Revision of the structure for an endo-beta-N-acetylglucosaminidase H substrate using a novel modification of the Smith degradation.

(Man)5(GlcNAc)2Asn was shown in a previous study (Trimnble, R. B., Tarentino, A. L., Plummer, T. H., Jr., and Maley, F. (1978) J. Biol. Chem. 253, 4508-4511) to be hydrolyzed by alpha-mannosidase to Man alpha 1 leads to 6Man alpha 1 leads to 6(Man alpha 1 leads to 3)Man beta 1 leads to 4GlcNAc beta 2 leads to 4GlcNAc-Asn. The latter is the most effective substrate for endo-beta-N-acetylglucosaminidase H tested to date. By employing a new and highly sensitive modification of the Smith degradation, it is shown that this compound is in reality Man alpha 1 leads to 6(Man alpha 1 leads to 3)Man alpha 1 leads to 6Man beta 1 leads to 4GlcNAc beta 1eads to 4GlcNAc-Asn. The method entails the conversion of a glycosyl asparagine derivative to its corresponding dimethylaminonaphthyl sulfonyl analogue, which after periodate oxidation is treated directly with Dowex 50-H+ to eliminate the modified carbohydrate residues. The dansylated products, which are eluted from the resin with ammonium hydroxide, can be identified rapidly by thin layer chromatography.

be hydrolyzed by a-mannosidase to Manal + 6Manal 6(Manal+ 3)Manbl+ 4GlcNAcPl-+ 4GlcNAc-Asn. The latter is the most effective substrate for endo-P-Nacetylglucosaminidase H tested to date. By employing a new and highly sensitive modification of the Smith degradation, it is shown that this compound is in reality Manal + 6(Manal+ 3)Manal -P 6Manpl+ 4GlcNAcfll -+ 4GlcNAc-Asn. The method entails the conversion of a glycosyl asparagine derivative to its corresponding dimethylaminonaphthyl sulfonyl analogue, which after periodate oxidation is treated directly with Dowex 50-H+ to eliminate the modified carbohydrate residues. The dansylated products, which are eluted from the resin with ammonium hydroxide, can be identified rapidly by thin layer chromatography.
Our previous study (1) on the substrate specificity of endo-P-N-acetylglucosaminidase H ( 2 ) extended the range of activity of this enzyme to asparaginyl glycopeptides containing only a single mannosyl residue. Although its rate of hydrolysis was lo5to 106-fold slower than that of (Man)5(GlcNAc)2Asn, Man(G1cNAc)zAsn was eventually hydrolyzed to completion. One of the compounds prepared for this study by a limited a-mannosidase digestion of (Man)5(GlcNAc)2Asn was (Man)r(GlcNAc)2Asn, with a structure determined from compositional and methylation analyses to be that of A in Fig. 1.
The proposed structure was based in part on that described for a similar compound isolated from an ovalbumin digest (3) but with the peripheral mannose linked a1 + 3.
However, it is evident from the structures of A and B in Fig. 1 that a methylation analysis would not distinguish between these compounds, and while our originally proposed structure of this compound was A , it could also be B. A means to resolve this problem presented itself in the form of the Smith degradation (4), which should yield Man(G1cNAc)pAsn from A but (GlcNAc)2Asn from B. In the course of this study, which proved eventually that B and not A is the correct * This work was supported in part by United States Public Health Service Grant GM 23900 from the National Institute of General Medical Sciences, Public Health Service, Department of Health and Human Services. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. structure, a novel and interesting modification of the Smith degradation was uncovered, one that should be valuable for the structural analysis of glycosyl asparagine derivatives and possibly glycopeptides.

EXPERIMENTAL PROCEDURES
The various asparagine-containing oligosaccharides used in these experiments were prepared as described earlier (1) and dansylated with [Me-3H]dimethylaminonaphththalene-l-sulfonyl chloride (New England Nuclear) which had been diluted with unlabeled dansyl' chloride. The final specific radioactivity of the dansyl-Asn oligosaccharides was from 13 to 60 X IO6 cpm/pmol. The modified Smith degradation was conducted as follows: 10 to 50 nmol of a 3H-labeled dansylated glycosyl asparagine derivative (larger than (GlcNAc)2Asn) were incubated overnight at 4°C with 0.2 ml of a solution containing 50 mM sodium acetate, pH 4.5, and 30 mM sodium periodate. An additional 0.1 ml of the latter was then added, and the incubation was continued for another 24 h at 4°C. The solution was brought to 2 ml with water and passed through a column of Dowex 50H+-X2 (100-200 mesh) (6 X 15 mm) in a Pasteur pipette. The column was washed with water until the eluate was neutral and then eluted after 18 to 24 h at room temperature with 5-ml aliquots of 5% NHsOH. Most of the radioactivity (40 to 60% of that added to the column) was eluted in the fist fraction, which was concentrated in uacuo to dryness in a 50-ml pear-shaped flask. The residue was taken up in 0.1 ml of water and 10 pl were applied to a thin layer cellulose (0.1-mm) glass plate (20 cm X 20 cm), which was developed in I-butanol/ethanol/HzO (2:l:l). Although the products in most instances were clearly identified on exposure of the chromatogram to UV light, even greater sensitivity could be achieved by fluorography after the region with the tritiated compounds had been irrigated with a solution containing 2-methylnaphthalene, toluene, and diphenyloxazole (5).

RESULTS AND DISCUSSION
As indicated previously (l), a limited a-mannosidase digestion of (Man)e(GlcNAc)pAsn yielded (Man)4(GlcNAc)2Asn and (Man)3(GlcNAc)2Asn. The structure of the tetramannosyl product was believed to be that of A in Fig. 1, based on a similar derivative isolated from a pronase digest of ovalbumin (3). However, the same methylated alditol acetates used to derive the structure in A would also be obtained from structure B. A possible means of distinguishing between A and B was suggested by the Smith degradation (4), wherein A should be converted to Man(GlcNAc)2Asn and B to (GlcNAc)nAsn.
The usual methodology employed in the Smith degradation is that of periodate oxidation, followed by sodium borohydride reduction and acid hydrolysis to remove the modified sugars. However, in the course of studies with GlcNAcAsn-dansyl, it was found that on attempting to isolate the periodate oxidation product of this compound on Dowex 50' by elution with NH40H, both asparagine and aspartic acid were present in the dansylated elution products (Fig. 2). This result contrasted with that obtained by purifying oxidized GlcNAcAsn-dansyl on Dowex 1 formate, which yielded the dialdehyde only, and suggested that Dowex 50+ was catalyzing the breakdown of the dialdehyde to the products observed in Fig. 2. The fact that dansyl-asparagine was present was surprising, since nor-' The abbreviation used is: dansyl, 5-dimethylaminonaphthalene-lsulfonyl.

-6
Manel -c 1GlcNAcE1 -1ClcNAc-Asn a-Mannosidase m d y only aspartic acid would have been obtained by applying the Smith degradation to GlcNAcAsn or its dansylated derivative. Thus treatment of the dansylated oxidized asparagine oligosaccharides with Dowex 50-H' suggested that this procedure could replace the reduction and acid hydrolysis steps required in the traditional Smith degradation, the net result being not only higher yields but greatly enhanced sensitivity in identifying the reaction products.
T o c o n f m t h a t t h e products resulting from the Dowex 50-H' procedure were not an anomaly associated with the breakdown of GlcNAcAsn, the reaction was applied to (GlcNA&Asn, and as shown in Fig. 2, only GlcNAcAsn was obtained. This finding suggested that Dowex 50-H' could successfully catalyze the acid hydrolysis of more complex carbohydrates than GlcNAcAsn. The reaction was therefore applied to even larger and more complex oligosaccharides than those employed in Fig. 2, such as dansylated Dowex 50-H+ columns. For the larger oligosaccharides it was necessary to use Dowex 50-H'-X2 (100-200 mesh) to ensure maximal retention of the dansylated compounds to the resin. On elution and thin layer chromatography of the oxidation products from (Man)3(GlcNAc)2Asn-dansyl and (Man)4-(GlcNAc)2Asn-dansyl, it was found that these compounds had been converted in high yields (at least 50%) to (Glc-NAc),Asn-dansyl (Fig. 3). The conversion of linear (Man)3-(G1cNAc)iAsn (lune I ) to mostly ( G~c N A c )~A s~ was anticipated from its structure (1); but the fact that (Man)4-(G1cNAc)gAsn (lune 2) yielded (G1cNAc)~Asn indicates that this compound, a product of the limited a-mannosidase digestion of (Man)~(GlcNAc)nAsn, is compound B and not A as reported previously (1). Of interest is the fact that (Man)S(GlcNAc)Asn (lune3) and (Man)s(Gl~NAc)~Asn (lune 4) were converted primarily to (Man)2(Gl~NAc)~Asn, which in effect c o n f m s t h e structures of the parent compounds (7).
In general it was not necessary to employ radioactive dansyl compounds, as the sensitivity of the method with fluorescent compounds was adequate for the detection of reaction products within a range of 1 to 5 nmol or less. It is possible that this procedure may also be applicable to glycopeptides (both dansylated and undansylated) which are retained by Dowex 50-H'.