Identification and molecular cloning of a novel glycoside hydrolase family of core 1 type O-glycan-specific endo-α-N-acetylgalactosaminidase from Bifidobacterium longum

We found endo-alpha-N-acetylgalactosaminidase in most bifidobacterial strains, which are predominant bacteria in the human colon. This enzyme catalyzes the liberation of galactosyl beta 1,3-N-acetyl-D-galactosamine (Gal beta 1,3GalNAc) alpha-linked to serine or threonine residues from mucin-type glycoproteins. The gene (engBF) encoding the enzyme has been cloned from Bifidobacterium longum JCM 1217. The protein consisted of 1,966 amino acid residues, and the central domain ( 590 - 1381 amino acid residues) exhibited 31 - 53% identity to hypothetical proteins of several bacteria including Clostridium perfringens and Streptococcus pneumoniae. The recombinant protein expressed in Escherichia coli liberated Gal beta 1,3GalNAc disaccharide from Gal beta 1,3GalNAc alpha 1pNP and asialofetuin, but did not release GalNAc, Gal beta 1,3(GlcNAc beta 1,6) GalNAc, GlcNAc beta 1,3GalNAc, and Gal beta 1,3GlcNAc from each p-nitrophenyl (pNP) substrate, and also did not release sialo-oligosaccharides from fetuin, indicating its strict substrate specificity for the Core 1-type structure. The stereochemical course of hydrolysis was determined by H-1 NMR and was found to be retention. Site-directed mutagenesis of a total of 22 conserved Asp and Glu residues suggested that Asp-682 and Asp-789 are critical residues for the catalytic activity of the enzyme. The enzyme also exhibited transglycosylation activity toward various mono- and disaccharides and 1-alkanols, demonstrating its potential to synthesize neoglycoconjugates. This is the first report for the isolation of a gene encoding endo-alpha-N-acetylgalactosaminidase from any organisms and for the establishment of a new glycoside hydrolase family (GH family 101).