Hemophilia A is the most common of the severe inherited bleeding disorders and is caused by defects or deficiencies in the blood coagulation protein, factor VIII. Factor VIII circulates as an inactive procofactor and is converted to factor VIIIa by thrombin or factor Xa-catalyzed cleavage at Arg740 (A2-B junction), Arg372 (A1-A2 junction), and Arg1689 (a3-A3 junction). Factor VIII cleavage by thrombin is suggested to be an ordered mechanism, with initial cleavage at Arg740 facilitating cleavages at Arg372 and Arg1689 (

Newell and Fay,
J. Biol. Chem.
). The catalytic efficiency for thrombin cleavage at Arg740 is greater than at either Arg1689 or Arg372, while cleavage at Arg372 appears rate-limiting. For this reason, we investigated whether cleavage at Arg1689 influenced catalysis at the Arg372 site. We constructed, stably transfected, and expressed two recombinant factor VIII mutants, Arg1689His and Arg1689Gln, to slow and eliminate, respectively, cleavage at this site. The specific activity of the Arg1689His and Arg1689Gln factor VIII variants were reduced to 47% and 9% the WT factor VIII value, respectively. This result was consistent with impaired cleavage rates of the light chain impacting this parameter. SDS-PAGE and Western blotting analysis of the Arg1689His variant showed a ~150-fold reduction in the rate of cleavage at Arg1689, while the Arg1689Gln variant was resistant to cleavage at this site. Furthermore, the Arg1689His and Arg1689Gln variants showed 2-fold and 7-fold rate reductions, respectively, in the generation of the A2 subunit following thrombin-catalyzed cleavage at Arg372 as compared to WT factor VIII. This result indicated a linkage between cleavage of the light chain and the primary activating cleavage in the factor VIII heavy chain. In monitoring thrombin activation time courses, we observed a 2-fold reduction in the peak activation for the Arg1689His variant and a 12-fold reduction for Arg1689Gln variant compared to wild type factor VIII. To investigate thrombin binding to Arg1689Gln factor VIII, this variant was used as a competitor of thrombin activation of WT factor VIII. The apparent Ki for Arg1689Gln factor VIII (59 nM) was similar to the Km (apparent) of WT factor VIII (55 nM), suggesting that the Arg1689Gln mutation does not impair binding of thrombin to factor VIII. Taken together, these results suggest thrombin cleavage of factor VIII Arg1689 is preferred before cleavage at Arg372 but not required. Since factor Xa also cleaves factor VIII at the same sites as thrombin, we evaluated factor Xa cleavage of factor VIII using SDS-PAGE and Western blot analysis. These results showed that factor Xa cleavage at Arg372, as judged by A2 subunit generation, for both Arg1689 variants was ~2-fold slower than that for WT factor VIII. Moreover, factor Xa cleavage of light chain at the Arg1721 site was reduced by ~2–3 fold for the variants suggesting that factor Xa cleavage at the Arg1689 site is preferred over the Arg1721 site. Together, these results suggest a less selective mechanism for cleavage order using factor Xa as compared with thrombin.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author