Mapping of solution components, pH changes, protein stability and the elimination of protein precipitation during freeze–thawing of fibroblast growth factor 20

This study discusses the effect of key factors like containers, buffers and the freeze (controlled vs. flash freezing) and thawing processes on the stability of a therapeutic protein fibroblast growth factor 20 (FGF-20). The freezing profiles monitored by 15 temperature probes located at different regions in a 2-L bottle during freezing can be grouped into three categories. A rapid drop in temperature was observed at the bottom followed by the top and middle center of the bottle. The freeze–thawing behavior in a 50 ml tube is considerably uniform, as expected. Among phosphate, HEPES (4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid), citrate and histidine (each containing 0.5 M arginine-sulfate) buffer systems, a minimum pH change (0.4 pH unit vs. ∼1.7 pH unit) was observed for the phosphate buffer system. Thawing in a 50 ml tube at room temperature standing resulted in a significant phase separation in citrate, histidine and HEPES buffers; however, phase separation was least in the phosphate buffer system. These phase separations were found to be temperature dependent. No effect of Polysorbate 80 on freeze–thawing of FGF-20 was observed. Significant concentration gradients in major buffer components and protein concentration were observed during freeze–thawing in a 2-L bottle. The segregation patterns of the various components were similar with the top and bottom layers containing lowest and highest concentrations, respectively. In the formulation buffer no pH gradient was formed, and the precipitation of FGF-20 during thawing at the top layer was related to an insufficient amount of arginine-sulfate and the precipitation at the bottom layer was due to a salting out effect. The precipitate generated during thawing goes into solution easily upon mixing whole solution of the bottle and the various gradient formations do not cause any irreversible change in structure, stability and isoform distribution of FGF-20. Comparison of slow freezing and flash freezing data suggests that the gradients in excipient and protein concentrations are mainly formed during thawing.