Quantitation of protein binding to the capillary wall in acidic, isoelectric buffers and means for minimizing the phenomenon

Notwithstanding the use of acidic, amphoteric, isoelectric buffers with isoelectric points (pI) in the pH 2-3 range, adsorption of proteins to the naked silica wall can be non-negligible. Two such buffers have been tested: iminodiacetic acid (IDA; pI 2.23, apparent pH 3.2 in 7 M urea) and aspartic acid (pI 2.77, apparent pH 3.7 in 7 M urea). Three potential quenchers of such interactions have been tested: hydroxyethylcellulose (HEC; number average molecular mass, M(r) 27 000), TEPA (tetraethylenepentamine) and a novel, quaternarized piperazine [N(methyl-N-ω-iodobutyl)-N'- methylpiperazine] (Q-Pip), either alone or in binary and ternary mixtures. Human α- and β-globin chains have been used as test proteins in capillary electrophoresis separations. It has been found that mixtures of these compounds are the worst possible remedy. E.g., a ternary mixture comprising 0.5% HEC, 0.5 mM TEPA and 1 mM Q-Pip still leaves behind 4.5% adsorbed protein onto the silica surface in runs in IDA buffer and 7 M urea (pH 3.2). Conversely, 0.5 mM TEPA or 1 mM Q-Pip, when used alone, minimize adsorption down to only 1.8% and 0.5%, respectively. When the same globin chain separations are performed in Asp and 7 M urea (pH 3.7), the situation is much worse: 44% protein is adsorbed in a ternary mixture of 0.5% HEC, 1 mM Q-Pip and 0.5 mM TEPA. However, when used alone, 0.5 mM TEPA and 1 mM Q-Pip reduce globin adsorption to levels of 8% and 5%, respectively. TEPA and Q-Pip are found to be in all cases the best quenchers of protein interaction to naked fused-silica; in addition they exhibit the unique property of smoothing the base-line and giving reproducible runs. The best method for desorbing bound protein was found to be an electrophoretic step consisting in driving sodium dodecylsulphate micelles from the cathodic reservoir. (C) 2000 Elsevier Science B.V.