Abstract
Introduction: An important coagulation regulatory mechanism is localization of clotting complexes to exposed phosphatidylserine (PS) on cell surfaces. All components of the intrinsic tenase complex (factor (F)VIIIa, FIXa, and FX) bind to PS. FVIIIa mimetic bispecific antibodies are drugs in development for hemophilia A that aim to mimic the cofactor function of FVIIIa by bringing together FIXa and FX to generate FXa. However, these antibodies differ from FVIII in many ways including no requirement of activation and a lack of direct PS binding. Emicizumab is a bispecific antibody currently on the market for hemophilia A patients with inhibitors. It binds to factor FIX, FIXa, FX, and FXa with micromolar affinities in solution. Previously, we have shown that in-house preparations of sequence-identical emicizumab (SI-Emi) showed similar weak affinities to its antigens and similar in vitro activity to published emicizumab results by one-stage clotting, chromogenic FXa generation, and thrombin generation. However, in chromogenic FXa generation using antibody concentrations in the range of the mean steady state plasma concentration of patients on emicizumab prophylaxis [~360 nM, (Oldenburg, et al., NEJM 2017)], SI-Emi maintained 28% of its activity even in the absence of PS-containing phospholipid vesicles. Another FVIIIa mimetic antibody, BS-027125, was discovered by our group and binds with low nanomolar affinity to FIX, FIXa and FX, with no detectable binding to FXa. In one-stage clotting, BS-027125 achieved clot times similar to physiological levels of FVIII, but had poor activity in thrombin generation at these concentrations. Furthermore, it too maintained small amounts of phospholipid-independent activity in chromogenic FXa generation. Given the artificial nature of the chromogenic FXa generation assay, and that activity of prothrombinase is PS-dependent thereby precluding omission of phospholipids from thrombin generation assays, we developed an assay to detect FXa generation in a plasma milieu by FVIIIa mimetic antibodies or FVIII with and without phospholipid vesicles.
Methods: FVIIIa mimetic antibodies or recombinant FVIII (rFVIII) were incubated with thrombin for 5 minutes, quenched with hirudin, then spiked into platelet-free congenital hemophilia A plasma treated with additional hirudin. FXIa (to generate FIXa in situ) with and without PC:PE:PS (40:40:20 molar ratio) phospholipid vesicles was added and reactions were triggered with a solution of CaCl2 and fluorogenic FXa substrate (Mes-D-LGR-ANSN(C2H5)2). Substrate cleavage was monitored kinetically on a fluorescent plate reader. Substrate cleavage by FXIa could not be detected, yet another unknown plasma peptidase did cleave substrate at a constant low rate that was background subtracted.
Results: In the absence of phospholipid vesicles, SI-Emi maintained 51±3.7% of its FXa generation activity at all concentrations tested (3.8±0.4 versus 8.0±1.1 RFU/min at 333 nM). BS-027125 showed very low activity (0.43±0.12 RFU/min at 50 nM) in the presence of phospholipid vesicles, however, in the absence of phospholipid vesicles, BS-027125 activity was not detectable above baseline. Nearly all rFVIII activity (>99%) was lost in the absence of phospholipid vesicles (0.14±0.04 versus 15.1±1.8 RFU/min at 0.3 IU/mL). Addition of annexin V was sufficient to block all rFVIII activity in the presence or absence of phospholipid vesicles, but could not block SI-Emi activity. Furthermore, addition of rivaroxaban, a direct FXa inhibitor, confirmed that detection of substrate cleavage was due to FXa activity.
Conclusions: In the absence of phosphatidylserine-containing phospholipid vesicles, SI-Emi promoted the generation of FXa in plasma triggered with FXIa. The activity of BS-027125 was too low in this assay to clearly determine its phospholipid-independent activity. These results suggest SI-Emi has mis-regulated (PS-independent) procoagulant activity due to a lack of phospholipid localization of the antibody-FIXa-FX complex. Given the weak affinity of SI-Emi for its antigens, the exact mechanism enabling this activity is unclear. Further study of this phenomenon and its relevance to overall thrombin generation and in vivo activity are needed.
Aleman:Bioverativ, a Sanofi company: Employment. Jindal:Bioverativ, a Sanofi company: Employment. Leksa:Bioverativ a Sanofi company: Employment. Peters:Bioverativ a Sanofi company: Employment, Equity Ownership. Salas:Bioverativ a Sanofi company: Employment, Equity Ownership.
Author notes
Asterisk with author names denotes non-ASH members.