Quantitative gel electrophoresis: sources of variation.

Gel electrophoresis is known for its often unsatisfactory precision. Percental relative standard deviations (RSD%) in a range of 15-70% have been reported. Therefore, an improvement of precision in quantitative 2-DE is necessary. In the present, study we have analyzed the work flow of 2-DE in detail to assess the main error sources. Potential major sources of variability for this technique include the transfer between first and second dimension, the analyst's expertise, and the staining or rather detection of separated proteins. The remarkable and completely irregular changes of the background signal from gel to gel were identified as one of the governing error sources. These background changes can be strongly reduced by the direct detection of the separated proteins using native fluorescence. More than a 3-fold better signal-to-noise ratio was found compared to Ruthenium-(II)-tris-(bathophenanthroline disulfonat) (RuBPS) and Coomassie staining, although the sample was used in an 800-fold lower concentration. This improvement together with well-defined peaks resulted in a better quantitative spot reproducibility of approximately 12-16% RSD%. Possibly, the variabilities due to detection and evaluation were already reduced to minor error components. However, according to the law of error propagation, the major error sources dominate the total error. To really prove the good detection and evaluation, these other sources of variability such as sample preparation, strip rehydration, protein loading, transfer between dimensions, interactions between gel and proteins, gel scanning, and spot integration have to be reduced next.