Prevention of Bleeding-Induced Vascular Abnormalities in the Hemophilic Mouse Joint By Increasing TAFI Levels or Inhibiting uPA Activity

Hemophilic arthropathy is characterized by vascular abnormalities in the joint soft tissue that may cause rebleeding and progressive joint destruction. Defective thrombin generation in hemophilia leads to impaired activation of Thrombin Activatable Fibrinolysis Inhibitor (TAFI). Activated TAFI (TAFIa) is a carboxypeptidase that cleaves C-terminal Lysine from fibrin or cell plasminogen receptors, thereby diminishing fibrinolysis or pericellular plasmin generation. In addition, TAFIa inactivates pro-inflammatory (i.e. C3a, C5a, bradykinin) and anti-angiogenic (i.e. VEGF-A, SDF-1α) substrates. We have recently shown that TAFI-gene deficiency causes vascular abnormalities in the mouse joints that phenocopy those in hemophilic arthropathy, either after joint bleeding or spontaneously with aging. Based on these mechanistic insights, we evaluated treatment with TAFI for its ability to mitigate bleeding-induced vascular abnormalities and associated bleeding susceptibility in the FVIII KO mouse joint.

Knee joint bleeding in FVIII KO mice was induced by a subpatellar puncture injury that was standardized to cause severe bleeding as defined by a hematocrit (Hct) drop from 47 ± 2 % at baseline (BL) to 29 ± 11 % two days after injury. The vascular abnormalities formed in the hemophilic mouse joint 2 weeks after injury capture those in patients with hemophilic arthropathy: i.e. increased vascular permeability and abnormal vessel enlargement.

Hydrodynamic gene delivery was used in mice for continuous treatment with human wild type TAFI (hTAFI-WT). To compensate for lower TAFI activation in hemophilia, two TAFI mutants were tested with extended enzymatic stability: hTAFI-CYQ (8-fold increased) and hTAFI-CIIYQ (180-fold increased). The TAFI constructs had comparable antigen expression profiles (reaching a similar level as in human plasma at day 2 and gradually returning to BL by week 2). Treatment efficacy to reduce vascular permeability, however, differed according to TAFIa's enzymatic stability in FVIII KO mice 2 weeks after injury. Without treatment, Evans blue-BSA extravasation in the joints was increased (1.8-fold compared to the contralateral joint; p= 0.0001 vs BL). hTAFI-WT overexpression initiated 2 days after injury had no effect, whereas hTAFI-CYQ and hTAFI-CIIYQ overexpression reduced the vascular permeability to 1.5-fold (n= 10; p= 0.10 vs untreated) and 1.3-fold (n= 8; p= 0.0032 vs untreated), respectively, indicating that hTAFI-CIIYQ was most effective. Histology at week 2 confirmed that hTAFI-CIIYQ reduced the number of excessively enlarged CD31⁺ vessels (with ≥ 20 µm diameter) by 3-fold (n= 6; p= 0.0011 vs untreated).

To determine whether the protective effect of TAFI overexpression on bleeding-induced vascular abnormalities was associated with inhibition of the plasminogen activation system, continuous administration of tranexamic acid (TXA; 50 mg/ml in drinking water) or antibody-mediated blockage of uPA's proteolytic activity (mU1; 60 mg/kg; 2x weekly) was initiated in FVIII KO mice 2 days after injury. Histology at week 2 showed mU1, but not TXA, reduced vessel enlargement to the same extent as hTAFI-CIIYQ.

Finally, the joint bleeding model was adopted to appreciate functional consequences of hypervascularity to changes in bleeding susceptibility. Aged TAFI KO mice (8 months) were used for their spontaneously (3-fold) increased number of enlarged vessels (≥ 40 µm diameter; p= 0.0076) in the joint that correlates with Power Doppler (PD) signal. A mild injury was optimized to cause minimal bleeding and showed that PD signals in joints of aged TAFI KO mice correlated negatively with the Hct 2 days after mild injury with injection of a FVIII inhibitor (r= -0.53, p= 0.024; n= 18), supporting hypervascularity as a major contributor to hemophilic joint bleeding. Furthermore, the same mild injury caused minimal bleeding in naive FVIII KO mice with low BL vascularity (median Hct 46 % vs 48 %; p= 0.81 vs BL; n= 7), but excessive bleeding in FVIII KO mice injured 2 weeks previously with vascular abnormalities (median Hct 38 %; p= 0.021 vs BL; n= 7). hTAFI-CIIYQ effectively reduced bleeding severity upon re-injury (median Hct 47 %; p> 0.99 vs BL; n=5).

Our data provide an experimental link between vascular abnormalities and bleeding susceptibility in hemophilic joints and conclude that the vascular abnormalities were alleviated by increasing TAFI levels or inhibiting uPA activity.