Plasminogen activator inhibitor-1, inflammation, obesity, insulin resistance and vascular risk
Irène Juhan-Vague, Laboratory Hematology, CHU Timone, 13385 Marseille cedex 5, France.
Tel.: +33 491492449; fax: +33 491942332; e-mail: email@example.com
Summary. Elevated plasma plasminogen activator inhibitor-1 (PAI-1) level is a core feature of insulin-resistance syndrome (IRS). Atherothrombotic complications in IRS are partly attributed to impaired fibrinolysis caused by increased plasma PAI-1 levels. Although the etiology of IRS is far from being explained, the clustering of inflammation, adipose tissue accumulation and insulin resistance suggests an etiopathological link. Proinflammatory cytokines might regulate PAI-1 expression in IRS; however, more studies are needed to confirm this complex mechanism in humans. Furthermore, modifying PAI-1 expression by PAI-1 inhibitors provides a new challenge and may reveal the true role of PAI-1 in atherosclerotic and insulin resistance processes.
Plasminogen activator inhibitor-1 (PAI-1) is the primary inhibitor of both tissue-type plasminogen activator (t-PA) and urokinase type plasminogen activator (u-PA) and thus limits the fibrinolytic process. Elevated plasma PAI-1 level is a core feature of insulin-resistance syndrome (IRS), also called metabolic syndrome or syndrome X . This syndrome is a major public health problem in western industrialized countries, as it represents a strong risk factor for vascular disease [2,3]. Atherothrombotic complications in IRS are partly attributed to impaired fibrinolysis caused by increased PAI-1 plasma levels .
In this article, potential mechanisms of increased PAI-1 synthesis in IRS and their relation with inflammation are discussed. Furthermore, data that have tried to delineate the role of PAI-1 in the development of atherosclerosis and its complications, and newer findings on a potential role of PAI-1 in the development of obesity and IRS, are reported.
IRS, inflammation and vascular risk
IRS is defined by a cluster of abnormalities, which includes obesity with a repartition of the fat in the central part of the body (visceral or android obesity), hypertension, glucose intolerance, hyperinsulinemia, dyslipidemia with elevated triglyceride, lowered high density lipoprotein (HDL) cholesterol concentration, and an increased proportion of small dense lipoproteins. Hemostatic abnormalities are also associated with IRS, and among these, increased PAI-1 expression has been extensively studied.
IRS has a large worldwide spread. Its prevalence in the USA increases from 7% for people aged 20–29 years, to 43% for people aged 60–69 years . Moreover, the prevalence of obesity has increased by 20% in the last 10 years in most western countries. People with IRS are at increased risk for developing diabetes mellitus and cardiovascular disease. The presence of IRS increases the cardiovascular mortality 6-fold . Fasting serum insulin level, a surrogate measure of insulin resistance, has been reported to be an independent risk factor for coronary heart disease .
Much effort has been made on revealing the mechanisms responsible for the development of insulin resistance that leads to the development of IRS. Low insulin sensitivity may have multiple origins. Genetic molecular defects in insulin-signalling pathways, and energetic imbalance due to increased energy-dense food consumption and reduced daily physical expenditure, are important factors involved in etiopathogenesis of IRS. Healthy adipose tissue is an efficient buffer against the daily flux of free fatty acids in the circulation ; in IRS, buffer function is impaired, accommodation of an increased energy influx is impeded, and expansion of subcutaneous adipose tissue is prevented. The prevention of adipocyte hypertrophy leads to deviation of lipid flux through other organs such as visceral fat, liver, skeletal muscle, pancreas and possibly vascular walls [8,9]. When lipids accumulate in such tissues, fatty acids enter deleterious pathways that cause toxicity and apoptosis .
Low-grade inflammation with prolonged cytokine-mediated acute-phase reaction is considered to be strongly related to IRS . Population studies have shown a strong association between indices of inflammation and abnormal lipid and carbohydrate metabolism, obesity and insulin resistance [11,12]. Indeed, there is evidence that obesity and insulin resistance may create an inflammatory state. Adipose tissue is an important source of interleukin (IL)-6 , and tumour necrosis factor (TNF) has been shown to be overexpressed in the adipose tissue of obese insulin-resistant patients . Interestingly, it was recently shown that thiazolidinedione, an insulin sensitizer compound that exerts its main action through activation of Peroxysome Proliferative Activator Receptor (PPAR)γ, has anti-inflammatory properties . The magnitude of the decrease in C-Reactive Protein (CRP) with thiazolidinedione was greater than that observed with statins in various trials . Thus, if insulin resistance triggers inflammation, it could be considered as a common precursor of both atherosclerosis  and diabetes .
In addition, several reports provide evidence that inflammation could initiate insulin insensitivity. IL-6 has been shown to induce cellular insulin resistance in hepatocytes . TNF could play a key role in mediating insulin resistance by inhibiting insulin action on adipocytes [20–24]. A high white blood cell count , and high levels of CRP, IL-6 and fibrinogen [26,27] are predictive for the development of Type 2 diabetes, but this predictive value is significantly attenuated after accounting for differences in body mass and insulin resistance. Interestingly, the classic anti-inflammatory drug, aspirin, causes a reduction in insulin resistance in animal models of obesity and in obese humans [28,29]. Regardless of the etiopathological mechanisms, all these data clearly demonstrate the clustering of inflammation, insulin resistance and obesity that strongly increases the cardiovascular risk.
Plasminogen activator inhibitor-1 and IRS
It is now well established that hypofibrinolysis due to elevated plasma PAI-1 levels is a core feature of IRS. In both obese insulin-resistant patients and non-insulin-dependent (Type 2) diabetic patients, PAI-1 levels are elevated; they are in the normal range in Type 1 diabetic patients [30,31]. Plasma PAI-1 levels are strongly associated with the parameters of IRS (body mass index [BMI], visceral fat, blood pressure, plasma levels of insulin or proinsulin, triglycerides, small dense low density lipoprotein particles, free fatty acids, HDL cholesterol) [1,4,31–41], whereas no correlation is observed with measures of glycemic control .
The modulation of IRS by weight loss with a hypocaloric diet [42,43], physical training, change in dietary composition, or use of oral antidiabetic drugs such as metformin , induces a decrease in PAI-1 levels that correlates with the decrease in weight and in plasma metabolic parameters. It was shown recently that the use of thiazolinediones such as troglitazone is associated with a decrease in plasma PAI-1 levels in patients with Type 2 diabetes .
Many cells/tissues could synthesize PAI-1. PAI-1 gene expression is inducible rather than constitutive and one of the foremost questions is to identify PAI-1 inducers in IRS. It was proposed that metabolic disturbances observed during this syndrome directly affect PAI-1 synthesis. Most cell-culture experiments showed that insulin [46–49], glucocorticoids , very low density lipoprotein , free fatty acid , glucose  and angiotensin II  increase PAI-1 production. However, these findings have not always been confirmed in vivo. An acute in vivo administration of insulin did not modify plasma PAI-1 concentration , and even decreased it .
Among other potential inducers of PAI-1 synthesis, the pro-inflammatory cytokine TNF appears of particular relevance in IRS [20,21]. Indeed, it can disrupt insulin-signaling pathways in pancreatic beta cells, liver and adipose tissue . In mice, the inactivation of TNF receptors (RI and RII) results in significantly reduced adipose tissue and plasma PAI-1 levels, suggesting that the TNF pathway could induce PAI-1 in IRS  by a possible direct transcriptional effect . Increased TNF expression has been described in adipose tissue from obese subjects, and thus adipose tissue was proposed as a privileged source of PAI-1 in IRS. Indeed, PAI-1 is synthesized by human adipose tissue [38,58–61]. We have shown a decrease of PAI-1 expression in adipose tissue correlating with weight loss , and a variability of PAI-1 expression depending on the areas of fat; visceral fat produces more PAI-1 than subcutaneous abdominal or femoral fat [58,60–62]. The main PAI-1-producing cells have been identified as stroma cells , some of them being of monocyte or smooth muscle cell origin. In human adipose tissue, PAI-1 antigen content correlates with TNF receptors and TGFβ antigen tissue contents, but not with that of TNF or IL-6, which are strongly correlated to each other [64,65]. Although adipose tissue could represent a target tissue for TNF and a reservoir of PAI-1 in IRS, some experimental and clinical findings clearly indicate that adipose tissue is not the main source of PAI-1. Insulin-resistant obese mice exhibit increased PAI-1 expression in several organs , and insulin-resistant patients suffering lipodystrophy present very high PAI-1 plasma levels (unpublished data). A classical finding of IRS is non-alcoholic steatohepatitis. It was shown that plasma PAI-1 levels are strongly correlated with levels of liver enzymes [67,68]. Furthermore, we have observed a highly significant association between plasma PAI-1 levels and the degree of liver steatosis in mice models of obesity and alcohol-induced liver damage as well as in obese humans . Thus, steatotic liver or the mechanisms that lead to steatosis may be involved in PAI-1 synthesis in IRS.
Consequences of PAI-1
Role of PAI-1 in coronary heart disease
High plasma PAI-1 levels have been considered as a risk factor for coronary heart disease [70–74]. PAI-1 has been shown to be predictive of events in univariate analysis. However, the predictivity disappeared after adjustment for BMI, triglycerides and HDL cholesterol, which are markers of insulin resistance . Sobel et al.  and Pandolfi et al.  have provided evidence that Type 2 diabetes is associated with an increased PAI-1 expression in arterial wall. This increased PAI-1 concentration in the vessel wall, as well as the increased PAI-1 level in plasma, could participate in increased cardiovascular risk and unfavourable plaque evolution in IRS and diabetes.
Several groups have used mouse models to test prospectively whether elevated PAI-1 expression promotes thrombosis and atherosclerotic lesion development. While it is clear that overexpression of PAI-1 favours the development of thrombosis , the exact role of PAI-1 in vascular remodeling remains controversial . Differences in results are probably related to the type of lesions, the duration of the lesion development and the genetic background of the mice . It seems that PAI-1 may limit cell migration in one early remodeling process, but enhance fibrin accumulation at later timepoints, promoting cell proliferation [78,80–82]. In apolipoprotein E-deficient mice lacking PAI-1, larger plaques at all sites of the vasculature were observed, but only at advanced stages of atherosclerosis. Rare fibrin deposits were observed in this study and therefore, the involvement of other mechanisms controled by PAI-1 was suggested, such as an effect on macrophage infiltration or on the level of active TGFβ known to control matrix deposition .
Role of PAI-1 in obesity and IRS
The high expression of PAI-1 in IRS raises questions on the PAI-1 contribution to this phenomenon. As PAI-1 is involved in tissue remodeling, it can be suggested that the elevated expression of PAI-1 observed in obesity is involved in adipose tissue development. The modulation of the PAI-1 gene in mice fed a high-fat diet induces change in weight gain; PAI-1 knockout mice gain weight faster  whereas PAI-1 transgenic mice with an overexpression of PAI-1 gain less weight . In the same way, plasminogen knockout mice fed a high-fat diet gained less weight . In vitro experiments support the fact that the hypofibrinolysis induced by excess of PAI-1  or the absence of plasminogen  leads to a decrease of preadipocyte differentiation, with consequences for adipose tissue growth. Thus, PAI-1 overexpression in IRS could control the fat mass mainly by reducing adiposity. In contrast to these studies, adiposity was reduced in genetically obese, leptin-deficient mice by disruption of the PAI-1 gene . This discrepancy should be further investigated.
Interestingly, it was recently demonstrated in a large cohort of healthy non-diabetic subjects that patients who developed an incident diabetes within 5 years presented higher levels of PAI-1 as well as of fibrinogen and C-reactive protein (CRP) at baseline than non-converters . PAI-1 predicted Type 2 diabetes independently of insulin resistance and other known risk factors for diabetes. In this way, treatment with an ACE inhibitor has been shown to decrease not only PAI-1 levels  and the rate of cardiovascular events  but also the incidence of Type 2 diabetes . This suggests that PAI-1 could play a role in the development of cardiovascular diseases, and in Type 2 diabetes.
PAI-1 could represent a potential target for therapeutic intervention that aims to decrease the risk of both cardiovascular disease and Type 2 diabetes. Studies with use of PAI-1 inhibitors in animal models of obesity and insulin resistance are needed to delineate the true contribution of PAI-1 to both cardiovascular and metabolic complications of insulin resistance.