Comparison between light induced and chemically induced oxidation of rhVEGF

The primary objective of this study was to compare the effects of light-and chemical-induced oxidation of recombinant human vascular endothelial growth factor (rhVEGF) and the impact of these reactions on protein formulation.

A liquid formulation of rhVEGF was exposed to fluorescent light (2 x 10(4) lux for up to 4 weeks), hydrogen peroxide (H2O2), or t-butythydroperoxide (t-BHP) to induce oxidation of rhVEGF. All samples were then treated by tryptic digest and analyzed by reversed phase HPLC to determine the extent of oxidation. Chemically treated samples were also examined by near-UV and far-UV circular dichroism spectroscopy to determine the effect of oxidation on the structure of the protein.

Exposure to light for 2 weeks resulted in 8 to 40% oxidation of all 6 methionine residues of rhVEGF (Met3 > Met18 > Met55 > Met78.81 > Met94). This amount of oxidation did not affect the binding activity of rhVEGF to its kinase domain receptor (KDR). Light exposure for 4 weeks increased metsulfoxide formation at Met3 and Met18 by an additional 16%, but did not affect the other residues. This oxidation decreased the receptor binding capacity to 73%. possibly due to the role of Met 18in receptor binding. Chemical oxidation of rhVEGF resulted in a greater extent of oxidation at all 6 methionines. Complete oxidation of Met3, Met18 and Met55 was observed after treatment with H2O2, while these residues underwent 40 to 60% oxidation after treatment with t-BHP. The receptor binding capacity was significantly reduced to 25% and 55% after treatment with H2O2 and t-BHP, respectively. After chemical oxidation, no changes in the secondary or tertiary structure were observed by far-UV and near-UV CD spectroscopy, respectively.

Methionine residues with exposed surface areas greater than 65 A2 and sulfur surface areas greater than 16 A2 were most susceptible to oxidation. Chemical oxidation resulted in higher metsulfoxide formation and decreased binding activity of the protein to KDR than light-induced oxidation. The reduction in KDR binding was not caused by measurable conformational changes in the protein. Photooxidation was dependent on the amount of energy imparted to the protein, while the ability of t-BHP or H2O2 to react with methionine was governed by solvent accessibility of the methionine residues and steric limitations of the oxidizing agent. Significant chemical oxidation occurred on sulfurs with minimum surface areas of 16 A2, while increased photooxidation occurred as a function of increasing surface areas of solvent exposed sulfur atoms. Such differences in the extent of oxidation should be considered during protein formulation since it may help predict potential oxidation problems.