PEGylation or Polysialylation Reduces FVIII Binding to LRP Resulting in Prolonged Half-Life in Murine Models.

The low-density lipoprotein (LDL)-receptor-related protein 1 (LRP1) binds FVIII and is involved in the in vivo clearance and regulation of FVIII plasma levels. The significance of LRP in the regulation of FVIII metabolism has been demonstrated both in conditional LRP1 knock-out mice resulting in increased endogenous levels and prolonged survival of FVIII and by blocking LRP in hemophilic mice leading to prolonged half-life of infused human FVIII. In the current study we investigated whether random chemical modification of lysine residues of recombinant FVIII with branched polyethylene glycol (PEG) or polysialic acid (PSA) polymers interferes with LRP interaction. The modified FVIII variants, both of which gave rise to an increased half-life in mouse PK studies, were compared with rFVIII for binding to LRP1 by surface plasmon resonance (SPR) methodology. The typical binding of FVIII to LRP1 was more reduced by polysialylation than by PEGylation. To further delineate the region within the binding domain of LRP1 that recognizes FVIII, the isolated clusters II and IV were analyzed for FVIII interaction in an ELISA-based assay. Interestingly, FVIII bound to both clusters with comparable affinity, indicating that the CR of cluster II and IV have the same ligand-binding specificity. The PEGylated FVIII showed a reduced binding to both immobilized clusters, but also for the modified protein there was no clear preference for one of the two domains. The specificity of FVIII binding to cluster II was further analyzed by competition experiments. Increasing concentrations of cluster II diminished the signal for both FVIII and the PEGylated variant with an IC₅₀ of about 1 mM. Also heparin had a weak competitive effect, suggesting that this inherently anticoagulant compound could contribute to increase the half-life of FVIII. While the LRP1 binding site in the light chain of FVIII is accessible also in non-activated FVIII, that within the heavy chain of FVIII is only exposed after its activation by thrombin. It was therefore additionally tested whether limited thrombin treatment would increase the affinity of modified FVIII for LRP1. While a 40% increase in binding of rFVIII was observed, this was not the case for modified FVIII, indicating that the attached polymers also interfere with this interaction site of FVIII. The presented data suggest that drastically reduced binding to LRP1 contributes to the increase in half-life of modified FVIII. Because polysialylation and PEGylation both involve surface-exposed lysine residues, the findings further support the view that lysine-containing surface loops in FVIII are key elements in the interaction with its clearance receptor LRP. The in vivo significance of inhibiting the FVIII receptor interaction was demonstrated in VWF-FVIII double knock-out mice. While FVIII, co-administered with VWF, had a half-life of 2.1 h, infusion of PEG-rFVIII alone in these animals lacking both VWF and FVIII led to a dramatic increase of FVIII half-life to 8.3 h. The prolonged half-life of PEG-rFVIII in the absence of VWF suggests that PEG-rFVIII interferes with clearance mechanisms and may substitute for the necessary VWF-FVIII complex formation and several of the physiological VWF-FVIII protective mechanisms. Hence, PEG-rFVIII has potential as the treatment of choice in patients with type Normandy von Willebrand’s disease.