Resolution of Structurally Similar Glycoproteins by Affinity-BasedReversed Micellar Extraction and Separation

Abstract

Affinity-based reversed micellar extraction and separation (ARMES) has proven effective in separating glycoproteins from nonglycosylated proteins from natural sources. The ability of ARMES to resolve closely related glycoproteins is of paramount importance if ARMES is to be used in glycoform resolution. It is demonstrated that ARMES can resolve the structurally similar soybean peroxidase (SBP; MW 37 kDa, pI 4.1) and αt-acid glycoprotein (AGP; MW 43 kDa, pI 3.7), both of which have affinity for Concanavalin A (Con A) (the affinity ligand). SBP was almost exclusively extracted at pH 8 and above, with a separation factor greater than 50 (resolution ∼ 20), far better than was possible using Con A affinity chromatography (R ∼ 0.25, separation factor ∼ 2). Model calculations suggest that differences in affinity measured by an equilibrium-building assay cannot account for the favorable extraction of SBP over AGP at higher pH. Hydrophobic interactions and/or charge shielding appear to affect partitioning of the lectin - glycoprotein complexes and add greatly to the selectivity of extraction in ARMES, especially at higher pH values.