Abstract
Sickle cell disease (SCD) is associated with chronic activation of coagulation. Previously, we demonstrated that inhibition of tissue factor (TF) attenuates thrombin generation (measured by plasma levels of thrombin-antithrombin complexes [TAT]) in a mouse model of SCD during steady state. Furthermore, we showed that neither inhibition of FXIIa-dependent activation of FXI (using 14E11 antibody) nor FXI deficiency reduces thrombin generation (TG) in sickle mice. In contrast, genetic deficiency of FXII or kininogen (HK) reduced plasma TAT levels. These data suggest that during steady state, FXIIa contributes to TG in sickle mice via activation of the kallikrein/HK pathway, but not FXI. In the present study, we further investigated the mechanisms of HK-induced TG at steady state, and increased TG observed during vaso-occlusive crisis (VOC). All experiments were performed using 4-5 month old Townes SS (sickle) and AA (control) mice.
Kallikrein cleaves HK into HK fragments (HKFs) and bradykinin (BK). First, we investigated whether a BK-mediated increase in vascular permeability contributes to TG by exposing perivascular TF. This hypothesis was disproved by data demonstrating no difference in vascular permeability (measured by the extravasation of Evans blue in the heart, lung, liver and kidney) between AA (n=8) and SS (n=10) mice. HKFs were shown to induce leukocyte TF expression in vitro via binding to CD11b/CD18 (Mac-1). Therefore, we investigated whether Mac-1 inhibition affects TG in SS mice. AA and SS mice were treated with an inhibitory anti Mac-1 (M1/70) or IgG control antibody on days 0, 3 and 6 (i.p. 1 mg/kg) and TG was analyzed 1 day after the last injection. In the control group, SS mice demonstrated higher plasma TAT levels compared to AA mice (8.1±1.6 vs 4.2±0.6 ng/mL, n=10-11, p<0.05), but inhibition of Mac-1 significantly reduced plasma TAT levels in SS mice (4.6±0.7 ng/mL, n=11, p<0.05). These data suggest that HK might contribute to TG during steady state via Mac-1-dependent induction of monocyte TF.
The steady state of SCD is interspersed with acute periods of VOC. Clinical data demonstrate that compared to the steady state, plasma levels of cell free DNA (cfDNA), activation of the contact system, and TG are further enhanced during VOC. To determine the mechanism of increased TG during VOC, we used the previously characterized mouse model of TNFα -induced VOC. Townes AA and SS mice were injected with recombinant TNFα (2 µg/g body weight) or the same volume of PBS, and plasma was collected 5 hours later. TNFα not only dramatically increased plasma levels of cfDNA in SS mice (14.78 ± 1.64 vs 679 ± 300 ng/mL; p<0.01), but also further increased plasma TAT levels compared to those observed in PBS-treated SS mice (2.9 fold, p<0.001, n=8). Importantly, there was a significant positive correlation between cfDNA and TAT in SS mice (r2 =0.65, p<0.001).
Since cfDNA can activate FXII, we determined whether FXIIa-dependent activation of FXI contributes to TG during VOC. AA and SS mice received 14E11 or IgG control (4 mg/kg) 30 minutes before TNFα (2 μg/g) or PBS injection, and plasma TAT was assessed 5 hours later. Strikingly, 14E11 attenuated the increased TAT level in TNFα-treated SS mice, to the level observed in SS mice injected with PBS and IgG (IgG/SS/PBS: 9 ng/mL ± 1.8 vs. IgG/SS/TNF: 18.9 ± 3.6, p<0.001; 14E11/SS/TNF: 9.86 ± 0.72, p<0.05 vs. IgG/SS/TNF). We also determined if TF activity is required for the increased TG observed during VOC. Interestingly, inhibition of TF with an inhibitory 1H1 antibody (25 or 75 mg/kg injected i.p. 1 or 18 hours prior to TNFα, respectively) had no effect on the increased TG observed in TNFα treated SS mice.
In aggregate, our data suggest that during the steady state of SCD, FXII-dependent TG is not FXI-dependent, but instead is mediated by a pathway involving HK, Mac-1 integrin and leukocyte TF. Furthermore, we propose that during VOC the massive release of cfDNA results in FXIIa-dependent FXI activation and enhances TG independently of TF. This study provides mechanistic insight into the initiators of TG in SCD. Moreover, it implicates FXIIa as a potential therapeutic target to reduce the prothrombotic state in SCD, during both steady state and VOC.
Disclosures
No relevant conflicts of interest to declare.
Plasma levels do not predict thrombin generation in patients taking direct oral anticoagulants