Glycosylation is a post-translation modification of proteins, which is associated with protein stability and function. Glycosylation patterns are different in different types of cells and some are unable to add glycosyl groups. For example, prokaryotic cells do not have the necessary machinery to add glycosyl groups to their proteins; therefore, human proteins expressed in bacteria are not glycosylated. Mammalian cells and plant cells can both glycosylate proteins but their glycosylation patterns are different. This is important because, recently, therapeutic human proteins have been expressed in plants. Although the expression of human antibodies in plants like tobacco or corn has been quite successful, the plant glycosylation machinery is often turned off because of differences in the glycosylation pattern between the human and the plant variety. Unfortunately, the non-glycosylated antibody is less stable and thus has a shorter shelf life than the native one. The decrease in stability is manifested by the formation of irreversible aggregates, which decreases the solubility of the preparation impeding their normal function.

The objective of this work is to investigate the effect of various stresses found during processing on the stability of fully glycosylated and non-glycosylated human antibodies. The stresses that we have chosen to investigate are: exposure to high temperatures, exposure to shear rate, exposure to freeze and thaw cycles and the effect of freeze drying. Experiments were performed in the presence and absence of various amounts of sucrose.

All the preparations were inspected by second derivative UV spectroscopy to detect coarse changes in the secondary structure. No major changes in protein structure were detected upon the applied stresses. Samples were run through a GPC column and monitor with a UV detector in tandem with a multiple angle light scattering detector. All the chromatograms show the presence of three species, the monomer (MW 130,000), the dimer (MW 260,000) and higher aggregates of molecular weights 1 million or higher. These relatively large aggregates consist of approximately ten antibody molecules. Assuming spherical shapes and closed packing (a density of 0.74), these aggregates are roughly 20 nm in diameter. The glycosylated antibody has a higher percent monomer than the non-glycosylated one (~92% vs. 88%). There are only minor differences between the ratio of the different species under various stresses with the exception of experiments performed at 50 C which seem to have less monomer.

Dynamic light scattering was used in this work to study the effect of various stresses on the formation of large aggregates (> 100 nm) of both antibody preparations. The experiments were performed with a backscattered instrument (Brookhaven) which is not very sensitive to the overall quality of the sample. Stirring (at either 20 C or 50 C) as well as freeze drying using an ethanol bath have the largest effects on both preparations. There is a minor effect by freeze and thaw (either one or two cycles) but a pronounced effect by freeze drying using both freezing methods (ethanol or liquid nitrogen). In the presence of sucrose large aggregates (larger than 100 nm) are absent in the control preparation but the protective effect of sucrose is more pronounced with the non-glycosylated preparation.

A few selected samples were inspected using a research goniometer equipped with a 2W laser and a digital correlator (Brookhaven). These experiments revealed details of the shape of the aggregates. Not only the non-glycosylated preparation has a larger number of aggregates but also they are more asymmetric than those of the glycosylated one. The presence of sucrose seems to have an “equalizer” effect for both preparations.