Physiochemical characterization of the BsAbs
Biochemical and biophysical properties have been shown to have an influence on mAb clearance in vivo .12, 19-23For the BsAbs, we evaluated these factors, which included solubility, isoelectric point (pI), thermal stability (Tm), as well as, the propensity for electrostatic (ie charge) and hydrophobicity-related interactions using heparin and hydrophobic interaction chromatography, respectively. The data from these analyses are summarized in Table 1.
Both BsAbs were readily concentrated to ~45 mg/mL in PBS buffer without any visible precipitation. To further compare the aggregation propensity of the two BsAbs, the samples were concentrated to 20 mg/mL in PBS buffer and held at 5°C for 10 days, when turbidity was measured by a micro-turbidity assay. Negligible differences were found between the two samples and with neither showing high turbidity, indicating comparable aggregation propensity at this condition. The samples were further stressed by incubating at 37°C for 7 days. Again, both samples exhibited low turbidity with negligible differences between them.
The experimental pI values for BsAb-1 and BsAb-2 were found to be 8.69 and 8.53, respectively. These pI values indicated that at physiological pH both BsAb molecules would be expected to have a similar weak overall positive charge. Additionally, the data indicated that the Fab/scFv conversion, or the ligand binding orientation switch between the two molecules did not grossly alter the pI (Table 1).
A previously developed heparin-based column assay was used to determine the degree of charge-based interaction for the BsAbs.11 In this experiment, BsAb-1 and BsAb-2 were injected over a column of heparin sepharose and then eluted with a linear gradient of increasing ionic strength. Retention of the molecule and elution time was then used to determine the heparin interaction potential (HpnIP) for charge-based interactions. Due to the small positive charge present on the BsAbs, they are somewhat retained by the column, eluting at [NaCl] of 218.1 mM and 191.1 mM, respectively, but the molecules displayed similar HpnIP (24.8%HpnIP versus 22.0%HpnIP) (Table 1).
Non-specific interactions of the BsAbs driven by hydrophobic association were evaluated using a HIC-based HPLC assay in which molecules were injected onto a solid phase hydrophobic resin pre-equilibrated in high concentrations of salt. Neither BsAb was retained on the HIC column indicating that both molecules are very hydrophilic (Table 1).
Thermal stability of the BsAbs were measured by differential scanning calorimetry (DSC). Analysis of BsAb-1 and BsAb-2 indicated that the first Tm value of BsAb-1 is much lower than that of BsAb-2 (59.0 oC versus 67.7 oC, respectively) (Table 1 and Supplemental Figure 1). In comparing DSC thermograms of BsAb-1 vs the matching IgG in BsAb-1, it is evident that the low Tm peak of BsAb-1 is due to the scFv portion of the molecule and that fusing the scFv domain to the C-terminal end of the HC showed no gross changes or perturbations in the thermal stability of the IgG portions (data not shown). This is consistent with our experience with other IgG-scFv fusion molecules that the IgG and scFv portions fold independently.