Search

Following the observation by Brown et al. (Am J Physiol 1997; 272: C937-49) that primary rat adipocytes in culture secrete gelatinase A (MMP-2), we have evaluated gelatinase expression in adipose tissue with the use of mouse models of obesity. Wild-type mice were kept on a standard fat diet (SFD) or on a high fat diet (42% fat, HFD) and- genetically obese db/db mice were kept on SFD; gonadal and subcutaneous fat pads were removed and analysed ex vivo. These studies revealed that: 1) the HFD induced adipocyte hypertrophy; 2) after 32 weeks, significantly higher levels of 70 kDa (p <0.05) and 65 kDa proMMP-2 (p <0.01) were observed in extracts of gonadal fat pads of mice on HFD; 3) the contribution of active MMP-2 to the total level was comparable in SFD and HFD groups (20 to 30%); and 4) gelatinase B (MMP-9) was not consistently detected. These findings were confirmed by gelatin zymography and by mRNA determination using competitive RT-PCR. The presence of MMP-2 in the adipose tissue was confirmed immunologically and its localization in adipocytes revealed by immunogold electron microscopy. The potential functional role of MMP-2 in adipose tissue remains to be determined.

The objective of this study was to assess the interest of antiphosphatidylethanolamine antibodies (aPE) in unexplained thrombosis (UT) defined as thrombotic episode without any of the main autoimmune and hereditary thrombophilic defects. Results from 98 UT were compared to those of (I) 142 patients with thrombophilia: 67 antiphospholipid syndrome (APS) and 75 hereditary hemostatic defects (HHD); (II) 110 patients without thrombosis: 60 with systemic lupus erythematosus (SLE) and 50 with infectious diseases (ID). As compared to controls (100 blood donors), aPE prevalence was significantly higher in both autoimmune contexts (APS: 43%; SLE: 40%, p<0.0001) and among non-autoimmune pathologies, only in UT (18%, p = 0.001) conversely to HHD (8%) or ID (10%). aPE prevalence in UT was not statistically different from that found in Primary APS (32%, p = 0.076) but lower than in Secondary APS (65%, p <0.005). In UT, aPE were mainly of IgM isotype like in Primary APS and they were found alone whereas in SLE they were always associated with classical antiphospholipid antibodies. No significant association was found between any isotype of aPE and a site of thrombosis in UT as well as in APS. In conclusion, this study demonstrates an increase of the prevalence of aPE in patients with unexplained thrombosis. Thus, aPE investigation appears to be of interest in UT and their persistent presence could define a biological variant of APS.

Plasminogen activator inhibitor-1 (PAI-1) is a specific inhibitor of plasminogen activators and may be the principal regulator of plasminogen activation in vivo. PAI-1 levels are elevated in insulin-resistant subjects and are associated with an increased risk of atherothrombosis. After adjustment for metabolic parameters, increased PAI-1 levels were no longer considered as a cardiovascular risk factor. The mechanisms underlying the strong association between PAI-1 levels and the metabolic disturbances found in insulin resistance are still not understood. Several studies have suggested that visceral adipose tissue may be a major source of PAl-1. Accordingly, adipose tissue PAI-1 production particularly that from visceral fat, was found to be elevated in obese human subjects. Within human adipose tissue, stromal cells appear to be the main cells involved in PAI-1 synthesis. This review discusses the potential mechanisms linking adipose tissue to plasma PAI-1 levels such as the intervention of cytokines (TNFalpha and TGFbeta), free fatty acids and hormones (insulin and glucocorticoids). Moreover alteration of adipose tissue cellular composition induced by the modulation of PAI-1 expression opens a novel field of interest.

The Thrombin Activatable Fibrinolysis Inhibitor (TAFI) is a recently described inhibitor of fibrinolysis. The physiological variations of plasma TAFI antigen are not well known. We studied TAFI antigen values in healthy populations with a commercially available kit from Milan Analytica (Switzerland). Broad range of TAFI antigen values (from 41% to 259%) was found in a population of 249 healthy individuals. Gender as well as pregnancy did not influence mean values of TAFI antigen. There was a positive correlation between TAFI antigen and age in female (r = 0.28; p <0.05) but not in male populations. Mean TAFI antigen value of a black African male group [mean +/- SD (range): 87 +/- 23 (39-144%)] was significantly lower than the one of age matched Caucasian men [114 +/- 34 (52-259%)] (p <0.0001). TAFI antigen values were very stable within individuals, they did not significantly vary on day time or at several months period. Thus, in contrast to large inter-individual variations, TAFI antigen levels are quite stable within individuals.

Background: Type 1 plasminogen activator inhibitor (PAI-1) is considered to be risk factor for acute myocardial infarction (AMI). A rebound of circulating PAI-1 has been reported after rt-PA administration. We investigated the relationships between PAI-1 levels before and after thrombolytic therapy with streptokinase (SK) as compared to rt-PA and the patency of infarct-related arteries., Methods and Results: Fifty five consecutive patients with acute MI were randomized to streptokinase or rt-PA. The plasma PAI-1 levels were studied before and serially within 24 h after thrombolytic administration. Vessel patency was assessed by an angiogram at 5+/-1days. The PAI-1 levels increased significantly with both rt-PA and SK as shown by the levels obtained from a control group of 10 patients treated with coronary angioplasty alone. However, the area under the PAI-1 curve was significantly higher with SK than with rt-PA (p<0.01) and the plasma PAI-1 levels peaked later with SK than with rt-PA (18 h versus 3 h respectively). Conversely to PAI-1 levels on admission, the PAI-1 levels after thrombolysis were related to vessel patency. Plasma PAI-1 levels 6 and 18 h after SK therapy and the area under the PAI-1 curve were significantly higher in patients with occluded arteries (p<0.002, p<0.04 and p<0.05 respectively). The same tendency was observed in the t-PA group without reaching significance., Conclusions: This study showed that the PAI-1 level increase is more pronounced after SK treatment than after t-PA treatment. There is a relationship between increased PAI-1 levels after thrombolytic therapy and poor patency. Therapeutic approaches aimed at quenching PAI-1 activity after thrombolysis might be of interest to improve the efficacy of thrombolytic therapy for acute myocardial infarction.

Circulating (PAI-1) levels are elevated in patients with coronary heart disease and may play an important role in the development of atherothrombosis. Many clinical studies have indicated that the insulin resistance syndrome, which is a situation predisposing to diabetes and ischemic heart disease, may be a major regulator of PAI-1 expression, especially in determining plasma PAI-1 levels. Central obesity is a characteristic of insulin resistance and is a well recognized risk factor for coronary heart disease. Recently the production of PAI-1 by adipose tissue, in particular by tissue from omentum, has been demonstrated and could be an important contributor to the elevated plasma PAI-1 levels observed in insulin resistant patients. Besides the effect of the metabolic status on plasma PAI-1 levels, the role of a genetic control has been emphasized, but according to recent results obtained in a family segregation study, its participation seems limited. Prospective cohort studies of patients with previous myocardial infarction or angina pectoris have underlined the association between increased plasma PAI-1 levels and the risk of coronary events, but the predictive capacity of PAI-1 disappears after insulin resistance marker adjustments. Taken together these results support the notion that PAI-1 can be a link between obesity, insulin resistance and cardiovascular disease.

Congenital deficiencies of protein S (PS) are associated with thrombophilia. Their characterization and classification have been hampered by the complex physiology of the protein C-protein S system and the poor standardization and reliability of laboratory assays. The free active form of protein S is usually determined by immunoassay using polyclonal antibodies in the plasma supernate after polyethyleneglycol (PEG) precipitation. A new one step ELISA using two monoclonal antibodies specific for distinct epitopes of the free form of protein S has been developed for the direct measurement of free PS in untreated plasma. We have tested two ELISA assays for free PS. One assay was based on the PEG precipitation (Asserachrom PS, Stago, Asnières, France) whereas the other was a one step ELISA assay (Asserachrom free PS, Stago). Values were obtained in 35 PS deficient patients recruited among 500 consecutive patients evaluated by the laboratory for diagnosis of congenital disorders of coagulation. Values were compared to those obtained in 50 patients with no PS deficiency matched for age and sex with the PS deficient patients as well as in 33 normal subjects and in 12 pregnant women. Strong correlation was found between the two tests (r = 0.81, p < 10(-5)) in the entire population (n = 130), as well as in the separate groups. The new one step ELISA was more accurate than the PEG free PS determination. Determination of PS activity and antigens allowed us to separate quantitative and qualitative deficiencies. Among the qualitative deficiencies, isolated decrease in PS activity was the most frequent defect observed (66%). This fact questions the substitution of PS activity assays by the one step antigenic free PS ELISA assay.

Insulin resistance represents a situation with a high risk of atherothrombosis and is accompanied by increased plasma plasminogen activator inhibitor-1 (PAI-1) levels. Fasting insulin level is highly correlated with PAI-1 levels in plasma. It has been shown that insulin increases PAI-1 synthesis by the human hepatoma cell line Hep G2. Moreover when Hep G2 cells expressing a down-regulation of insulin receptors by incubation with 10(-7) M insulin, were stimulated by 10(-9) M insulin, an overexpression of PAI-1 synthesis was observed despite a reduced number of insulin receptors. Insulin-like growth factor 1 (IGF-1) shares many properties with insulin. The aim of the present study was to evaluate the effect of IGF-1 on PAI-1 synthesis by Hep G2 cells down-regulated either by insulin or IGF-1. Incubation of Hep G2 cells with increasing doses from 10(-9) to 10(-7) M IGF-1 induced a dose-dependent stimulation of PAI-1 synthesis up to 4.5-fold the control level. When cells were first pre-incubated with 10(-7) M IGF-1 for 18 h, acid washed, and then stimulated with 10(-9) M IGF-1, the expression of IGF-1 receptors was greatly reduced (up to 70%). In contrast PAI-1 secretion was increased 3.4-fold the level of control cells and by 1.9-fold the level of cells first stimulated with 10(-9) M IGF-1. Both transcripts of PAI-1 mRNA were also increased.(ABSTRACT TRUNCATED AT 250 WORDS)

Platelets have been described to contain a large proportion of the circulating plasminogen activator inhibitor type 1 (PAI-1) which is released on platelet activation. This protein could be taken up by platelets from the plasma or synthesized by megakaryocytes (MKs). Recently, PAI-1 mRNA has been detected in a human megakaryoblastic leukemia cell line (MEG-01) by the polymerase chain reaction (PCR). However, a direct-demonstration of its presence in normal human MKs is lacking. In order to prove directly the megakaryocytic origin of platelet PAI-1, the MEG-01 cell line, human bone marrow enriched in MKs, and bone marrow smears from allogeneic bone marrow transplantation donors were investigated for the presence of PAI-1 mRNA using in situ hybridization (ISH). Specimens of bone marrow were first stained with May-Grünwald Giemsa (MGG) for cell identification according to their morphology. Subsequently, the same slides were used for ISH. PAI-1 mRNA was clearly demonstrated in the MEG-01 cell line and in MKs, and its presence correlated with the detection of PAI-1 antigen by immunocytochemistry. PAI-1 mRNA was also detected in morphologically characterized mature granulocytes of marrow samples.

Individuals with elevated levels of plasminogen activator inhibitor type 1 are at risk of developing atherosclerosis. The mechanisms leading to increased plasma PAI-1 concentrations are not well understood. The link observed between increased PAI-1 levels and insulin resistance has lead workers to investigate the effects of insulin or triglyceride rich lipoproteins on PAI-1 production by cultured hepatocytes or endothelial cells. However, little is known about the contribution of these cells to PAI-1 production in vivo. We have studied the expression of PAI-1 in human liver sections as well as in vessel walls from different territories, by immunocytochemistry and in situ hybridization. We have observed that normal liver endothelial cells expressed PAI-1 while parenchymal cells did not. However, this fact does not refute the role of parenchymal liver cells in pathological states. In healthy vessels, PAI-1 mRNA and protein were detected primarily at the endothelium from the lumen as well as from the vasa vasorum. In normal arteries, smooth muscle cells were able to produce PAI-1 depending on the territory tested. In deeply altered vessels, PAI-1 expression was observed in neovessels scattering the lesions, in some intimal cells and in smooth muscle cells. Local increase PAI-1 mRNA described in atherosclerotic lesions could be due to the abundant neovascularization present in the lesion as well as a raised expression in smooth muscle cells. The increased PAI-1 in atherosclerosis could lead to fibrin deposit during plaque rupture contributing further to the development and progression of the lesion.

The mechanism underlying diurnal variations in PAI-1 as well as the cellular origin of PAI-1 in subjects with high PAI-1 levels are unknown. We evaluated diurnal changes (8:00 am vs 4:00 pm) in PAI-1 (functional and immunological assays), t-PA Ag and t-PA/PAI-1 complex levels in controls and subjects with high PAI-1 levels. Three test groups were recruited among obese hyperinsulinemic subjects, emergency care unit patients with inflammatory syndrome or infection and pregnant women. The classical afternoon decrease of PAI-1 level was observed in controls and obese subjects but its amplitude was greater in the latter. The decrease in t-PA Ag and t-PA/PAI-1 complex levels was the same in controls and in obese. As, in previous studies, elevated PAI-1 levels have been correlated with insulin resistance and a decrease in insulin sensibility has been described in the early morning, it is proposed that this "dawn phenomenon" could be implicated in the circadian variations of PAI-1 in controls and could be amplified in obese subjects. Great variability in PAI-1, t-PA Ag or t-PA/PAI-1 complex levels was observed in patients with acute inflammatory syndrome or infection for whom classical biorhythms are suppressed. No diurnal changes in PAI-1 and other fibrinolytic parameters were observed in patients with inflammatory syndrome or in pregnant women suggesting that other sources and/or other regulatory mechanisms of PAI-1 production are involved.

Capillary cells represent 95% of the vascular bed, and cells from large and micro-vessels do not express identical functions. In order to study the hormonal regulation of plasminogen activator inhibitor 1 (PAI-1) secretion by human capillary cells we used epithelial cells from omental tissue (HOTMEC). As their endothelial origin is subject to controversy, we attempted to determine their characteristics by comparing them to human umbilical vein endothelial cells (HUVEC). Morphological and biological criteria were studied. By phase contrast microscopy HOTMEC elicited a cobblestone pattern similar to HUVEC. Weibel-Palade bodies were not found in the cytoplasm with electron microscopy. Fluorescence microscopy studies indicated that HOTMEC took up acetylated-LDL more intensely than HUVEC, and showed no staining for von Willebrand factor. The phenotype of HOTMEC was studied by flow cytometry using monoclonal antibodies (mo Ab) directed against epitopes either specific for endothelial cells or for mesothelial cells. We showed that in our preparations only 10% of cells reacted with mo Ab specific for endothelial cells. About 60% of the HOTMEC were labelled with an antibody directed against mesothelial cells. HOTMEC expressed fibrinolytic factors. Tissue plasminogen activator (t-PA) levels in HOTMEC conditioned medium were 50 fold higher than those of HUVEC, and the PAI-1 secretions were identical in both cell types. Insulin which is known to increase PAI-1 synthesis by hepatocytes did not enhance the PAI-1 level either in HOTMEC or in HUVEC conditioned media. Our results suggested that morphological and functional methods did not allow discrimination between the cell types present in the omentum tissue.(ABSTRACT TRUNCATED AT 250 WORDS)

It has been reported that the level of PA-inhibitor increases in postoperative patients and on the other hand that glucocorticoids increase the PA-inhibitor level in cell culture. Because surgery is associated with increased plasma cortisol level, a relation between the postoperative increase in plasma cortisol and PA-inhibitor levels was looked for. Blood samples were collected from 8 patients undergoing extensive abdominal surgery, before operation and postoperatively at 2 hr, 4 hr, 24 hr and daily for 7 days. Plasma cortisol and PA-inhibitor were increased 2 hr after surgery, when there was a significant correlation (p less than 0.05). The maximum increase was at 24 hr and the values fell to normal on day 6. An increase in t-PA related antigen (t-PA R:Ag) and a decrease in euglobulin fibrinolytic activity (EFA) also occurred. In 7 controls 0.25 mg ACTH was given intravenously and blood was collected after 1/2, 1, 2, 4, 6 hr. Although the increase in plasma cortisol level following ACTH was comparable to that observed after surgery the increase was not associated with significant change in PA-inhibitor level, t-PA R:Ag or EFA. A cause-effect relationship between the increased plasma cortisol and PA-inhibitor level could not be shown. The mechanism of the postoperative increase in PA-inhibitor thus remains unknown.

Secretion of plasminogen activator inhibitor 1 (PAI-1) by cultures of human umbilical vein endothelial cells and human hepatocellular cell line Hep G2 was evaluated after insulin stimulation. The secretion of PAI-1 antigen and activity was measured in the conditioned medium and the cellular extracts after incubation of confluent cultures with 1% serum medium for 24 hours. Insulin induced a dose dependent increase of the PAI-1 secretion by Hep G2 cell line. At 10(-8) M a two fold increase of PAI-1 antigen and activity were observed whereas alpha 2 antiplasmin and fibrinogen were not significantly modified. No effect of insulin was observed on PAI-1 antigen and PAI activity production by human endothelial cells whereas endotoxin resulted in a two fold increase in PAI-1 secretion. In recent clinical studies we have demonstrated that the level of plasma insulin correlated with that of PAI-1. Thus we hypothesize that hepatocytes represent a physiological source of plasma PAI-1 which is modulated by plasma insulin level.

The effect of heparin and of two low molecular weight (low Mr) fractions of heparin on thrombolysis with recombinant human tissue-type plasminogen activator (rt-PA, Genentech Inc., So. San Francisco, CA) or human single chain urokinase-type plasminogen activator (scu-PA, Sandoz AG, Basle, Switzerland) was measured in a rabbit jugular vein thrombosis model. Four bolus injections of 200 anti-Factor Xa units/kg body weight of heparin (Liquemine, Hoffmann-La Roche, Basle, Switzerland), of 90 units/kg of CY 216 (Choay, Paris, France) or of 90 units/kg of CY 222 (Choay, Paris, France) were given intravenously, immediately after the start of the infusion of rt-PA or scu-PA and at hourly intervals during their intravenous infusion over 4 hours. The bolus injections resulted in anti-Factor Xa levels in plasma of 5.7 +/- 1.2 units/ml just before the repeat bolus injections of heparin with corresponding values of 3.9 +/- 0.2 units/ml for CY 216 and 1.6 +/- 0.2 units/ml for CY 222. Thrombolysis with 0.25 mg/kg rt-PA was 36 +/- 1 percent (n = 9) in the absence of anticoagulant, 40 +/- 1 percent (n = 7, p less than 0.05) in the presence of heparin, 49 +/- 5 percent (n = 7, p less than 0.02) with CY 216 and 62 +/- 5 percent (n = 7, p less than 0.01) with CY 222.(ABSTRACT TRUNCATED AT 250 WORDS)

The fibrinolytic system was investigated in 120 patients with spontaneous or recurrent deep vein thrombosis (DVT) without any known organic disease able to explain by itself the occurrence of a thrombosis and without any known defect of antithrombin III, Heparin Cofactor II, Protein C, or Protein S. The assays included: Euglobulin fibrinolytic activity (EFA), tissue-type plasminogen activator related antigen (t-PA-Ag) and plasminogen activator inhibitor activity (PA inhibitor), which were measured before and after 10 min of venous occlusion (V.O.). On the basis of the results, the patients could be classified in 3 groups: good responders with an at least two-fold increase of EFA after venous occlusion (n = 76), poor responders with a lesser increase of EFA due to deficient release of t-PA (n = 12), and poor responders with a normal t-PA release but an increased level of PA-Inhibitor (n = 32). The poor responders due to deficient t-PA release (10% of total) had a higher incidence of recurrence of deep vein thrombosis, than the other groups (p less than 0.01). An overall correlation was found between the level of PA-Inhibitor activity and the triglyceride level (r = 0.40, p less than 0.01), suggesting that these elevations may be due to a common cause, at least in some of the patients. It is concluded that a poor fibrinolytic response to venous occlusion occurs in 35 percent of DVT patients.(ABSTRACT TRUNCATED AT 250 WORDS)

We have previously observed a positive correlation between Plasminogen Activator Inhibition capacity (PA Inhibition), Body Mass Index (BMI) and plasma insulin levels in a population of non diabetic subjects. The anti diabetic biguanide Metformin which decreases insulin resistance has been reported to increase the blood fibrinolytic activity. Therefore we have studied the effect of Metformin on PA Inhibition levels in obese subjects with normal glucose tolerance. Eighteen obese women (O) (BMI: 31.4 +/- 1.13, m +/- S.E.M.) were compared to age matched controls (C) (BMI: 20.2 +/- 0.8) and randomized to a 15 days treatment by Metformin (M) (1.7 g/day) or placebo (P) in a double blind study while on a weight maintaining diet. O compared to C had higher levels (m +/- S.E.M.) of PA Inhibition (9 +/- 1.8 IU/ml, versus 2.88 +/- 0.29 p less than 0.01), lower euglobulin fibrinolytic activity (EFA) (4.95 +/- 1.17 mm versus 9 +/- 0.29 p less than 0.05), higher plasma insulin (24.1 +/- 2.1. uU/ml), versus 12 +/- 1 p less than 0.01) and triglyceride (1.32 +/- 0.16 mmol/l, versus 0.8 +/- 0.08 p less than 0.05). After 15 days of treatment, in group M a significant decrease in PA Inhibition (5.51 +/- 1.4, versus 9.48 +/- 2.1 p less than 0.05) in plasma insulin (18.5 +/- 0.1, versus 24.5 +/- 3.5, p less than 0.05) and plasma triglyceride (1.08 +/- 0.1, versus 1.47 +/- 0.3 p less than 0.05) and an increase in EFA (6.50 +/- 0.28, versus 5.25 +/- 0.35 p less than 0.05) were observed. No significant variation was observed in group P.

Elderly patients have previously been shown to have an increased plasma concentration of tissue plasminogen activator (t-PA) antigen (t-PA Ag). Since the concentration of t-PA Ag depends on both free t-PA and t-PA complexed with inhibitors, mainly plasminogen activator inhibitor (PA inhibitor), we have investigated the relationship between the plasma concentration of PA inhibitor and age in 20 elderly and 20 young individuals. Elderly individuals showed a slight increase in PA inhibitor, in parallel with increase on others, acute-phase proteins, fibrinogen, von Willebrand factor, factor VIII:C, and the erythrocyte sedimentation rate. The increase in PA inhibitor as well as other acute-phase proteins in the elderly may be significant in relation to the increased incidence of thrombotic disease.

Type 2 diabetic patients are known to frequently have a high insulin level and were recently described as having high plasminogen activator inhibitor (PAI) activity, compared to normal controls. As we have shown in several clinical conditions (normal subjects, obese patients, angina pectoris patients) that plasma PAI activity was linked with plasma insulin, we have studied in 38 type 2 diabetic patients the relationship between PAI activity, insulin and other parameters. Patients showed higher level of PAI activity, as well as plasma glucose, insulin, triglyceride, cholesterol and Apolipoprotein B levels than normal controls; highest values were observed with diabetic patients also affected by coronary artery disease. A significant correlation was found between PAI activity and insulin (r = 0.60, p less than 0.001), body mass index (r = 0.32, p less than 0.05) and Apolipoprotein B (r = 0.33, p less than 0.05). The two latter correlations disappeared after adjustment for insulin. These results are in agreement with our previous report showing an in vitro effect of insulin on the synthesis of PAI by a hepatocellular cell line. Hyperinsulinemia presented by type 2 diabetic patients may increase the hepatic synthesis of PAI, inducing an hypofibrinolysis, which could play a role in the development of the vascular complications. Attempts to reduce hyperinsulinemia could have a favorable effect by lowering PAI activity.