that the hypothalamus becomes insensitive or ‘resistant’ to the effects of leptin. However whether the concept of leptin resistance also applies to all tissues remains to be determined. In this regard, leptin exerts multiple effects in the vascular system and accordingly, leptin receptors have been identified in many types of vascular cells, including endothelial cells, macrophages, and platelets (Figure 2).
Mice and human platelets express the long form of the leptin receptor (ob) and both leptin and leptin-receptor-deficient mice have been reported to be protected from experimental thrombosis.69 However, whereas many animal model studies have proved a causative pro-thrombotic effect of leptin, contradictory findings have been reported for human platelets. Leptin is unable to aggregate human platelets per se, yet, when co-incubated at a concentration range that corresponds to that of plasma leptin levels in obese individuals it has shown to synergize with sub-threshold concentrations of ADP and thrombin to promote platelet aggregation, an effect not observed at lower concentrations found in normal individuals (Figure 2).70,71 Consistent with these observations clinical trials have found a strong association between plasma leptin levels and vascular thrombosis.72 Moreover, hyperleptinemia has been shown to carry an increased risk for thrombotic events such as ischemic stroke and AMI, independent of other risk factors.73 The mechanisms by which leptin may induce platelet activation are the following: (i) activates Janus Kinase 2 (JAK2) which through the PI3-K/Akt signaling pathway activates PDE3A and consequently hydrolyses cAMP or (ii) through PLCγ2, PKC, and PLA2 induces GPIIb/IIIa activation, Ca2+ increase and TX synthesis. On the other hand, other studies have argued that the effect of leptin on ADP-induced platelet aggregation is found to be attenuated in obesity, a condition where leptin receptors are found to be desensitized.74
3.5.2 Adiponectin
In contrast to the other adipokines, adiponectin is an anti-inflammatory adipokine and contributes to increase insulin sensitivity protecting, therefore, against diabetes, atherosclerosis and thrombosis (Figure 2). Accordingly, high concentrations of adiponectin have been associated with a reduction in the risk of coronary artery disease (CAD) and an increase in endothelial NO production. Conversely, patients with CAD, increased carotid intima–media thickness and obesity exhibit low plasma adiponectin levels.75 Regarding the impact of adiponectin on platelet function, both mice and human platelets express adiponectin receptors AdipoR1 and AdipoR2. Yet, whereas adiponectin deficient mice have shown enhanced thrombus formation after photochemically induced arterial injury and increased agonist-induced platelet aggregation,76 the addition of adiponectin to human platelets has shown to exert no effects on platelet activation/aggregation.67 Adiponectin has also shown to inhibit macrophage-related TF expression and activity likely affecting the coagulation cascade.77
4. Adipose tissue effects in the coagulation cascade and fibrinolysis
Systemic inflammation, endothelial dysfunction, disturbances of lipid and glucose metabolism, and IR contribute to the hypercoagulable state and the impaired fibrinolysis found in obesity.78 However, obesity is characterized by the elevation of several clotting factors and PAI-1 directly affecting coagulation and fibrinolysis independent of genetic factors (Figure 1).79 Moreover, a recent study has reported that subcutaneous adipose tissue shows a stronger relationship with functional measures of hypercoagulability as compared to visceral adipose tissue, suggesting that the anatomic location of adipose deposition may influence the type of thrombotic event, with visceral adipose tissue being associated with arterial thrombosis whereas subcutaneous adipose tissue predisposing to venous thrombosis.80
4.1 Adipose tissue and alterations in the coagulation factors
4.1.1 Tissue factor (TF)
TF, a 47 kDa transmembrane glycoprotein mainly found in atherosclerotic plaque VSMCs and lipid-loaded macrophages, is considered to be the initial trigger of thrombus formation upon atherosclerotic plaque rupture.27 Recent studies have also suggested the contribution of neutrophil extracellular traps (NETs)-bound TF in the progression of coronary thrombosis.81 TF activates the extrinsic coagulation cascade by serving as the cell surface receptor for coagulation factor VIIa (FVIIa). Activation of the TF/FVIIa complex activates FX with the consequent FXa-mediated generation of thrombin and eventual conversion of fibrinogen to fibrin.
Adipose tissue expansion during obesity induces the recruitment of pro-inflammatory cells (macrophages, neutrophils, CD8 T cells, B cells, mast cells, and INF-γ-Th1) which perpetuate the secretion of pro-inflammatory adipokines favoring the chronic low-grade inflammatory state and endothelial dysfunction. Yet, macrophages, which have suffered a polarization towards a pro-inflammatory phenotype, are also capable to express and release TF favoring a pro-coagulable state.68 Indeed, plasma levels of TF have been found increased in obese subjects as compared to healthy controls.82 Moreover, studies in obese mice have demonstrated a significant up-regulation in TF transcript levels in adipocytes and stromal vascular cells.83 The elevated TF expression observed with increased body weight may also occur in response to several factors, including insulin, TNF-α, transforming growth factor-β and leptin. In this latter regard, leptin has been found to promote the generation of active TF in human neutrophils and peripheral blood mononuclear cells83 as well as in coronary endothelial cells.84 Moreover, a recent observational study by Petrini et al., has demonstrated the ability of leptin to induce the shedding of pro-coagulant TF-bearing MP from human peripheral blood mononuclear cells.85
4.1.2 Molecules involved in clot formation
Levels of fibrinogen, von Willebrand factor, and factors VII and VIII have been found increased in obesity (Figure 1) likely because of the stimulating effect of inflammatory mediators on hepatocytes and endothelial cells.86 A clinical observational study reported that patients with type 2 diabetes and abdominal fat displayed higher plasma activity of clotting factors VII and VIII as well as increased plasma levels of fibrinogen and von Willebrand factor antigen as compared to type 2 diabetic patients with healthy body weight.87 However, whether elevated levels of coagulation factors/elements directly contribute to thrombosis or are merely biomarkers of inflammation remains uncertain.
Obesity has also been associated with higher levels of fibrin degradation products (FDP).88 Increased plasma cytokines are known to induce hepatic fibrinogen production, which further exacerbates inflammation, especially following fibrinolytic cleavage into FDP-D and FDP-E. Yet, interestingly, fibrinogen has shown to predict weight gain in middle-aged adults reinforcing the positive feedback loop between inflammation and the development of obesity and/or metabolic syndrome.89
4.2 Adipose tissue-related disturbances in the fibrinolytic system
4.2.1 PAI-1
The conversion of fibrinogen into fibrin and the development of an insoluble fibrin clot are the final steps of the coagulation cascade. PAI-1, a member of the serpin superfamily (serine protease inhibitors), is mainly secreted from the liver and the adipose tissue and blocks the activation of fibrinolysis. As such, PAI-1 inhibits plasminogen activators (PAs) including t-PA and urokinase-type plasminogen activator (u-PA) thereby limiting the dissolution of the fibrin clot (Figure 2). Rodent studies have revealed that obesity is associated with a marked upregulation of PAI-1 expression in visceral adipose tissue90 and higher PAI-1 plasma levels.91–93 Moreover, Nagai et al. 91 demonstrated, in an mice model of photochemically induced thrombus formation, that obese mice with high plasma PAI-1 levels displayed shorter times to achieve complete artery occlusion as compared to their lean counterparts with lower PAI-1 levels. These observations indicate that higher PAI-1 levels may impair fibrin clearance thereby enhancing the risk of thrombosis. PAI-1 levels have also been found to be increased in obese patients.94 The existing correlation between PAI-1 (and in general with the fibrinolysis) and body weight is likely explained through the increase levels in inflammatory cytokines found in obesity.78 In counterpart, extremely obese patients undergoing Roux-en-Y gastric bypass have shown a clear reduction in PAI-1 levels which nicely correlated with changes in leptin concentrations and with BMI.61 Interestingly, in vitro and in vivo studies have also demonstrated that besides its role in atherothrombosis, PAI-1 is implicated in adipose tissue development and in the control of insulin signaling in adipocytes. Nevertheless, uncertainties remain whether the observed associations between elevated PAI-1 levels and pro-thrombotic, metabolic, and inflammatory parameters are correlative or causative.
5.2.2 TAFI
Thrombin not only contributes to clot formation but also stabilizes the clot by activating thrombin activatable fibrinolysis inhibitor (TAFI). Once activated TAFI protects the fibrin clot against lysis. Thus, defects in TAFI activation might contribute to the severity of bleeding disorders and, conversely, increased TAFI activation due to an increased rate of thrombin generation might lead to thrombotic disorders. In fact, TAFI overexpression has been considered a potential biomarker for the risk prediction of ACS.95 So far, there are no animal model studies or clinical trials supporting a correlation between adipose tissue and TAFI, yet, one study has suggested a role for TAFI in the association between obesity and thrombosis.96 Further studies are needed to determine both the true clinical impact of this fibrinolytic inhibitor and its link with adipose tissue depots.
5. Conclusions
The involvement of adipose tissue to increase the thrombotic tendency has been proposed through several direct and indirect mechanisms involving endothelial dysfunction together with platelet function abnormalities, enhanced coagulation, and hypo-fribrinolysis. Direct effects include the synthesis and secretion of TF and PAI-1 by adipose tissue, whereas the indirect effects are mainly mediated through the release of adipokines/inflammatory mediators and the oxidative environment which enhance platelet activity, induce liver production of coagulation factors and changes the vessel wall overall contributing to the thrombotic propensity found in obesity. Although many pro-thrombotic factors associated with obesity have been shown to improve with weight reduction, it remains de be determined the effect that the degree of weight loss following different interventions has on the alterations found in platelets and the coagulation and fibrinolytic systems. In addition, a better understanding of the pathways and components linking adipose tissue (visceral and ectopic) with circulating and vascular components involved in thrombus formation is required in order to identify those obese individuals at particularly high thrombotic risk and develop new targets for intervention.97
Acknowledgements
This work was supported by PNS 2015-71653-R (to G.V.) PNS SAF2016-76819-R (to L.B.) from the Spanish Ministry of Science and Innovation and FEDER Una Manera de Hacer Europa funds; from Instituto de Salud Carlos III CIBERCV (CB16/11/00411 to L.B.) and Red TerCel (RD16/0011/018), and a grant from Fundació la Marató TV3 (to G.V.) and from the Spanish Society of Cardiology (Beca FEC Investigación Básica/2016 to G.V.). We thank the continuous support of the Generalitat of Catalunya (Secretaria d’Universitats i Recerca del Departament d’Economia i Coneixement de la Generalitat; 2014SGR1303) and the Fundacion Investigación Cardiovascular.
Conflict of interest: none declared.
References
Author notes
This article is part of the Spotlight Issue on Dysfunctional Adipocyte and Cardiovascular Disease.