K. Ito et al., HUMAN SALIVARY ENDO-BETA-N-ACETYLGLUCOSAMINIDASE HS SPECIFIC FOR COMPLEX TYPE SUGAR CHAINS OF GLYCOPROTEINS, The Journal of biological chemistry, 268(21), 1993, pp. 16074-16081
The enzyme that catalyzed the conversion of human salivary alpha-amyla
se family A (HSA-A) to family B (HSA-B) was identified. It was partial
ly purified from the precipitate obtained by centrifugation of human s
aliva at 105,000 x g for 60 min by solubilization with 3[(3-cholamidop
ropyl)dimethylammonio]-1-propane sulfonate and column chromatographies
with Sephacryl S-300-HR and hydroxylapatite. The enzyme preparation w
as practically free from contaminating exoglycosidases and proteases.
The enzyme cleaved the N,N'-diacetylchitobiose moiety of the sugar cha
in of HSA-A, as shown by the isolation of the protein moiety which con
tained 1 GlcNAc and 1 Fuc residue and the sugar chain (Gal)2(Fuc)1(Glc
NAc)2(Man)3(GlcNAc). This enzyme also cleaved the N,N'-diacetylchitobi
ose moiety of the sugar chain of human transferrin tetraglycopeptide A
sn-Tyr-Asn(GlcNAc)2(Man)3(GlcNAc)2-(Gal)2-Lys to yield equimolar amoun
ts of peptide Asn-Tyr-Asn(GlcNAc)Lys and sugar chain (Gal)2(GlcNAc)2-(
Man)3(GlcNAc). The enzyme was identified as an endo-beta-N-acetylgluco
saminidase. The enzyme acted on HSA-A with desialylated and defucosyla
ted outer chain moieties of the sugar chains at a similar rate as that
of native HSA-A. The enzyme activity was reduced to 13 and 5% using H
SA-A with the sugar chains whose outer chain moieties lacked Gal and G
lcNAc, respectively, from the nonreducing end. The enzyme also acted o
n human transferrin, calf fetuin, and asparagine oligosaccharides of t
ransferrin and fetuin. On the other hand, the enzyme did not act on ov
albumin, RNase B, Taka-amylase, yeast invertase, and ovalbumin asparag
ine oligosaccharides. These results indicate that human salivary endo-
beta-N-acetylglucosaminidase is specific for complex type sugar chains
and can release the sugar chains from native glycoproteins and glycop
eptides regardless of the existence of a Fuc residue on the proximal G
lcNAc of the N,N'-diacetylchitobiose core of their sugar chains. The s
ource of the enzyme was epithelial cells peeling from the oral cavity
epithelium into saliva. The enzyme was thought to be integrated on the
surface of the epithelial cell membrane. This enzyme was named endo-b
eta-N-acetylglucosaminidase HS. Thus, these studies indicate that the
properties of the enzyme are distinct from those of known endo-beta-N-
acetylglucosaminidase and endo-beta-N-acetylglucosaminidase HS is a no
vel endo-beta-N-acetylglucosaminidase.