Heme Induces Systemic Activation of Coagulation in Vivo in a Tissue Factor-Dependent Manner

Abstract 820

Heme is released from red blood cells and hemoproteins during several pathological states, including bacterial infections, ischemia-reperfusion and hemolytic anemias. An excess of free heme is toxic, causing oxidative stress that activates endothelial cells and the inflammatory response. Heme has been reported to increase the expression of tissue factor (TF) on endothelial cells in vitro. TF is the primary initiator of the coagulation cascade. We tested the hypothesis that heme induces TF-dependent activation of coagulation in vivo. Eight week old C57 Bl/6 mice were given a bolus injection (0.035ml/g i.v.) of hemin ranging from 0 – 35 μmols/kg, and plasma was collected 6 hours after treatment. Coagulation activation was assessed by measuring plasma thrombin-antithrombin (TAT) levels. Administration of 17.5 and 35 μmols/kg of hemin significantly increased TAT from 5.3 ± 0.7 ng/L to 9.7 ± 1.3 ng/L (p<0.001) and 13.8 ± 1.9 ng/L (p<0.0001), respectively. Mice were treated with 35 μmols/kg of hemin in all further experiments. In a time course study, plasma was collected after 1, 3 and 6 hrs. Plasma TAT levels were significantly increased as early as 1 hour after treatment (14.1 ± 2.4 ng/L compared to 6.46 ± 0.6 ng/L, p<0.0001), and remained elevated at 3 (12.5 ± 0.6 ng/L, p<0.0001) and 6 hours (14.5 ± 1.0 ng/L, p<0.0001). To determine the role of TF in hemin-induced activation of coagulation, mice were pretreated with the rat anti-mouse TF antibody (1H1) or IgG control (single i.p. injection; 20mg/kg). Hemin significantly increased plasma TAT levels in IgG treated mice (4.6 ± 1.3 ng/Lto 9.6 ± 1.3 ng/L, p<0.0001). Importantly, inhibition of TF with 1H1 significantly attenuated this increase (6.8 ± 0.4 ng/L, p<0.01 compared to IgG/ hemin group). Moreover, plasma TAT levels were also significantly reduced in hemin injected low TF mice (expressing 1–2% of normal TF levels) compared to control mice injected with hemin (p<0.01). These data indicate that heme-induced activation of coagulation is TF-dependent. Next, we used TF ^flox/flox, Tie-2 Cre mice to determine the effect of TF gene deletion in all hematopoietic and endothelial cells on hemin-induced activation of coagulation. Interestingly, plasma TAT levels in TF ^flox/flox, Tie-2 Cre mice (11.9 ± 2.7 ng/L) were not different from that observed in control TF ^flox/flox mice 6 hours after hemin injection (11.6 ± 3.2 ng/L), indicating that TF expressed by other cell types contributes to the activation of coagulation. It has been previously shown that heme induces vascular damage and increases vascular permeability which could expose perivascular TF. Therefore perivascular cells are the likely source of TF that may contribute to the activation of coagulation in hemin treated mice.

In addition to the activation of coagulation, hemin (35 μmols/kg) increased total number of white cells (2.3 fold, p = 0.000032), monocytes (1.75, p=0.07) and neutrophils (3.5 fold, p=0.0002) as well as reduced platelet counts (0.86 fold, p=0.003) 6 hours after injection. Furthermore, plasma levels of the proinflamatory cytokine interleukin-6 were also increased (8.5-fold, p=0.002). However, inhibition of TF with 1H1 antibody did not affect any of these parameters.

Our data indicate that heme-induced activation of coagulation is TF-dependent whereas heme-induced inflammation is independent of TF.