Zein proteins are a complex mixture of polypetides that belong to the alcohol soluble storage proteins group (prolamines) in corn. These proteins constitute about 50-60% of the total endosperm protein and are classified in different groups on the basis of differences in their solubility and sequence. Among them, zein proteins are considered the majority group showing a high tendency to aggregate what makes very difficult their analysis by any analytical method. Thus, capillary zone electrophoresis (CZE) of these proteins require the use of very complex background electrolytes (BGEs) non-compatible with on-line ESI-MS analysis. The aim of this work was to find a new BGE for the CZE separation of zein protein fully compatible with ESI-MS while providing further light on the complex CZE separation of aggregatable proteins. Thus, it is demonstrated in this work that efficient and reproducible CZE separations of zein proteins can be achieved by using a BGE composed of water, acetonitrile, formic acid and ammonium hydroxide. Besides, it is shown that zeins analysis is significantly improved by including the effect of an ammonium gradient during their separation. It is experimentally verified that the ammonium gradient can easily be achieved in CZE by either working with a sample zone with a low concentration of ammonium and a background electrolyte (BGE) with a high concentration, or conversely, working with a sample zone with high ammonium concentration and a BGE with low concentration of ammonium, giving rise in both cases to a significant improvement in the CZE separation of these proteins. It is demonstrated that this procedure can give rise to efficiency improvements of up to 20-folds in the CZE separation of zein proteins. Under optimized conditions, 20 proteins could be separated with average efficiencies higher than 400000 theoretical plates/m. Some possible explanations of this effect are discussed including stacking, protein-capillary wall adsorption, protein solubility and protein-salt interactions.

Authors are grateful to the AGL2005-05320-C02-01 Project (Ministerio de Educacion y Ciencia) and the S-505/AGR-0153 Project (Comunidad Autonoma de Madrid) for financial support of this work.

Peer reviewed