Abstract
Formamidase [EC 3.5.1.9] from rat liver proved to be microheterogenous. After preparative isoelectric focusing in density gradient columns, 2 peaks of formamidase with identical substrate specificity were identified. By analytical focusing in thin layers of polyacrylamide or Sephadex G-75 SF, even 5 bands could be separated. Their isoelectric points were 4.75, 4.78, 4.82, 4.92 (main band) and 5.11, but their Km did not differ significantly (54-62 .mu.mol/l). An identical MW of 34,700 .+-. 3200 for all bands was determined by disc electrophoresis. This value was confirmed by sedimentation analyses .**GRAPHIC**. = 3.00 S) and electrophoresis in the presence of sodium dodecylsulfate (MW = 34,900 .+-. 2300), which only gave a single band. The homogeneity was also confirmed by electrophoresis in the presence of 6 M urea. Repeated disc electrophoresis or focusing under native conditions with single, isolated formamidases again resulted in different bands which were identified, not only by Coomassie Blue, but also by their hydrolytic cleavage of naphthyl acetate. Formamidase showed neither proteolytic nor asparagine-amidohydrolase activity and oligosaccharide conjugates were not detectable. Ampholytes, buffer ions, pH and peroxodisulfate did not affect the heterogeneity. Initial burst measurements with diethyl(4-nitrophenyl)phosphate yielded an equivalent weight of 36,300. Formylkynurenine reduced this inhibition very effectively. Thus, an extraordinary reactive serine residue appeared to be located in the catalytic site of formamidase. A participation of sulfhydrylgroups in the inactivating reaction of arsenite was excluded although 2 such groups were detected by 5,5''-dithiobis(2-nitrobenzoic acid). N-bromosuccinimide reacted primarily with 1 of the 9 tryptophan residues without loss of enzymatic activity, but a 18.6-fold excess of this reagent resulted in a complete loss of activity.