Evaluation of the metal binding sites in a recombinant coagulation factor VIII identifies two sites with unique metal binding properties.

Coagulation factor VIII is a glycosylated, non-covalent heterodimer consisting of a heavy chain (A1-A2-B domains) and a light chain (A3-C1-C2 domains). The association of the chains, and the stability and function of the dimer depend on the presence of metal ions. We applied X-ray fluorescence, X-ray crystallographic structure determination with anomalous signals at different wavelengths, and colorimetric measurements to evaluate the metal binding sites in a recombinant factor VIII molecule, turoctocog alfa. We identified a metal binding site in domain A3 dominated by Cu(+) binding and a site in domain A1 dominated by Zn(2+) binding.

Investigating clearance mechanisms for recombinant activated factor VII in a perfused liver model.

Clearance mechanisms for recombinant activated human FVII (rFVIIa; NovoSeven), a heterogeneously glycosylated protein, have yet to be fully elucidated, but may involve the liver. The effects of the gamma-carboxy glutamic acid (Gla) domain and the sialic acid content of the protein on rFVIIa clearance were investigated following intravenous administration of rFVIIa lacking the Gla domain, des(1-44) rFVIIa and asialo-rFVIIa in pharmacokinetic (PK) studies and perfused rat livers. PK parameters for both rFVIIa and des(1-44) rFVIIa had similar biphasic clearance profiles, as well as half-lives ([t(1/2)]=80 and 88 minutes, respectively), while asialo-rFVIIa was cleared quickly (t(1/2)=21 minutes) with a linear clearance profile. Perfused liver studies with all proteins (10 nM) mirrored the trends in profiles observed in the PK study. rFVIIa and des(1-44) rFVIIa were cleared to a similar extent, 41% and 35%, respectively, after 1 h, whereas plasma-derived FVII from humans (which has a higher sialylation content than rFVIIa) was cleared to a lesser extent (21%). Asialo-rFVIIa, on the other hand, was almost totally cleared and when an excess of asialo-orosomucoid was added to the perfusate, its clearance was significantly reduced (by 34%) and also for rFVIIa, albeit to a lesser extent (by 14%). Together these data suggest that carbohydrate receptor(s) (e.g. the asialoglycoprotein receptor, ASGPR) play a role in asialo-rFVIIa and rFVIIa clearance. In vivo and liver clearance data correlated well showing similar trends and indicated that rFVIIa clearance is not affected by the Gla domain, but rather by a subpopulation of N-glycosylated structures on rFVIIa.

Evaluation of potential antigenicity of active-site-inhibited recombinant human FVIIa (FFR-rFVIIa) in an immune-tolerant rat model.

Recombinant human FVIIa (rFVIIa) was inactivated by coupling Phe-Phe-Arg-CK- (FFR) covalently to the active site of the enzyme. To test the chemically-modified human protein for potential antigenicity prior to clinical trial an immune-tolerant rat model was established. Intraperitoneal injection of the parent compound, human rFVIIa, within 30 h after birth, followed by repeated subcutaneous challenge with rFVIIa in Freunds incomplete adjuvant resulted in 79% non-responding rats at day 32. Monthly subcutaneous challenge showed that the induced tolerance was stable over the 3 months study period in 80% of the rats. The clinically relevant route, intravenous administration, was used for evaluating the potential antigenicity of FFR-rFVIIa. Repeated intravenous administration of different dosages of FFR-rFVIIa did not break tolerance, indicating that FFR-rFVIIa might not be antigenic, for a limited number of intravenous administrations in a clinical setting.