Tumor necrosis factor-α (TNF-α) plays a critical role in the pathogenesis of vascular injury in diabetic patients (1). Increased circulating levels of TNF-α have been reported in diabetic patients (2,3). Hyperglycemia stimulates TNF-α secreted from monocytes and endothelial cells (4,5). Moreover, TNF-α may cause vascular injury by affecting the balance between coagulation and fibrinolysis. For example, TNF-α stimulates the expression of tissue factor that is the initiator of blood coagulation activation and the secretion of plasminogen activator inhibitor-1 that inhibits fibrinolysis (1).
Activated protein C (APC) is a serine protease that inhibits activation of the blood coagulation system by proteolytically inactivating factors Va and VIIIa and by stimulating fibrinolysis (6,7). APC may indirectly promote fibrinolysis by inhibiting thrombin generation and by inhibiting the action of plasminogen activator inhibitor-1 (8). Recently, it was reported that APC has an anti-inflammatory effect and that it inhibits vascular injury induced by TNF-α (9,10). Therefore, APC may inhibit hypercoagulability and inflammatory response induced by TNF-α, which increases during hyperglycemia. However, the relationship between circulating TNF-α and APC levels has not yet been reported in diabetic patients. In the present study, we investigated the relationship between the plasma levels of TNF-α and APC in normotensive type 2 diabetic patients.
Twenty-four normotensive (<140/90 mmHg) nonobese type 2 diabetic patients (16 men and 8 women, aged 53.0 ± 2.0 years [mean ± SE], BMI 23.1 ± 0.5 kg/m2, diabetes duration 6.7 ± 1.1 years, systolic blood pressure 131.0 ± 2.1 mmHg, diastolic blood pressure 78.1 ± 1.9 mmHg, fasting blood glucose levels 7.8 ± 0.32 mmol/l, and HbA1c 8.9 ± 0.3%) enrolled in the present study. All patients had normoalbuminuria and were being treated with diet therapy alone. Data obtained in 18 age-matched nonobese healthy subjects (14 men and 4 women) were used as control. The plasma levels of TNF-α were measured with an enzyme immunosorbent assay (EIA) kit (TNF-α EIA kit; Biosourse International, Camarillo, TX). As a marker of APC generation, APC-protein C inhibitor (PCI) complex was measured by enzyme-linked immunoassay as previously described (11). Protein C (PC) antigen was measured by solid-phase immunoassay as previously described (11). Total protein S (PS), a cofactor for activation of PC, was measured as previously reported (11). As a marker of coagulation activation, the plasma levels of thrombin-antithrombin complex (TAT) were measured by an EIA using anti-human monoclonal TAT antibody. As a marker of fibrinolysis, d-dimer was measured with a commercial EIA kit (d-dimer test F; Kokusai Shiyaku, Kobe, Japan). The plasma levels of TNF-α were significantly higher (1.60 ± 0.13 vs. 0.81 ± 0.32 pg/ml, P < 0.05) in diabetic patients than in normal subjects. The plasma levels of APC-PCI were significantly higher (4.63 ± 0.38 vs. 2.58 ± 0.60 pmol/l, P < 0.005) in diabetic patients than in normal subjects. There was a significant and positive correlation between the plasma levels of TNF-α and TAT in diabetic patients (r = 0.46, P < 0.05). There was a significant and inverted correlation between the plasma levels of TNF-α and d-dimer in diabetic patients (r = −0.52, P < 0.01). The plasma levels of TNF-α were positively and significantly correlated with the plasma levels of APC-PCI (r = 0.42, P < 0.05) in diabetic patients. There was no significant correlation between TNF-α and PC antigen (r = 0.33) and PS levels (r = 0.07).
The elevation of TNF-α in diabetic patients observed in our study is in agreement with previous data (2,3). Proportional correlation of TNF-α with TAT and inverse correlation with d-dimer suggests the occurrence of hypercoagulability and hypofibrinolysis in association with TNF-α in diabetic patients. Interestingly, the circulating levels of TNF-α were significantly correlated with APC-PCI complex, a marker of APC generation. These data suggest that APC may regulate fibrinolysis and hypercoagulability induced by TNF-α. It was reported that APC reduces vascular injury and hypercoagulability by inhibiting TNF-α production in rats treated with lipopolysaccaride (LPS) (9). APC may also inhibit production of TNF-α in LPS-treated culture cells (10). Another report (12) demonstrated that APC activates protease activated receptor-1 and induced gene expression of A20 (TNF-α induced protein 3) and tristetraprolin. A20 is cytoplasmic zinc finger protein that inhibits TNF-α-induced nuclear factor κB activity (13). Tristetraprolin inhibits TNF-α production by destabilizing its messenger RNA (14). Therefore, these mechanisms may be involved in the protective effect of APC in diabetic patients. However, the significant increase of TNF-α in diabetic patients suggests that APC may not be sufficient for regulating TNF-α expression. In brief, PC activation may be important for the regulation of TNF-α-induced coagulation abnormalities and inflammation in diabetes.