Purification and Characterization of UDP-GlcNAc: GlcNAcβ1–6(GlcNAcβ1–2)Manα1-R [GlcNAc to Man]-β1, 4-N-acetylglucosaminyltransferase VI from Hen Oviduct*

A new β1,4-N-acetylglucosaminyltransferase (GnT) responsible for the formation of branched N-linked complex-type sugar chains has been purified 64,000-fold in 16% yield from a homogenate of hen oviduct by column chromatography procedures using Q-Sepharose FF, Ni2+-chelating Sepharose FF, and UDP-hexanolamine-agarose. This enzyme catalyzes the transfer of GlcNAc from UDP-GlcNAc to tetraantennary oligosaccharide and produces pentaantennary oligosaccharide with the β1–4-linked GlcNAc residue on the Manα1–6 arm. It requires a divalent cation such as Mn2+and has an apparent molecular weight of 72,000 under nonreducing conditions. The enzyme does not act on biantennary oligosaccharide (GnT I and II product), and β1,6-N-acetylglucosaminylation of the Manα1–6 arm (GnT V product) is essential for its activity. This clearly distinguishes it from GnT IV, which is known to generate a β1–4-linked GlcNAc residue only on the Manα1–3 arm. Based on these findings, we conclude that this enzyme is UDP-GlcNAc:GlcNAcβ1–6(GlcNAcβ1–2)Manα1-R [GlcNAc to Man]-β1,4-N-acetylglucosaminyltransferase VI. This is the only known enzyme that has not been previously purified among GnTs responsible for antenna formation on the cores of N-linked complex-type sugar chains.

Determination of GnT VI Activity-GnT VI activity was assayed as described previously (21) with minor modifications. The standard incubation mixture contained the following components in a total volume of 10 l: 150 mM HEPES (pH 8.0), 10 mM UDP-GlcNAc, 100 mM GlcNAc, 30 mM MnCl 2 , 0.5% Triton X-100, 2 mg/ml bovine serum albumin, 24 or 120 pmol of substrate ([(2,4),(2,6)]-PA), and 2 l of enzyme fraction. After incubation at 37°C for 4 h, 40 l of water was added, and the enzyme reaction was stopped by boiling for 1 min. After centrifugation at 13,000 rpm for 5 min, 10 l of the supernatant from the reaction mixture was applied to a TSK-Gel ODS-80TM column (4.6 ϫ 75 mm; Tosoh, Tokyo, Japan). Elution was performed at 55°C with 20 mM ammonium acetate (pH 4.0) at a flow rate of 1.6 ml/min. Fluorescence was monitored with excitation and emission wavelengths of 320 and 400 nm, respectively. The specific activity of the enzyme is expressed as picomoles of product formed per h/mg of protein. The protein concentration was determined with a BCA kit (Pierce) or a Bio-Rad protein assay kit using bovine serum albumin as the standard. The activities of GnTs III-V were measured according to the method of Nishikawa et al. (22).
Homogenization and Preparation of the Microsomal Fraction (Step 1)-All purification steps were carried out at 4°C. Frozen hen oviduct (270 g) was homogenized in a Waring Blendor in Buffer A. After centrifugation at 8000 rpm for 10 min, the supernatant was pooled, and the pellet was subjected to two more extractions, after which all the supernatants (1200 ml) were combined. Following centrifugation at 78,000 ϫ g for 2 h, a microsomal fraction of ϳ13 g was obtained.
Solubilization of GnT VI (Step 2)-The microsomal fraction was suspended in 150 ml of Buffer B, gently stirred for 2 h, and then centrifuged at 105,000 ϫ g for 1 h. The supernatant fraction was collected, and the residual pellet was subjected to another extraction, followed by the same ultracentrifugation. The first and second Triton X-100 extracts were combined and used for further enzyme purification.
Q-Sepharose FF Column Chromatography (Step 3)-The Triton X-100 extracts were applied to a column of Q-Sepharose FF (10 ϫ 4 cm) that had been equilibrated with Buffer C. Fractions of 40 ml were collected through this column chromatography. The column was washed with Buffer C until the protein concentration was reduced to 1.0 mg/ml. Elution was then carried out with a linear gradient established between 1000 ml of Buffer C and 1000 ml of 0.8 M NaCl/Buffer C. The fractions containing GnT VI activity were combined.
Ni 2ϩ -chelating Sepharose FF Column Chromatography (Step 4)-The pool of fractions from Step 3 was applied directly to a column of Ni 2ϩ -chelating Sepharose FF (2.5 ϫ 10 cm) that had been equilibrated with Buffer D. Ni 2ϩ -chelating Sepharose FF resin was layered on the chelating Sepharose FF resin without metal ions (2.5 ϫ 5 cm) to avoid any possible leakage of Ni 2ϩ . Fractions of 11 ml were collected through this column chromatography. All of the GnT VI activity was retained by the column. After washing the column with Buffer D until the protein concentration was reduced to 0.2 mg/ml, GnT VI activity was eluted with a linear gradient established between 300 ml of Buffer D and 300 ml of 0.1 M glycine/Buffer D. The fractions containing GnT VI activity were combined, and the buffer in this fraction was replaced by Buffer E by means of an Amicon Diaflow Ultrafiltrater using a YM-30 membrane (Amicon, Inc., Beverly, MA).
UDP-hexanolamine-Agarose Affinity Column Chromatography (Step 5)-After the above step, the column was siliconized with Sigmacote (Sigma), and siliconized tubes (Assist, Tokyo) were used for fractionation. The UDP-hexanolamine-agarose column (1.5 ϫ 15 cm) had been equilibrated with 0.05 M NaCl/Buffer E, and the concentrated enzyme fraction from Step 4 was applied to the column, followed by washing with 100 ml of Buffer E. GnT VI activity was eluted with Buffer F. At the loading and washing steps, fractions of 5 ml were collected. At the elution step, fractions of 1.4 ml were collected, and those containing GnT VI activity were pooled.
UDP-hexanolamine-Agarose Affinity Column Chromatography (Step 6)-An equal volume of Buffer G was added to the pool of fractions from Step 5, and this solution was rechromatographed by the same procedure as described for Step 5. This purified GnT VI fraction was used for the enzyme characterization.
Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was performed by the method of Laemmli (23) using 10% gels. Molecular markers (Amersham Pharmacia Biotech, Little Chalfont, United Kingdom) were used for size standards. Proteins on the gels were stained with a silver staining kit (DAIICHI 2D silver stain II, Daiichi Pure Chemicals, Tokyo).

RESULTS
Purification of GnT VI-The activity of GnT VI was assayed using the fluorescently labeled agalactotetraantennary sugar chain as an acceptor substrate basically according to the original method of Taguchi et al. (21). Since bovine serum albumin (2 mg/ml) was found to be effective in preserving enzyme activity at 37°C, especially for the highly purified enzyme fraction (Steps 5 and 6), it was routinely included in the standard assay mixture.
Like other glycosyltransferases, GnT VI activity was found in the microsomal fraction. At least 80% of the GnT VI activity was associated with the microsomal fraction. GnT VI activity was successfully solubilized from the microsomal fraction by extraction with Triton X-100. Substantial amounts of proteins were separated from GnT VI by Q-Sepharose FF and Ni 2ϩchelating Sepharose FF chromatographies (Fig. 3, A and B). After the Ni 2ϩ -chelating Sepharose FF column chromatography (Step 4), the enzyme was more stable than in Steps 2 and 3. As opposed to results with GnTs III and IV (11,13), no GnT VI activity was eluted after application to a Cu 2ϩ -chelating column. Following the above two column chromatography steps, the GnT VI active fraction still contained GnT III and ␤1,4-galactosyltransferase activities. Cation exchange, dye affinity, gel filtration, and lectin column chromatographies did not prove to be effective for further purification of the enzyme fraction after Step 4. The use of an affinity column (Steps 5 and 6) (Fig. 3C) packed with an analog of the common donor substrate for GnTs (UDP-hexanolamine) as a ligand was effective. This affinity column has been proven to be very effective for purification of GnTs and was first used for the purification of GnT I by Oppenheimer and Hill (5). The activity of GnT III was not bound to this column under the conditions described under "Experimental Procedures." The majority of GnT VI activity was bound to this affinity column and was eluted by the buffer lacking MnCl 2 . Since the eluted fraction (after Step 5) still contained several proteins as revealed by SDS-polyacrylamide gel electrophoresis (data not shown), this fraction was rechromatographed (Step 6) under the same conditions as described for Step 5. The addition of DTT was crucial in preserving GnT VI activity after Steps 5 and 6, whereas it was without effect in Step 4. The enzyme fraction of Step 6 showed a single broad band with a molecular weight of 72,000 on SDS-polyacrylamide gel electrophoresis under nonreducing conditions (Fig. 4) and of 60,000 under reducing conditions (data not shown), suggesting that the enzyme was a glycoprotein. This band was also correlated with GnT VI activity by elution with UDP (5 mM) or NaCl (0.4 M) from a UDP-hexanolamine-agarose column, suggesting that this band was GnT VI (data not shown). No GnT III-V activities were detected in this fraction when [2,2]-PA was used as an acceptor substrate. All experiments for enzyme charac-terization described below were performed using the chromatographic Step 6 material. Table I summarizes the purification of GnT VI, which was purified 64,000-fold in 16% yield.
pH Optimum-GnT VI possessed activity over a relatively broad pH range with an optimum at pH 7.75.
Effect of MnCl 2 Concentration on GnT VI Activity-GnT VI activity was Mn 2ϩ -dependent and was high around 5 mM. When the MnCl 2 concentration was increased, the activity gradually decreased.
Effect of Divalent Cations on GnT VI Activity-The effects of divalent cations on GnT VI activity were examined. The activity was maximal with Mn 2ϩ . Co 2ϩ , Mg 2ϩ , and Ni 2ϩ could partially substitute for Mn 2ϩ , whereas Ca 2ϩ , Zn 2ϩ , Fe 2ϩ , and Cu 2ϩ showed no significant effect (Table II).
Acceptor Substrate Specificity-The acceptor substrate specificity of GnT VI was examined using complex-type sugar chains. Purified GnT VI was revealed to select its substrate clearly (Table III). It could transfer GlcNAc to [(2,4),(2,6)]-PA and [2,(2, 6)]-PA (GnT V product), whereas no activity was observed when [2,2]-PA or [(2,4),2]-PA (GnT IV product) was used as a substrate. DISCUSSION Six different GnTs (I-VI) are known to be mainly involved in antenna formation on the cores of N-linked complex-type sugar chains (Fig. 1). In addition, Raju and Stanley (24) recently identified two other distinct GnT activities designated GnTs VII and VIII from Chinese hamster ovary mutant cells. GnT VI activity is defined as that catalyzing the transfer of GlcNAc to the Man␣1-6 arm and forms a GlcNAc␤1-4Man␣1-6 linkage. Only GnT VI has not been purified in the GnTs that act on ␣-Man residues. The substrate specificity of GnT VI has not been clearly determined, although it has been suggested, by using hen oviduct microsomes as an enzyme source, that GnT VI acts after assembly of GlcNAc␤1-2Man␣1-3-, GlcNAc␤1-2Man␣1-6-, and GlcNAc␤1-6Man␣1-6-antennae by the action of GnTs I, II, and V, respectively (20). The tissue distribution of this enzyme seems to be highly restricted. So far, only avian oviduct (20) and fish ovary (25) have been shown to express this enzyme activity. The transferrin synthesized by the human hepatocarcinoma cell line HepG2 was reported to have a pentaantennary glycan chain (26), which is the product of GnT VI activity.
The substrate specificity of purified GnT VI agrees with the results reported by Brockhausen et al. (20) in a study using hen oviduct microsomes as the enzyme source and GlcNAc␤1-6(Gl-cNAc␤1-2)Man␣1-6Man␤1-(CH 2 ) 8 COOCH 3 as an acceptor substrate. Some differences from their results in terms of optimal conditions for activity, e.g. the effect of Mn 2ϩ concentration and the effects of divalent cations, are to be noted. Differences are also noted for the optimal conditions of enzyme from hen oviduct microsomes (21) and this purified GnT VI using the same substrate, [(2,4),(2,6)]-PA. Such discrepancies could be due to the presence of another (or other) GnT VI(s) in hen oviduct microsomes.