AFFINITY PURIFICATION AND CHARACTERIZATION OF GLUCOSIDASE-II FROM PIG-LIVER

Glucosidase II has been purified from crude pig liver microsomes by a convenient procedure involving DEAE-Sephacel, Con A-Sepharose and affi nity chromatography on N-5-carboxypentyl-1-deoxynojirimycin-AH-Sepharo se. Specific binding of glucosidase II to the affinity matrix required its prior separation from glucosidase I, which was accomplished by fr actional Con A-Sepharose chromatography. The three-step procedure yiel ded, with approximate to 15% enzyme recovery, a > 190-fold enriched gl ucosidase II, consisting of two proteins (107 kDa and 112 kDa). Both p olypeptides are N-glycosylated with probably one glycan chain, in line with their binding to Con A-Sepharose. Immunological cross-reactivity and other experimental data indicate that the 107 kDa N-glycoprotein is derived from the 112 kDa species by partial proteolysis. The occasi onal presence of a 60 kDa peptide co-eluting with the catalytic activi ty suggests that glucosidase II may be associated with other protein s ubunit(s) in a heteromeric membrane complex. Glucosidase II hydrolyzes the alpha 1,3-glucosidic linkages in Glc(2-1)-Man(9)-GlcNac(2), as we ll as synthetic alpha-glucosides, efficiently but does not remove the distal alpha 1,2-linked glucose in Glc(3)-Man(9)-GlcNAc(2). The enzyme has a pH optimum close to 6.5 and is not metal ion-dependent. Catalyt ic activity is strongly inhibited by basic sugar analogues including 1 -deoxynojirimycin (dNM; app. K-i approximate to 7.0 mu M)I N-5-carboxy pentyl-dNM (app. K-i = 32 mu M) and castanospermine (app. K-i approxim ate to 40 mu M) Substitution of the 3-OH or 6-OH group in dNM by a flu oro group reduces the inhibitory potential drastically. We conclude fr om these observations that the two hydroxy groups are essential for in hibitor/substrate binding due to their ability to interfere as hydroge n bond donors. A polyclonal antibody raised against the 107 kDa polype ptide reacted specifically with two proteins from different cell types on Western blots. Their molecular masses were identical with those fr om pig liver microsomes, pointing to a highly conserved amino acid seq uence of glucosidase II. This suggests that the variance in molecular mass for glucosidase II reported for the enzyme from other tissues and species may be due to partial proteolysis.