Abstract
A review of past specific electrophoretic studies in a barbiturate buffer revealed that guanidinated proteins had reduced mobilities when compared to those of their native precursors. This phenomenon was ascribed to an increased affinity of the derivatives for anions. Since the derivatives, when compared to the native proteins, showed little change in affinities for common anionic matrices, the consideration arose that the electrophoretic changes discovered in each case were due to the binding of a specific anion. Barbiturate was considered to be the likely ion involved. Accordingly, the hypothesis was examined that the covalent linkage of a barbiturate to an insoluble matrix might provide a new chromatographic means of separating cationic proteins on the basis of an affinity for their content of guanidino groups. This proved to be the case; the technique was applicable to both homoarginine and arginine residues. It was found that anionic proteins also bound to the matrix; only however, at pH values below 7. This led to the consideration and confirmation that protonated imidazole groups of histidine residues also had an affinity for the matrix. Theoretical considerations and studies with model compounds suggest that this affinity of barbiturates for guanidino and imidazole groups rests on a combination of electrostatic and pi-electron interactions. Therefore, cationic proteins can be separated with this matrix at pH values of either 8 and higher, because of differences in arginine content, or below 7, because of differences in the sums of arginine and surface histidine content. In contrast, anionic proteins are probably separable largely on the latter basis only.A corollary study with a cationic matrix containing covalently linked guanidino groups demonstrated a particular affinity for the phenolate anions of tyrosine residues and could be applied at pH values of 8 or greater. This matrix, with pi-electrons, is also a simple ion-exchanger for carboxyl groups. The strengths of these interactions are such that not only anionic proteins but also cationic proteins can bind and be resolved on this cationic matrix.