Background: Therapy for hemophilia A is replacement of FVIII. Up to 40% of patients who receive FVIII develop inhibitory antibodies to FVIII. Likewise, dogs with hemophilia A develop inhibitory antibodies to canine FVIII providing a model to identify mechanisms of inhibitor formation and to develop and test novel approaches to prevent and reverse inhibitor formation. However these inhibitor dogs, like humans with inhibitors, need effective hemostatic therapy. Emicizumab does not bind effectively to canine FIXa or FX.

Aims: To develop a bispecific monoclonal antibody that will bind canine coagulation FIXa and FX and prevent bleeding in hemophilia A dogs with and without inhibitory antibodies to canine FVIII.

Methods: Starting with the bispecific antibody emicizumab that binds human FIXa and FX, we assayed its ability to enhance FIXa activation of FX. To create a canine therapeutic, the antigen binding regions of emicizumab were transferred to canine IgG4 constant regions. To create a bifunctional antibody on this canine IgG4 framework (k-IX-X), positive charges were introduced on the FIXa binding arm (L323K and L325K) and negative charges on the FX binding arm (K237E, L323E, and L325E). Residues in complementarity determining regions were mutated and activity of the mutations was assessed by activity assay.

Results: Emicizumab activity with individual factors was probed by mixing human and canine factors. Binding is reduced ~10-fold each for canine FIXa and FX (Figure 1). Moving the antigen binding regions of emicizumab to the k-IX-X framework did not change the activity. Selected mutations in k-IX-X increased the activity relative to emicizumab (Figure 2).

Conclusion(s): Improvements in binding to canine factors by introducing mutations into the antigen binding regions could enhance the activity of the canine bispecific monoclonal antibody into the range where it is a safe therapeutic product that could reduce bleeding events in hemophilia A dogs.

Image

Figure 1. Assay conditions: varied emicizumab and 2 nM FIXa, 140 nM FX, 20 µM lipid, 5 mM calcium, and 500 µM Pefachrome FXa. Absorbance vs time curves were analyzed as rate of factor Xa generation. h – human, c – canine, IXa – factor IXa, X – factor X. Curve fits are to a single site ligand binding model.

Image

Figure 2. Assay conditions: varied k-IX-X and 2 nM canine FIXa, 140 nM canine FX, 20 µM lipid, 5 mM calcium, and 500 µM Pefachrome FXa. Changed residues in the FIXa binding arm are indicated. Absorbance vs time curves were analyzed as rate of factor Xa generation.

« Back to ISTH 2022 Congress

ISTH Congress Abstracts - https://abstracts.isth.org/abstract/production-of-a-bispecific-monoclonal-antibody-that-enhances-canine-fixa-activation-of-canine-fx/