CGE, also known as SDS-CGE, is being established in the pharmaceutical industry replacing SDS-PAGE. In most cases, the method is applied for the identity and purity control of proteins, for example monoclonal antibodies. In order to quantify these components with sufficient precision using the same quality control method, a RSD for the quantitative analysis under 2% is required. A reliable and highly precise CGE method could be obtained after thorough optimization. It was crucial to increase the sample concentration and the injection volume in order to achieve sufficiently high S/N ratios (>70). The application of hydrodynamic injection is beneficial for the precision of the method compared to the traditionally used electrokinetic one. Linearity was demonstrated and LOD and LOQ were estimated. Both injection modes were compared in long series runs (n = 48). Furthermore, the use of an internal standard was investigated. Thus, the RSD% of the migration time was reduced from 0.9 to 0.2% and the RSD% of peak areas was greatly improved. However, the normalization to the total area further reduced the influence of the injection error. RSD% for the peak area ratios of typically between 1 and 2% was provided.

Single proteins separated by 2-DE often show multiple spots spreading along the first dimension. In many cases, such charge trains are explained by isoform differences or by putative post-translational modifications including phosphorylation, glycosylation and others. We now report that individual spots of such charge trains on 2-D gels in fact often represent the same protein, but, apparently due to conformational changes, segregate to different isoelectric points. If MS analysis reveals protein identity, we therefore suggest integrating all individual spots within a charge train for quantification. Especially in quality control of pharmaceutical proteins, the integration of the spot groups of all active contents is preferable in order to obtain reproducible and reasonable quantitative results. However, most commercial software packages for gel analysis integrate the signals spot-wise. We provide an improved quantification tool for proteins with charge train groups. This calculation can be implemented using the MATLAB software and the self-developed "Correct Integration Software System" or the commercial software package Delta2D.

In this study, size-exclusion chromatography and high-resolution atomic absorption spectrometry methods have been developed and evaluated to test the stability of proteins during sample pretreatment. This especially includes different storage conditions but also adsorption before or even during the chromatographic process. For the development of the size exclusion method, a Biosep S3000 5 μm column was used for investigating a series of representative model proteins, namely bovine serum albumin, ovalbumin, monoclonal immunoglobulin G antibody, and myoglobin. Ambient temperature storage was found to be harmful to all model proteins, whereas short-term storage up to 14 days could be done in an ordinary refrigerator. Freezing the protein solutions was always complicated and had to be evaluated for each protein in the corresponding solvent. To keep the proteins in their native state a gentle freezing temperature should be chosen, hence liquid nitrogen should be avoided. Furthermore, a high-resolution continuum source atomic absorption spectrometry method was developed to observe the adsorption of proteins on container material and chromatographic columns. Adsorption to any container led to a sample loss and lowered the recovery rates. During the pretreatment and high-performance size-exclusion chromatography, adsorption caused sample losses of up to 33%.

Human papillomavirus 6b L1 virus-like particles (VLPs) were successfully expressed using Bombyx mori nucleopolyhedrovirus (BmNPV) bacmid expression system and rapidly purified using size exclusion chromatography after ultracentrifugation procedure and characterized by capillary zone electrophoresis (CZE). The average capillary electrophoresis migration time was 11 min with the relative standard deviation (RSD) of 0.3% of human papillomavirus 6b L1 VLPs. After this threefold fractionation, the CZE samples were still further investigated by dynamic light scattering and immuno blotting. The versatile technique, CZE not only proved to be a valuable tool for VLP characterization, but was also found to be reliable and precise. Thus CZE will also be an important option for the quality control of VLPs in pharmaceutical research level.

No abstract