Suppression of TAFI by siRNA inhibits invasion and migration of breast cancer cells.

Human thrombin activatable fibrinolysis inhibitor (TAFI), also known as carboxypeptidase B2 (CPB2), is a procarboxypeptidase enzyme. The purpose of the present study was to observe the expression of TAFI in breast cancer (BC) and breast cancer cell (BCC) lines and to investigate the effect of TAFI suppression by small interfering (si)RNA gene silencing on invasion and migration of BCC lines. A significant increase in TAFI level was identified by immunohistochemical analysis in BC tissues compared with normal breast tissues. TAFI suppression also inhibited cell viability, invasion and migration ability as demonstrated by MTT, Transwell chamber, and wound scratch assays, respectively (P<0.05). The data suggested that suppression of TAFI by siRNA inhibits invasion and migration of breast cancer cells and that TAFI may be a new target for breast cancer therapy.

Pro-inflammatory cytokines reduce human TAFI expression via tristetraprolin-mediated mRNA destabilisation and decreased binding of HuR.

Thrombin activatable fibrinolysis inhibitor (TAFI) is the zymogen form of a basic carboxypeptidase (TAFIa) with both anti-fibrinolytic and anti-inflammatory properties. The role of TAFI in inflammatory disease is multifaceted and involves modulation both of specific inflammatory mediators as well as of the behaviour of inflammatory cells. Moreover, as suggested by in vitro studies, inflammatory mediators are capable of regulating the expression of CPB2, the gene encoding TAFI. In this study we addressed the hypothesis that decreased TAFI levels observed in inflammation are due to post-transcriptional mechanisms. Treatment of human HepG2 cells with pro-inflammatory cytokines TNFα, IL-6 in combination with IL-1ß, or with bacterial lipopolysaccharide (LPS) decreased TAFI protein levels by approximately two-fold over 24 to 48 hours of treatment. Conversely, treatment of HepG2 cells with the anti-inflammatory cytokine IL-10 increased TAFI protein levels by two-fold at both time points. We found that the mechanistic basis for this modulation of TAFI levels involves binding of tristetraprolin (TTP) to the CPB2 3'-UTR, which mediates CPB2 mRNA destabilisation. In this report we also identified that HuR, another ARE-binding protein but one that stabilises transcripts, is capable of binding the CBP2 3'UTR. We found that pro-inflammatory mediators reduce the occupancy of HuR on the CPB2 3'-UTR and that the mutation of the TTP binding site in this context abolishes this effect, although TTP and HuR appear to contact discrete binding sites. Interestingly, all of the mediators tested appear to increase TAFI protein expression in THP-1 macrophages, likewise through effects on CPB2 mRNA stability.

Evidence that fibrinolytic changes in paediatric obesity translate into a hypofibrinolytic state: relative contribution of TAFI and PAI-1.

Paediatric obesity, like adulthood obesity, is associated with an increase of fibrinolysis inhibitors. No study, however, has evaluated the impact of these changes on plasma fibrinolytic capacity. We investigated plasma fibrinolysis and the role therein of the fibrinolytic changes associated with obesity in 59 obese children (body mass index > 95th percentile) and 40 matched controls. Fibrinolysis was investigated by measuring 1) the plasma levels of relevant fibrinolytic factors; 2) the in vitro fibrinolytic capacity under different conditions, using a microplate plasma clot lysis assay; 3) the circulating levels of markers of clotting and fibrinolysis activation. Plasminogen activator inhibitor 1 (PAI-1), total thrombin activatable fibrinolysis inhibitor (TAFI) and fibrinogen levels were higher in obese children as compared to controls (p<0.01). Plasma clots from obese children lysed significantly slower than control clots when exposed to exogenous plasminogen activator, indicating a greater resistance to fibrinolysis. By the use of a selective inhibitor of activated TAFI and by regression analyses we found that fibrinolysis resistance in obese samples was attributable to PAI-1 increase and to enhanced TAFI activation. The ratio between the circulating levels of D-dimer and thrombin-antithrombin complex, a marker of in vivo fibrinolysis, was significantly lower in obese children, suggesting a reduced fibrinolytic efficiency. These data indicate that paediatric obesity is associated with a hypofibrinolytic state which might contribute to the increased thrombotic risk associated with this condition.

Plasma levels of unactivated thrombin activatable fibrinolysis inhibitor (TAFI) are down-regulated in young adult women: analysis of a normal Japanese population.

Thrombin-activatable fibrinolysis inhibitor (TAFI) is an anaphylatoxin-inactivating enzyme generated by proteolytic cleavage of its zymogen, and is the same enzyme as that first designated by our group as procarboxypeptidase R (proCPR). TAFI in plasma is presumed to influence vascular disease in its role as a fibrinolysis inhibitor. The activity of TAFI is strongly influenced by genetic polymorphism, especially at amino acids Thr/Ala-147 and Thr/Ile-325. In this study, we analyzed 202 healthy controls who were not on any medication, had no unusual medical history and whose blood data were normal. In a previous report, we established an enzyme-linked immunosorbent assay (ELISA) specific for non-activated TAFI (proCPR), and investigated levels of unactivated TAFI as an estimate of anti-fibrinolytic capacity. In this study, we determined normal Japanese TAFI levels for each age, sex, and genetic polymorphism of Thr/Ala-147 and Thr/Ile-325, and also showed that the TAFI level in young adult women is lower than in aged women.

Plasma activity of thrombin activatable fibrinolysis inhibitor in Crimean-Congo hemorrhagic fever.

OBJECTIVES: Crimean-Congo hemorrhagic fever (CCHF) is a potentially fatal viral infection. The exact mechanism for hemorrhage remains unknown. Thrombin activatable fibrinolysis inhibitor (TAFI) is a plasma procarboxypeptidase B-like proenzyme and synthesized in the liver, down-regulating fibrinolysis. In this study, we measured the TAFI activity in plasma of patients with CCHF to examine the relationship between hemorrhage and the pathogenesis of CCHF. METHODS: Twenty-one patients and similar number of controls were included in the study. The diagnosis of CCHF was confirmed through detection of IgM by ELISA and/or PCR. TAFI activity was measured in plasma samples. RESULTS: TAFI activity in CCHF patient group was mean 7.2+/-2.3 microg/ml (range: 0.95-10.31 microg/ml) and in the control group was mean 11.7+/-4.1 microg/ml (range: 3.07-23.9 microg/ml). There was a significant decrease of TAFI activity in CCHF patients when compared to controls. A positive correlation between CRP, PT, INR, serum albumin and TAFI activity levels were found. We suggest that the decrease of TAFI activity may be due to liver dysfunction during viral active disease state. CONCLUSIONS: Low TAFI activity may be an attributable factor, leading to imbalance in fibrinolysis, resulting in bleeding complications in CCHF.

Generation of a stable activated thrombin activable fibrinolysis inhibitor variant.

Activated thrombin activable fibrinolysis inhibitor (TAFIa), generated upon activation of TAFI, exerts an antifibrinolytic effect. TAFIa is a thermolabile enzyme, inactivated through a conformational change. The objective of the current study was to generate a stable variant of human TAFIa. Using a site-directed as well as a random mutagenesis approach to generate a library of TAFI mutants, we identified two mutations that increase TAFIa stability, i.e. a Ser305 to Cys and a Thr329 to Ile mutation, respectively. Combining these mutations in TAFI-Ala147-Ile325, the most stable isoform of TAFIa (half-life of 9.4 +/- 0.4 min), revealed a TAFIa half-life of 70 +/- 3.1 min (i.e. an 11-fold increase versus 6.3 +/- 0.3 min for TAFIa-Ala147-Thr325, the most frequently occurring isoform of TAFI in humans) at 37 degrees C. Moreover, clot lysis (induced by tissue plasminogen activator) experiments in which TAFI-Ala147-Cys305-Ile325-Ile329 was added to TAFI-depleted plasma revealed a 50% clot lysis time of 313 +/- 77 min (i.e. a 3.0-fold increase versus 117 +/- 10 min for TAFI-Ala147-Thr325). The availability of a more stable TAFIa variant will facilitate the search for inhibitors and allow further structural analysis to elucidate the mechanisms of the instability of TAFIa.

Optimisation and evaluation of a high-throughput mammalian protein expression system.

Transient transfection of mammalian cells with episomal vectors is a very useful method for producing high levels of recombinant proteins. Transient systems remove the need for the laborious and time-consuming process of creating stable cell lines. Here, we describe the optimisation and evaluation of a high-throughput transient expression system in HEK293-EBNA cells. The process was developed for the expression of 10 constructs simultaneously in deep-well plates and subsequent purification using 96-well plate affinity chromatography. This enabled multiple combinations of different constructs, vectors, and expression conditions to be studied in parallel.

Properties of recombinant human plasma procarboxypeptidase U produced in mammalian and insect cells.

BACKGROUND: Carboxypeptidase U (EC 3.4.17.20, TAFIa) is a new member of the metallocarboxypeptidase family circulating in human plasma as a zymogen. It is activated during coagulation and is considered as an important player in the regulation of fibrinolysis. METHODS: Heterologous expression of human plasma procarboxypeptidase U (proCPU, TAFI) was obtained in mammalian cells (C127 and DON) and in insect cells (Sf21 and H5 cells). Conditioned media were purified by cation-exchange chromatography and plasminogen affinity chromatography to yield an essentially pure protein. RESULTS: All systems gave high expression levels (6-20 mg/l). Due to differences in glycosylation of the activation peptide, the recombinant variants of proCPU migrated differently on SDS-PAGE (52-65 kDa). However, after activation, all active recombinant enzymes migrated at 35 kDa, similar to native CPU and no evidence for post-translational modification of the catalytic domains could be detected. For the mammalian cell produced variants, activation was more efficient after desialylation. After activation, CPU showed low solubility (0.2 mg/ml) but was inhibited similarly as native CPU. CONCLUSIONS: Mammalian cell systems were the most efficient for the production of human plasma recombinant proCPU. The obtained zymogen differs with respect to the extent and the heterogeneity of glycosylation but, after activation, the experiments did not reveal any alteration between the recombinant and native protein.

Erythropoietin-induced thrombosis as a result of increased inflammation and thrombin activatable fibrinolytic inhibitor.

Chronic inflammation is a major cause of morbidity and mortality in end-stage renal disease. The associated anemia in these patients due to renal cortical atrophy and erythropoietin deficiency is treated with recombinant erythropoietin. Recent reports suggest a growing incidence of symptomatic venous thrombosis in cancer patients treated with recombinant erythropoietin. Several investigators have reported on different mechanisms of thrombosis in these patients. We hypothesize that thrombosis in patients with end-stage renal disease due to increased expression of C-reactive protein (CRP) as a result of chronic inflammation promotes the release of thrombin activatable fibrinolytic inhibitor causing fibrinolytic deficit and eventually thrombosis. Furthermore, because endothelial nitric oxide is responsible for the maintenance of the normal vascular function, the decreased levels of nitric oxide in chronic inflammation cause endothelial damage and result in thrombosis. To test this hypothesis, blood samples were collected from 106 patients (49 male and 57 female, aged 59.8+/-15.7 years) with end-stage renal disease undergoing hemodialysis and treated with recombinant erythropoietin at a mean dose of 201.8 U/kg/week. Blood samples were drawn in 5-mL tubes containing 3.2% sodium citrate just before the hemodialysis procedure. These blood samples were immediately centrifuged to obtain platelet-poor plasma, which was aliquoted and frozen at -70 degrees C until further analysis. Erytropoietin antibodies were measured using an anti-EPO enzyme-linked immunosorbent assay (ELISA) method developed in our laboratory. Nitric oxide was measured using a NO analyzer (Sievers 280I, Ionics, Boulder, CO). Plasma CRP levels were measured with a highly sensitive ELISA method IMUNOCLONE CRP ELISA (American Diagnostica, Greenwich, CT). TAFI antigen levels in plasma were analyzed with an IMUCLONE TAFI ELISA kit (American Diagnostica, Greenwich, CT). TAFI functional activity was assayed with an ACTICHROME TAFI activity kit. The measured levels of nitric oxide, CRP, TAFI antigen, and TAFI functional were 37.36+/-36.8 (normal value, 37.49+/-18.96; range, 19.3-102 microM), 12.27+/-10.6 (normal value, < 1 microg/mL), 146.9+/-28.4% NHP (normal, 100% NHP), and 102.55+/-37% NHP (normal range, 22.3-165.7; mean, 89.5% NHP), respectively. The erythropoietin antibody was detected in 9.4% of the patient group. While 20% of the erythropoietin antibody-positive and 27.1% of the erythropoietin antibody-negative patients experienced chest pain, thrombotic events developed in 9.4% of the erythropoietin antibody-negative patients. These data provide the rationale for a novel mechanism of thrombosis through increased activity of CRP, nitric oxide, and TAFI, leading to fibrinolytic deficit and thrombosis in patients treated with erythropoietin.

Acute phase mediators modulate thrombin-activable fibrinolysis inhibitor (TAFI) gene expression in HepG2 cells.

Thrombin-activable fibrinolysis inhibitor (TAFI) has recently been identified as a positive acute phase protein in mice, an observation that may have important implications for the interaction of the coagulation, fibrinolytic, and inflammatory systems. Activated TAFI (TAFIa) inhibits fibrinolysis by removing the carboxyl-terminal lysines from partially degraded fibrin that are important for maximally efficient plasminogen activation. In addition, TAFIa has been shown to be capable of removing the carboxyl-terminal arginine residues from the anaphylatoxins and bradykinin, thus implying a role for the TAFI pathway in the vascular responses to inflammation. In the current study, we investigated the ability of acute phase mediators to modulate human TAFI gene expression in cultured human hepatoma (HepG2) cells. Surprisingly, we found that treatment of HepG2 cells with a combination of interleukin (IL)-1 and IL-6 suppressed endogenous TAFI mRNA abundance in HepG2 cells (~60% decrease), while treatment with IL-1 or IL-6 alone had no effect. Treatment with IL-1 and/or IL-6 had no effect on TAFI promoter activity as measured using a luciferase reporter plasmid containing the human TAFI 5'-flanking region, whereas treatment with IL-1 and IL-6 in combination, but not alone, decreased the stability of the endogenous TAFI mRNA. Treatment with the synthetic glucocorticoid dexamethasone resulted in a 2-fold increase of both TAFI mRNA levels and promoter activity. We identified a functional glucocorticoid response element (GRE) in the human TAFI promoter between nucleotides 92 and 78. The GRE was capable of binding the glucocorticoid receptor, as assessed by gel mobility shift assays, and mutation of this element markedly decreased the ability of the TAFI promoter to be activated by dexamethasone.

Different mechanisms contribute to the biphasic pattern of carboxypeptidase U (TAFIa) generation during in vitro clot lysis in human plasma.

Carboxypeptidase U (CPU,TAFIa) recently gained interest as a significant player in dampening the fibrinolytic rate. The aim of this study was to investigate the time course of the generation of CPU activity during coagulation and fibrinolysis using an in vitro clot lysis model in human plasma. A first peak of CPU activity appeared after initiation of the coagulation phase and a second rise in CPU activity was observed during the fibrinolysis. The decrease in the proCPU plasma concentration followed the same trend as the appearance of the CPU activity. The direct thrombin inhibitor inogatran eliminated the CPU generation during coagulation but not during fibrinolysis. Addition of the plasmin inhibitor aprotinin during fibrinolysis resulted in a decrease in CPU activation during the lysis phase. These results demonstrate that proCPU was activated during coagulation by thrombin and during fibrinolysis by plasmin. Addition of a CPU inhibitor before initiation of clotting decreased the clot lysis time as expected. However, addition in the time period between the two peaks of CPU activity had no apparent effect on the clot lysis time.

Overall haemostatic potential can be used for estimation of thrombin-activatable fibrinolysis inhibitor-dependent fibrinolysis in vivo and for possible follow-up of recombinant factor VIIa treatment in patients with inhibitors to factor VIII.

Thrombin generation induced by recombinant factor VIIa (rFVIIa) in patients with haemophilia and/or inhibitors to factor VIII/IX could enhance generation of thrombin-activatable fibrinolysis inhibitor (TAFI), a recently described link between coagulation and fibrinolysis. TAFI is unstable and it is not easy to measure its active form in vivo. Overall haemostatic potential (OHP) is a novel method for haemostasis estimation, based on determination of the fibrin aggregation curve in which tiny amounts of thrombin are used for activation of clotting. We measured OHP in six patients with inhibitors to factor VIII before injection of rFVIIa and 10 and 120 min thereafter. Overall fibrinolytic potential (OFP) and clot lysis time (CLT) analysed by this method could be used for indirect estimation of TAFI generation. We found no change in pro-TAFI and total TAFI antigen before and after treatment with rFVIIa. OHP was almost undetectable before treatment but increased into the range of normal pooled plasma 10 and 120 min after rFVIIa treatment, as did CLT. However, after addition of potato tuber carboxypeptidase inhibitor, a specific inhibitor of TAFI, the shortening of CLT was lower than that in NPP. OFP was increased in patient plasma both 10 and 120 min after treatment compared with NPP. There was a strong positive correlation between pro-TAFI concentration and shortening of CLT after PTCI addition and a negative correlation between pro-TAFI concentration and OFP 10 min after rFVIIa injection. Thus, rFVIIa normalizes OHP and CLT 10 min after injection. While this improvement slightly decreases, but still exists after 2 hours, it suggests efficacy in bleeding prevention using a protocol based on rFVIIa administration every 2 hours.

Co-segregation of thrombophilic disorders in factor V Leiden carriers; the contributions of factor VIII, factor XI, thrombin activatable fibrinolysis inhibitor and lipoprotein(a) to the absolute risk of venous thromboembolism.

BACKGROUND AND OBJECTIVES: The clinical expression of factor V Leiden varies widely within and between families and only a minority of carriers will ever develop venous thromboembolism. Co-segregation of thrombophilic disorders is a possible explanation. Our aim was to assess the contributions of high levels of factor VIII:C, factor XI:C, thrombin activatable fibrinolysis inhibitor (TAFI) and lipoprotein (a) (Lp(a)) to the risk of venous thromboembolism in factor V Leiden carriers. DESIGN AND METHODS: Levels of the four proteins were measured, in addition to tests of deficiencies for antithrombin, protein C and protein S, and the prothrombin G20210A mutation, in 153 factor V Leiden carriers, derived from a family cohort study. The (adjusted) relative risk and absolute risk of venous thromboembolism for high levels of each protein were calculated. RESULTS: Of carriers, 60% had one or more concomitant thrombophilic disorders. Crude odds ratios (95% CI) of venous thromboembolism for high protein levels were: 3.2 (1.1-9.3) (factor VIII:C); 1.7 (0.6-4.9) (factor XI:C); 3.0 (1.1-8.2) (TAFI); and 1.9 (0.7-5.7) (Lp(a)). Adjusted for age, sex, other concomitant thrombophilic disorders and exogenous risk factors, the odds ratio for venous thromboembolism were 2.7 (0.8-8.7) for high factor VIII:C levels and 1.8 (0.6-5.3) for high TAFI levels. Annual incidences in subgroups of carriers were 0.35% (0.09-0.89), 0.44% (0.05-1.57) and 0.94% (0.35-2.05) for concomitance of high levels of factor VIII:C, TAFI and both, respectively, as compared to 0.09% (0.00-0.48) in single factor V Leiden carriers and 1.11% (0.30-2.82) for other concomitant disorders. INTERPRETATION AND CONCLUSIONS: High levels of factor VIII:C and TAFI, in contrast with factor XI:C and Lp(a), are mild risk factors for venous thromboembolism, and substantially contribute to the risk of venous thromboembolism in factor V Leiden carriers. Our data support the hypothesis that the clinical expression of factor V Leiden depends on co-segregation of thrombophilic disorders.