Pre-Clinical Characterization and Analysis of Long Lasting Recombinant Factor IXFc

Abstract 4415

A recombinant Factor IXFc fusion molecule has been demonstrated to prolong half-life in several species (normal mice, rats, monkeys and FIX-deficient mice and dogs). The recombinant fusion molecule consists of a single rFIX moiety fused to the Fc region of Immunoglobulin G. A number of post-translational modifications exist on the rFIX portion of the fusion protein. These PTMs consist of γ-carboxylation, in the Gla domain, of the first 12 glutamic acids; the majority of which occupy glutamic acids 1–10 and partial occupation of glutamic acids 11–12. Other possible modifications include: hydroxylation of an aspartic acid residue, sulfation of a tyrosine residue and phosphorylation of a serine residue. In addition, there are several N- and O- linked glycosylation sites. One of the more important PTMs that may impact biological activity is γ-carboxy-glutamic acid (Gla) content thus several analytical methods were developed to monitor Gla content. Analytical methods include: amino acid analysis, AAA, (total Gla content), peptide mapping by HPLC-MS (Gla occupation), FIX coagulation activity by aPTT and analytical anion-exchange HPLC (iso-form separation). Material from the polishing ion-exchange step; purified peak and strip peak fractions, were analyzed. The average Gla content, measured by AAA and quantified on a molar basis relative to plasma derived FIX, of the purified peak fraction was 11.0 (3.9 % RSD) and, relative to plasma derived FIX which was fixed at 12.0 Gla. The average Gla content and biological activity of the strip peak fraction was reduced relative to the purified peak fraction. Gla occupancy was determined by digesting FIXFc with Lys-C and separating the peptides by HPLC with detection by a Mass Spectrometer (HPLC-MS). The primary peptides monitored were K1K2 and K3, each peptide has a maximum occupancy of 6 Gla's. The K1K2 peptide with 6 Gla was detected in the purified peak fraction, and 5 Gla was below detection whereas the strip peak fraction exhibited a reduced level of 6 Gla's and consequently K1K2 peptide with 5 Gla's was detected. The K3 peptide exhibits a distribution of Gla's ranging from 4 to 6, for both purified and strip peak fractions. An anion-exchange HPLC method was used to separate intact rFIXFc iso-forms. Purified peak fraction was injected onto an anion-exchange column and was separated into three species. Each species was collected and analyzed by aPTT, AAA and peptide mapping. Each species retained biological activity as well as Gla content ≥9.5. Peptide mapping indicated 6 Gla's in the K1K2 peptide (5 Gla's if present were below detection), while 4, 5 and 6 Gla's were detected in the K3 peptide. Strip peak fraction was also injected on to the anion-exchange column and two primary species were separated, each species was collected and analyzed as the purified peak fraction. No biological activity was measured for species 1, and species 2 had equivalent activity as the strip peak fraction (starting material) at 0.40 IU/nmol. Gla content by AAA indicated the strip peak fraction (starting material) to be reduced relative to the purified peak fraction; species 1 and 2 were determined to have a Gla content of 8.5 and 9.9, respectively. Peptide mapping of the strip peak fraction (starting material and species 1 and 2, derived from analytical anion-exchange HPLC), indicated peptides corresponding to K1K2 with 5 Gla's as well as 6 Gla's, and the K3 peptide exhibited 4, 5 and 6 Gla's. In summary, a number of analytical methods were developed and utilized to assess the Gla content of rFIXFc in purified peak fractions as well as the strip peak fraction. The data suggests the purified peak fraction is highly γ-carboxylated. Further fractionation of the rFIXFc iso-forms derived from the purified peak fraction was performed by analytical anion-exchange into several species. The separated species (species 1 and 2) retained biological activity with a reduction in Gla content, relative to the un-fractionated material and species 3. Analysis of the strip peak fraction exhibited a reduction of biological activity and Gla content, as well as an increased population of K1K2 5 Gla's. These analytical techniques were used to provide analytical as well characterization methodologies to support process development for designing a manufacturing process for production of clinical material.