Advertisement

Purification and Characterization of the GalNAc-4-sulfotransferase Responsible for Sulfation of GalNAcβ1,4GlcNAc-bearing Oligosaccharides *

  • Lora V. Hooper
    Footnotes
    Affiliations
    Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110
    Search for articles by this author
  • Ole Hindsgaul
    Affiliations
    Department of Chemistry, University of Alberta, Edmonton, Alberta T6G 2G2, Canada
    Search for articles by this author
  • Jacques U. Baenziger
    Correspondence
    To whom correspondence should be addressed. Tel.: 314-362-8730; Fax: 314-362-8888
    Affiliations
    Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110
    Search for articles by this author
  • Author Footnotes
    * This work was supported by National Institutes of Diabetes and Digestive and Kidney Diseases Grant R01-DK41738 (to J. U. B.) The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
    § Howard Hughes Medical Institute Predoctoral Fellow.
Open AccessPublished:July 07, 1995DOI:https://doi.org/10.1074/jbc.270.27.16327
      The pituitary glycoprotein hormone lutropin is characterized by its pulsatile appearance in the bloodstream which is important for the expression of its biological activity in the ovary. We have previously shown that lutropin bears unique Asn-linked oligosaccharides terminating with GalNAc-4-SO4 which allow the hormone to be rapidly cleared from the bloodstream via a specific receptor in the liver, thus contributing to its pulsatile appearance in the circulation. Furthermore, we have found that carbonic anhydrase VI, synthesized by the submaxillary gland and secreted into the saliva, also bears Asn-linked oligosaccharides terminating with GalNAc-4-SO4, suggesting that this unique sulfated structure mediates other biological functions in addition to rapid clearance from the circulation. We report here the purification of a GalNAc-4-sulfotransferase which transfers sulfate to terminal β1,4-linked GalNAc on Asn-linked oligosaccharides. We show that the purified submaxillary gland enzyme has kinetic parameters identical to the pituitary enzyme, indicating that the same sulfotransferase is responsible for the sulfation of lutropin oligosaccharides in pituitary and carbonic anhydrase VI oligosaccharides in submaxillary gland. This GalNAc-4-sulfotransferase has an apparent molecular mass of 128 kDa and can be specifically photoaffinity radiolabeled with 3′,5′-ADP, a competitive inhibitor of sulfotransferase activity. The acceptor specificity of this GalNAc-4-sulfotransferase indicates that it is able to transfer sulfate to terminal GalNAcβ1,4GlcNAc on both N- and O-glycosidically linked oligosaccharides, suggesting that this enzyme is also responsible for the sulfation of O-linked glycans on proopiomelanocortin.
      The pituitary glycoprotein hormone lutropin (LH)1(
      The abbreviations used are: LH
      lutropin
      GalNAc
      N-acetylgalactosamine
      GlcNAc
      N-acetylglucosamine
      Man
      mannose
      PAPS
      3′-phosphoadenosine 5′-phosphosulfate
      3′,5′-ADP
      adenosine 3′,5′-diphosphate
      GGnM-MCO
      GalNAcβ1,4GlcNAcβ1,2Manα-O(CH2)8COOCH3
      CG
      chorionic gonadotropin
      WGA
      wheat germ agglutinin
      HPLC
      high performance liquid chromatography
      PAGE
      polyacrylamide gel electrophoresis.
      )
      1The abbreviations used are: LH
      lutropin
      GalNAc
      N-acetylgalactosamine
      GlcNAc
      N-acetylglucosamine
      Man
      mannose
      PAPS
      3′-phosphoadenosine 5′-phosphosulfate
      3′,5′-ADP
      adenosine 3′,5′-diphosphate
      GGnM-MCO
      GalNAcβ1,4GlcNAcβ1,2Manα-O(CH2)8COOCH3
      CG
      chorionic gonadotropin
      WGA
      wheat germ agglutinin
      HPLC
      high performance liquid chromatography
      PAGE
      polyacrylamide gel electrophoresis.
      is essential for the regulation of a number of physiological processes involved in reproduction, including follicular maturation, ovulation, and the secretion of estradiol and progesterone. LH exerts its effects by binding to and stimulating the lutropin/chorionic gonadotropin (LH/CG) receptor in the ovary. One of the distinctive features of the biology of LH is its pulsatile pattern of appearance in the bloodstream which is thought to be necessary for the in vivo expression of its bioactivity (
      • Lincoln D.W.
      • Fraser H.M.
      • Lincoln G.A.
      • Martin G.B.
      • McNeilly A.S.
      ,
      • Crowley W.F.
      • Hofler J.G.
      ). Like other G protein-coupled receptors, the LH/CG receptor is desensitized upon ligand binding(
      • Wang H.
      • Segaloff D.L.
      • Ascoli M.
      ); thus, the rise and fall of LH levels in the circulation may be crucial for maintaining maximal stimulation of the LH/CG receptor in the ovary. Many factors are important in producing this pulsatile rise and fall in the bloodstream, including frequency of stimulation of LH release from the pituitary by LH-releasing hormone, the amount of LH released during the secretory burst, and the circulatory half-life of LH(
      • Veldhuis J.D.
      • Carlson M.L.
      • Johnson M.L.
      ,
      • Evans W.S.
      • Sollenberger M.J.
      • Booth R.A.J.
      • Rogol A.D.
      • Urban R.J.
      • Carlsen E.C.
      • Johnson M.L.
      • Veldhuis J.D.
      ).
      We have determined previously that LH from four different animal species bears unique Asn-linked oligosaccharides terminating with the sequence SO4-4-GalNAcβ1,4GlcNAcβ1,2Manα in contrast to the more commonly found terminal sequence sialic acid-Galβ1,4GlcNAcβ1,2Manα(
      • Green E.D.
      • Van Halbeek H.
      • Boime I.
      • Baenziger J.U.
      ,
      • Green E.D.
      • Baenziger J.U.
      ,
      • Green E.D.
      • Baenziger J.U.
      ,
      • Smith P.L.
      • Bousfield G.S.
      • Kumar S.
      • Fiete D.
      • Baenziger J.U.
      ). Subsequent clearance studies showed that native bovine (b) LH, bearing oligosaccharides which terminate with GalNAc-4-SO4, is cleared from the circulation 4-fold faster than recombinant bLH bearing oligosaccharides which terminate with sialic acid(
      • Baenziger J.U.
      • Kumar S.
      • Brodbeck R.M.
      • Smith P.L.
      • Beranek M.C.
      ). The rapid clearance of native bLH is mediated by a receptor expressed in hepatic reticuloendothelial cells which recognizes oligosaccharides terminating with GalNAc-4-SO4 and removes the hormone from the bloodstream (
      • Fiete D.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). The presence of sulfated rather than sialylated oligosaccharides on LH results in a shorter circulatory half-life, which contributes to the pulsatile appearance of the hormone in the bloodstream and has a significant impact on in vivo hormone bioactivity. The presence of sulfated oligosaccharides on LH is therefore crucial to the biologic function of this hormone.
      The presence of sulfated oligosaccharides on LH reflects the sequential action of two highly specific enzymes expressed in the pituitary. A glycoprotein hormone-specific GalNAc-transferase, which recognizes a specific peptide motif in the underlying protein(
      • Smith P.L.
      • Baenziger J.U.
      ,
      • Mengeling B.J.
      • Manzella S.M.
      • Baenziger J.U.
      ), transfers GalNAc to the synthetic intermediate GlcNAc2Man3GlcNAc2Asn. The specificity of this transferase accounts for the addition of GalNAc to LH and thyroid stimulating hormone (
      • Green E.D.
      • Baenziger J.U.
      ,
      • Green E.D.
      • Baenziger J.U.
      ) but not to other pituitary glycoproteins. A GalNAc-4-sulfotransferase, which utilizes 3′-phosphoadenosine 5′-phosphosulfate (PAPS) as the sulfate donor, accounts for the addition of sulfate to the 4-hydroxyl of terminal GalNAc residues of LH(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). Unlike the GalNAc-transferase, the GalNAc-4-sulfotransferase does not appear to be protein specific(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). Both GalNAc-transferase and GalNAc-4-sulfotransferase levels are up-regulated in rat pituitary in concert with rising LH levels following ovariectomy, resulting in the maintenance of terminal glycosylation of LH oligosaccharides with GalNAc-4-SO4 (
      • Dharmesh S.M.
      • Baenziger J.U.
      ). Thus, synthesis of the sulfated oligosaccharides on LH is highly regulated, further supporting the importance of these unique oligosaccharides to the biological function of LH. In order to better understand the regulation of the synthesis of this sulfated structure, these transferases must be purified and cloned.
      We have shown that both the GalNAc-transferase and the GalNAc-4-sulfotransferase are present in tissues other than the pituitary, including submaxillary gland, lacrimal gland, and kidney (
      • Dharmesh S.M.
      • Skelton T.P.
      • Baenziger J.U.
      ), suggesting the existence of terminal GalNAc-4-SO4 on glycoproteins from other tissues. Furthermore, we have recently shown that oligosaccharides terminating with GalNAc-4-SO4 are present on a secreted form of carbonic anhydrase from submaxillary gland where there are also high levels of both transferases(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ). Here we report the purification of the GalNAc-4-sulfotransferase to homogeneity from bovine submaxillary gland and present evidence that this is the same enzyme which is responsible for the sulfation of LH oligosaccharides in the pituitary. In addition, we demonstrate that this sulfotransferase is not limited in specificity to Asn-linked oligosaccharides, but also has the ability to transfer sulfate to β1,4-linked GalNAc on O-glycosidically linked oligosaccharide acceptors. These results suggest that this sulfotransferase is therefore responsible for the sulfation of N-linked oligosaccharides on LH and carbonic anhydrase VI and O-linked oligosaccharides on proopiomelanocortin(
      • Skelton T.P.
      • Kumar S.
      • Smith P.L.
      • Beranek M.C.
      • Baenziger J.U.
      ).

      MATERIALS AND METHODS

      Bovine submaxillary glands were purchased from Pel-Freez (Rogers, AR). Wheat germ agglutinin (WGA)-Sepharose (10 mg of WGA/ml of gel) was made by coupling WGA (Sigma) to CNBr-activated Sepharose 4B (Pharmacia Biotech Inc.). 3′,5′-ADP-agarose was obtained from Sigma. [35S]PAPS was enzymatically synthesized as described previously (
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ) using [35S]SO4 from ICN (Costa Mesa, CA). [32P]3′,5′-ADP (3000 Ci/mmol) was purchased from ICN. The synthesis of GGnM-MCO has been previously published(
      • Vandana
      • Hindsgaul O.
      • Baenziger J.U.
      ,
      • Srivastava V.
      • Hindsgaul O.
      ).

      Purification of GalNAc-4-sulfotransferase from Bovine Submaxillary Gland

      All steps were carried out at 4°C.

      Step 1: Preparation of Triton X-100 Extract

      600 g of bovine submaxillary glands were thawed, passed through a meat grinder, homogenized, and extracted in 1% (v/v) Triton X-100 as described previously by Schwyzer and Hill (
      • Schwyzer M.
      • Hill R.L.
      ) for purification of a porcine submaxillary gland GalNAc-transferase.

      Step 2: DEAE-Sepharose

      The Triton X-100 extract (1010 ml) was diluted to 6 liters in Buffer A (25 mM Tris, pH 7.4, 13% glycerol, 0.1% Triton X-100) and was applied to a 1-liter DEAE-Sepharose (Pharmacia) column equilibrated in Buffer A. After washing in 4 liters of Buffer A the column was eluted in 4 liters of 0.5 M NaCl in Buffer A.

      Step 3: WGA-Sepharose

      The DEAE-Sepharose eluate was applied to a 100-ml WGA-Sepharose column which was washed with 1 liter of 0.25 M KCl in Buffer A and then with 500 ml of Buffer B (15 mM HEPES, pH 7.4, 13% glycerol, 0.1% Triton X-100). The column was batch-eluted in 500 ml of 0.5 M GlcNAc, 0.25 M NaCl in Buffer B.

      Step 4: Sulfonamide-agarose

      The WGA-Sepharose eluate was passed over a 20-ml p-aminomethylbenzene sulfonamide-agarose column (Sigma) and the flow-through was collected. The column was washed with 100 ml of 0.25 M NaCl in Buffer B, and the wash fraction was combined with the flow-through.

      Step 5: Phenyl-Sepharose

      The combined flow-through and wash fractions from the previous step were diluted to 3 liters in Buffer C (15 mM HEPES, pH 7.4, 13% glycerol) and loaded at a flow rate of 5 ml/min onto an 80-ml phenyl-Sepharose column (Pharmacia) equilibrated in Buffer C. The column was washed with 500 ml of Buffer C and batch-eluted in 200 ml of 2% (v/v) Triton X-100 in Buffer C.

      Step 6: 3′,5′-ADP-agarose: NaCl Gradient Elution

      1 mM GDP was added to the phenyl-Sepharose eluate, which was incubated overnight with 10 ml of 3′,5′-ADP-agarose (1.9 μmol/ml gel). The resin was collected in a 1.5-cm diameter column and washed first with 50 ml of 1 mM GDP in Buffer D (15 mM HEPES, pH 7.4, 4 mM magnesium acetate, 0.1% Triton X-100, 13% glycerol) and then with 100 ml of Buffer D. The column was eluted in a 0-1 M NaCl gradient in Buffer D at a flow rate of 0.5 ml/min in a total volume of 80 ml. 1-ml fractions were collected and assayed for protein and for GalNAc-4-sulfotransferase activity.

      Step 7: Phenyl-Sepharose Chromatography

      The active fractions from step 6 were pooled and diluted with 2 volumes of Buffer C and adsorbed overnight to 6 ml of phenyl-Sepharose equilibrated in Buffer C. The resin was collected in a 1.5-cm diameter column, washed with 30 ml of Buffer C, and eluted in 30 ml of 2% Triton X-100 in Buffer C at a flow rate of 0.5 ml/min. 1-ml fractions were collected and assayed for GalNAc-4-sulfotransferase activity.

      Step 8: 3′,5′-ADP-agarose Chromatography: Elution with 5′-ADP

      The pooled active fractions from the previous step were brought to 1 mM GDP and were adsorbed to 2 ml of 3′,5′-ADP-agarose overnight. The resin was collected in a 1.0-cm diameter column and was washed with 20 ml of 1 mM GDP in Buffer D, followed by 50 ml of Buffer D. The column was eluted in 20 ml of 20 mM 5′-ADP in Buffer D at a flow rate of 0.5 ml/min. 0.5-ml fractions were collected and assayed for the presence of protein and GalNAc-4-sulfotransferase activity.

      Step 9: DEAE-Sepharose Chromatography

      The pooled active fractions from step 8 were applied to a 0.3-ml DEAE-Sepharose column. The column was washed with 1 ml of 20 mM 5′-ADP in Buffer D and the flow-through and wash fractions were combined.

      Step 10: Hydroxylapatite Chromatography

      The combined flow-through and wash fractions from the previous step were applied to a 0.3-ml hydroxylapatite column (Bio-Rad) which was washed with 2 ml of Buffer D and eluted with 2 ml of 0.2 M sodium phosphate, pH 7.4, 0.1% Triton X-100, 13% glycerol.

      Assay of GalNAc-4-sulfotransferase

      GalNAc-4-sulfotransferase reactions (50 μl) were carried out as described (
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ) at 28°C for 2 h, and contained 15 mM HEPES, pH 7.4, 1% Triton X-100, 40 mM β-mercaptoethanol, 10 mM NaF, 1 mM ATP, 4 mM magnesium acetate, 13% glycerol, protease inhibitors, 2 μM unlabeled PAPS, 1 × 106 cpm of [35S]PAPS, 20 μM GGnM-MCO, and sulfotransferase source. [35S]SO4-GGnM-MCO was separated from [35S]PAPS and from labeled endogenous acceptors by passage over a Sep-Pak C18 cartridge (Waters) as described previously(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). Control reactions were done in the absence of GGnM-MCO.

      GalNAc-4-sulfotransferase Product Proof

      A standard GalNAc-4-sulfotransferase reaction was done as described above, omitting unlabeled PAPS and using partially-purified bovine pituitary sulfotransferase or homogeneous bovine submaxillary sulfotransferase as the enzyme source. Reactions were incubated at 28°C overnight and the sulfated GGnM-MCO product was isolated on Sep-Paks as described previously(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). Partial mild acid hydrolysis and separation of sulfated monosaccharides by HPLC on a CarboPak PA1 column (Dionex) were carried out as described previously(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ).

      Photoaffinity Labeling with [32P]3′,5′-ADP

      Photolabeling reactions (50 μl) were carried out in Buffer B containing the indicated amounts of protein, 4 mM MgCl2, and 1 μCi of [32P]3′,5′-ADP, in the presence or absence of 1 mM 3′,5′-ADP or 1 mM ATP. Reactions were incubated on ice in a 96-well microtiter plate for 5 min and then exposed to ultraviolet light for 30 min at 1-cm distance using a hand-held UV lamp set at 254 nm. Duplicate reactions were not exposed to UV. Reactions containing DEAE-Sepharose and WGA-Sepharose elutions were precipitated with 10% trichloroacetic acid and the pellets were boiled in SDS-PAGE sample buffer (10% glycerol, 5% 2-mercaptoethanol, 2% SDS, 0.003% bromphenol blue, and 62.5 mM Tris, pH 6.8) and loaded onto a 7.5% SDS gel. The reaction containing 3′,5′-ADP-agarose NaCl gradient elution was not trichloroacetic acid precipitated, but was boiled in 1 × sample buffer following the reaction and was loaded directly onto the gel. Due to differences in background, autoradiogram exposure times were optimized individually for each of the purification steps and varied between 12 and 48 h.

      Acceptor Substrate Specificity

      Oligosaccharide acceptors attached to a hydrophobic aglycone tail (-MCO) were chemically synthesized as described previously(
      • Vandana
      • Hindsgaul O.
      • Baenziger J.U.
      ,
      • Srivastava V.
      • Hindsgaul O.
      ), terminating with GlcNAc. β1,4-Linked GalNAc was added using β1,4-galactosyltransferase (Sigma) and UDP-GalNAc (Sigma) as described previously(
      • Palcic M.M.
      • Hindsgaul O.
      ). GalNAc-4-sulfotransferase reactions were carried out using pure GalNAc-4-sulfotransferase as described above, except that the acceptor substrate was varied. Km and Vmax values for each oligosaccharide were calculated from double-reciprocal plots of 1/V versus 1/[S], where [S] varied from 3.3 to 40 μM. For oligosaccharides on which there are two terminal β1,4-linked GalNAc residues, the Kmvalues were calculated based on whole oligosaccharide concentration. Correlation coefficients for double-reciprocal plots ranged from 0.96 to 0.99.

      RESULTS

      Purification of GalNAc-4-sulfotransferase

      Table I summarizes the purification of GalNAc-4-sulfotransferase from 600 g of bovine submaxillary glands. The transferase was purified 1576-fold in 10 steps, with a yield of 0.24% over the total homogenate. Bovine submaxillary gland was chosen as the source for this purification because of its high expression of GalNAc-4-sulfotransferase relative to other tissues, including pituitary(
      • Dharmesh S.M.
      • Skelton T.P.
      • Baenziger J.U.
      ).
      TABLE IPurification of GalNAc-4-sulfotransferase
      After solubilization of GalNAc-4-sulfotransferase activity in Triton X-100, the extract was chromatographed on DEAE-Sepharose resulting in 7.1-fold purification over the homogenate. An additional 13-fold purification was achieved by chromatography on WGA-Sepharose with elution in 0.5 M GlcNAc. This was followed by passage of the partially-purified enzyme over a sulfonamide-agarose column which removes carbonic anhydrase VI from the preparation resulting in an additional 1.6-fold purification. Removal of carbonic anhydrase at this stage is essential to the overall success of this purification scheme, and is necessitated by the fact that carbonic anhydrase VI is a major contaminant of the affinity-purified enzyme preparation if not removed. Carbonic anhydrase VI, which we have shown is a substrate for the GalNAc-4-sulfotransferase(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ), is an abundant protein synthesized in the submaxillary and parotid salivary glands (
      • Fernley R.T.
      • Coghlan J.P.
      • Wright R.D.
      ). Sulfonamides are specific inhibitors of carbonic anhydrases and have been used extensively as affinity ligands in the purification of several members of this class of enzymes(
      • Fernley R.T.
      • Coghlan J.P.
      • Wright R.D.
      ,
      • Feldstein J.B.
      • Silverman D.N.
      ,
      • Murakami H.
      • Sly W.S.
      ). Step 6, hydrophobic chromatography on phenyl-Sepharose, facilitated removal of the salt from the enzyme mixture; however, no fold enrichment is given in for this step because the high concentrations of Triton X-100 required for elution prevented accurate protein quantitation. Analysis by SDS-PAGE of the proteins present in purification steps 1-5 is shown in Fig. 2, lanes 1-5.
      Figure thumbnail gr2
      Figure 2:SDS-PAGE of GalNAc-4-sulfotransferase purification fractions. Lane 1, 10 μg of Triton extract; Lane 2, 10 μg of DEAE-Sepharose eluate; Lane 3, 5 μg of WGA-Sepharose eluate; Lane 4, 5 μg of sulfonamide-agarose flow-through fraction; Lane 5, phenyl-Sepharose 1 eluate; Lane 6, 2 μg of pooled fractions from 3′,5′-ADP-agarose NaCl gradient elution; Lane 7, phenyl-Sepharose 2 eluate; Lane 8, 0.5 μg of pooled fractions from 3′,5′-ADP-agarose 5′-ADP elution. Lane 9, 2 μg of hydroxylapatite column elution. Proteins were visualized by silver nitrate staining.
      The partially-purified GalNAc-4-sulfotransferase eluted from phenyl-Sepharose was affinity-purified on 3′,5′-ADP-agarose with elution in a 0-1 M NaCl gradient (Fig. 1A), resulting in a total purification of 695-fold over the homogenate and enrichment of a predominant protein which migrates at 128 kDa on SDS-PAGE (Fig. 2, lane 6). A significant proportion (>50%) of the GalNAc-4-sulfotransferase activity does not bind to the 3′,5′-ADP-agarose during this first round of affinity chromatography resulting in a significant loss of enzyme activity at this step (only 12% of the activity loaded is recovered in the gradient elution). We have identified and raised monoclonal antibodies to a specific GalNAc-4-sulfotransferase binding protein which is present in the partially-purified enzyme mixture and which passes through the 3′,5′-ADP-agarose column. Preliminary results indicate that this binding protein, when complexed with the sulfotransferase, prevents it from interacting efficiently with the affinity ligand thus explaining the low yields obtained in this purification step.2(
      L. V. Hooper and J. U. Baenziger, unpublished observations.
      ) Following hydrophobic chromatography to remove salt, the GalNAc-4-sulfotransferase was subjected to a second round of affinity chromatography on 3′,5′-ADP-agarose with elution in 20 mM 5′-ADP (Fig. 1B). We utilized 5′-ADP rather than 3′,5′-ADP as the affinity eluent because GalNAc-4-sulfotransferase activity is not detectable in the presence 3′,5′-ADP, a competitive inhibitor, even at concentrations below 1 mM.2 The pooled active fractions from this step were next chromatographed on a small (0.3 ml) DEAE-Sepharose column. Although the transferase is bound to and eluted from DEAE-Sepharose earlier in the purification (Step 2), in the presence of 20 mM 5′-ADP, 50% of the GalNAc-4-sulfotransferase activity loaded passes through DEAE-Sepharose at this later stage. This step results in the removal of relatively minor contaminants at 150 and 200 kDa which remain bound to the resin. In order to concentrate the enzyme, the combined flow-through and wash fractions from this step were next applied to a hydroxylapatite column. The eluate from this step yielded GalNAc-4-sulfotransferase which was purified 1576-fold over the total homogenate, and gave a single, broad band centered around an apparent molecular mass of 128 kDa (Fig. 2, lane 9; this lane was overloaded in order to enhance detection of any minor contaminating proteins). The intensity of staining of this 128-kDa band in the individual hydroxylapatite column elution fractions also correlated with the amount of GalNAc-4-sulfotransferase activity in these fractions (data not shown). The diffuse nature of the purified sulfotransferase may be due to a large number of O-linked chains since quantitative amino acid analysis of the purified transferase shows serine and threonine combined to comprise 11% of the total amino acid content.
      Figure thumbnail gr1
      Figure 1:Affinity chromatography of GalNAc-4-sulfotransferase on 3′,5′-ADP-agarose. A, partially-purified sulfotransferase from step 5 was bound to a 3′,5′-ADP-agarose column which was eluted in a gradient of 0-1 M NaCl. B, enzyme from step 7 was applied to a second 3′,5′-ADP-agarose column which was eluted with 20 mM 5′-ADP as indicated by the arrow. Pooled fractions are indicated with a bar.
      To confirm that the purified sulfotransferase transfers sulfate exclusively to the 4-OH of terminal GalNAc, we characterized the linkage of the sulfate transferred to GGnM-MCO during the assay reaction. [35S]SO4-GGnM-MCO was subjected to mild acid hydrolysis under conditions which cleave glycosidic bonds more rapidly than sulfate esters, and the sulfated monosaccharides were isolated and analyzed by HPLC on a CarboPak PA1 column as described previously(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). A single sulfated monosaccharide peak which comigrated with authentic GalNAc-4-SO4 standard was obtained from the assay product generated by purified submaxillary sulfotransferase (Fig. 3A), thus confirming the identity of this enzyme as a GalNAc-4-sulfotransferase. Sulfated monosaccharide generated using a partially-purified pituitary extract as the sulfotransferase source gave an identical pattern (Fig. 3B).
      Figure thumbnail gr3
      Figure 3:GalNAc-4-sulfotransferase product proof. Purified GalNAc-4-sulfotransferase (A) or partially-purified pituitary extract (B) was used to generate [35S]SO4-GGnM-MCO which was subjected to mild acid hydrolysis. Sulfated monosaccharides were analyzed by HPLC on a CarboPak PA1 column (Dionex). The elution positions of authentic standards are: 1, GlcNAc-3-SO4; 2, SO42-; 3, GalNAc-3-SO4; 4, GalNAc-4-SO4; 5, GlcNAc-6-SO4; 6, SO4-GGNM-MCO; 7, GalNAc-6-SO4.

      Photoaffinity Labeling with [32P]3′,5′-ADP

      [32P]3′,5′-ADP was used as a specific photoaffinity probe to confirm that the purified GalNAc-4-sulfotransferase activity corresponds to a protein of 128 kDa. Radiolabeled PAPS and PAPS analogs have been used as specific photolabeling probes in studies of several other sulfotransferases (
      • Otterness D.M.
      • Powers S.P.
      • Miller L.J.
      • Weinshilboum R.M.
      ) as well as in the identification and purification of the PAPS transporter(
      • Mandon E.C.
      • Milla M.E.
      • Kempner E.
      • Hirschberg C.B.
      ). 3′,5′-ADP is a PAPS analog which, in addition to being a by-product of the sulfotransferase reaction, is a potent competitive inhibitor of GalNAc-4-sulfotransferase activity (Ki = 2.4 μM). A band migrating at 128 kDa which is photolabeled with [32P]3′,5′-ADP copurifies with GalNAc-4-sulfotransferase activity during DEAE-Sepharose, WGA-Sepharose, and 3′,5′-ADP-agarose chromatography (Fig. 4A, lanes 2, 4, and 6). This labeling is UV-dependent, as control reactions done in the absence of UV do not show any radiolabeling (Fig. 4A, lanes 1, 3, and 5). In addition, the 128-kDa radiolabeled protein becomes enriched relative to other proteins which are also radiolabeled during this procedure (Fig. 4A, compare lanes 2 and 6). UV-dependent radiolabeling of the 128-kDa band can be competed by 1 mM PAPS but not by 1 mM ATP, demonstrating the specificity of the labeling reaction (Fig. 4B). These results imply that the 128-kDa labeled protein has a relatively high affinity for 3′,5′-ADP consistent with the identification of this protein as the GalNAc-4-sulfotransferase.
      Figure thumbnail gr4
      Figure 4:Photoaffinity labeling of a 128-kDa protein with [32P]3′,5′-ADP. A, composite autoradiogram of protein from several purification steps subjected to photolabeling with [32P]3′,5′-ADP. Lanes 1 and 2, 20 μg of DEAE-Sepharose eluate; Lanes 3 and 4, 20 μg of WGA-Sepharose eluate; Lanes 5 and 6, 2 μg of pooled fractions from 3′,5′-ADP-agarose NaCl gradient elution. Control reactions (Lanes 1, 3, and 5) were done in the absence of UV light. B, 20 μg of protein eluted from WGA-Sepharose was subjected to photolabeling in the absence of any nucleotide (Lanes 1 and 2), in the presence of 1 mM PAPS (Lane 3), or in the presence of 1 mM ATP (Lane 4). The reaction in Lane 1 was not subjected to UV. The 128-kDa labeled protein is indicated with an arrow.

      Acceptor Substrate Specificity of GalNAc-4-sulfotransferase

      The substrate specificity of the GalNAc-4-sulfotransferase was assessed using a series of oligosaccharide acceptors chemically and enzymatically synthesized as described previously(
      • Vandana
      • Hindsgaul O.
      • Baenziger J.U.
      ,
      • Srivastava V.
      • Hindsgaul O.
      ). Table II summarizes the results obtained using purified submaxillary gland transferase. Each of these synthetic oligosaccharides terminates with GalNAc in a β1,4-linkage to an underlying GlcNAc. The sequences underlying terminal GalNAcβ1,4GlcNAc are representative of the sequences observed in dibranched and tribranched Asn-linked oligosaccharides with the exception of the sequence GalNAcβ1,4GlcNAcβ1,6(Galβ1,3)GalNAc, which has been found O-glycosidically linked to the 16-kDa amino-terminal fragment of bovine proopiomelanocortin(
      • Siciliano R.A.
      • Morris H.R.
      • Bennett H.P.J.
      • Dell A.
      ). Previous results using crude bovine pituitary membrane extracts indicated that the trisaccharide GalNAcβ1,4GlcNAcβ1,2Manα (GGnM) contains sufficient information to allow recognition and transfer by the GalNAc-4-sulfotransferase(
      • Skelton T.P.
      • Hooper L.V.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). Purified submaxillary GalNAc-4-sulfotransferase has a Km of 15.0 μM for GGnM (Table II), in good agreement with our previous results. The submaxillary sulfotransferase has a Km of 31.1 μM for the terminal disaccharide GalNAcβ1,4GlcNAcβ (GGn), indicating that the underlying mannose is not essential for recognition or transfer of sulfate. The purified transferase has a Km of between 10 and 51 μM for each of the remaining acceptor oligosaccharides tested, all of which terminate with GGn. Up to 20 mM monomeric GalNAc did not inhibit GalNAc-4-sulfotransferase activity (not shown), consistent with the disaccharide GGn as the fundamental unit necessary for GalNAc-4-sulfotransferase recognition. Interestingly, GalNAc-4-sulfotransferase has a Km of 17.9 μM for GalNAcβ1,4GlcNAcβ1,6(Galβ1,3)GalNAc, an O-glycosidically linked structure(
      • Siciliano R.A.
      • Morris H.R.
      • Bennett H.P.J.
      • Dell A.
      ), as compared to a Km of 15.0 μM for GGnM. The GalNAc-4-sulfotransferase is, therefore, capable of recognizing and transferring sulfate to terminal β1,4-linked GalNAc on O-glycans as well as N-glycans.
      TABLE IIGalNAc-4-sulfotransferase acceptor substrate specificity
      Similar results were obtained when these acceptor oligosaccharides were tested using partially-purified bovine pituitary sulfotransferase as the enzyme source. The pituitary enzyme was found to have a Km of 19.2 and 42.5 μM for GGnM and GGn, respectively, and a Km of 18 μM for the O-linked structure GalNAcβ1,4GlcNAcβ1,6(Galβ1,3)GalNAc. The purified submaxillary sulfotransferase and partially-purified pituitary sulfotransferase have similar Km values for PAPS (1.1 μM for purified submaxillary transferase and 4 μM for pituitary). These and other data strongly suggest that the submaxillary and the pituitary GalNAc-4-sulfotransferase are the identical enzyme.

      DISCUSSION

      The pituitary glycoprotein hormone lutropin is crucial to the regulation of a number of physiological processes involved in reproduction. Essential to the expression of LH bioactivity is its pulsatile rise and fall in the bloodstream, which is in turn dependent on its ability to be rapidly cleared from the circulation(
      • Lincoln D.W.
      • Fraser H.M.
      • Lincoln G.A.
      • Martin G.B.
      • McNeilly A.S.
      ,
      • Crowley W.F.
      • Hofler J.G.
      ). Our laboratory has previously shown that this rapid clearance is attributable to the presence of unique sulfated oligosaccharides on LH and a GalNAc-4-SO4-specific receptor in hepatic endothelial cells(
      • Baenziger J.U.
      • Kumar S.
      • Brodbeck R.M.
      • Smith P.L.
      • Beranek M.C.
      ,
      • Fiete D.
      • Srivastava V.
      • Hindsgaul O.
      • Baenziger J.U.
      ). We have also shown that carbonic anhydrase VI, an abundant protein synthesized by the submaxillary gland and secreted into the saliva(
      • Fernley R.T.
      • Coghlan J.P.
      • Wright R.D.
      ), bears N-linked oligosaccharides which terminate with GalNAc-4-SO4(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ). We have now purified the sulfotransferase responsible for synthesis of terminal GalNAc-4-SO4 on the N-linked oligosaccharides of LH and carbonic anhydrase VI to apparent homogeneity from bovine submaxillary gland.
      In our purification scheme, two rounds of affinity chromatography on 3′,5′-ADP-agarose resulted in the isolation of a protein which migrates at 128 kDa on SDS-PAGE. The identification of this protein as the GalNAc-4-sulfotransferase is supported by experiments in which a 128-kDa protein was specifically photolabeled with [32P]3′,5′-ADP. The fact that a 1576-fold purification is sufficient to achieve homogeneity indicates that the submaxillary gland GalNAc-4-sulfotransferase is an abundant enzyme relative to other purified glycosyltransferases and sulfotransferases(
      • Brandan E.
      • Hirschberg C.B.
      ), and corresponds to our finding that submaxillary gland homogenates have a 15-fold higher specific activity than pituitary homogenates(
      • Dharmesh S.M.
      • Skelton T.P.
      • Baenziger J.U.
      ). This is not surprising given the relative abundance of carbonic anhydrase VI, which we have shown to be an endogenous substrate for the GalNAc-4-sulfotransferase in submaxillary gland(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ). The salivary gland-specific form of carbonic anhydrase, which is stored in granules and secreted into the saliva(
      • Feldstein J.B.
      • Silverman D.N.
      ,
      • Murakami H.
      • Sly W.S.
      ,
      • Fernley R.T.
      • Darling P.
      • Aldred P.
      • Wright R.D.
      • Coghlan J.P.
      ,
      • Parkkila S.
      • Kaunisto K.
      • Rajaniemi L.
      • Kumpulainen T.
      • Jokinen K.
      • Rajaniemi H.
      ,
      • Ogawa Y.
      • Chang C.-K.
      • Kuwakara H.
      • Hong S.-S.
      • Toyosawa S.
      • Yagi T.
      ), is expressed at very high levels(
      • Fernley R.T.
      • Coghlan J.P.
      • Wright R.D.
      ), and furthermore, greater than 50% of its Asn-linked oligosaccharides terminate with GalNAc-4-SO4(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ). Thus, the relative abundance of GalNAc-4-sulfotransferase in submaxillary gland correlates with the efficient sulfation of carbonic anhydrase VI oligosaccharides.
      Several lines of evidence indicate that the GalNAc-4-sulfotransferase purified from submaxillary gland is the same enzyme responsible for the synthesis of sulfated oligosaccharides on LH in the pituitary. 1) Analysis of sulfated monosaccharides derived from sulfotransferase assay products shows that both the submaxillary and the pituitary transferases are highly specific enzymes which transfer sulfate exclusively to the 4-hydroxyl of terminal β1,4-linked GalNAc. 2) Submaxillary and pituitary sulfotransferase have identical acceptor substrate specificities and virtually identical Kmvalues for these acceptors. Furthermore, the Kmvalues for PAPS are similar. 3) Carbonic anhydrase VI, an endogenous glycoprotein of the submaxillary gland, has been shown to bear oligosaccharides terminating with GalNAc-4-SO4(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ), thus demonstrating the in vivo presence of sulfated oligosaccharides in the submaxillary gland. 4) Carbonic anhydrase VI synthesized in bovine parotid glands, which express the GalNAc-transferase but not the GalNAc-4-sulfotransferase, bears oligosaccharides terminating with GalNAc rather than GalNAc-4-SO4(
      • Hooper L.V.
      • Beranek M.C.
      • Manzella S.M.
      • Baenziger J.U.
      ).
      Using a panel of synthetic oligosaccharide acceptors to investigate acceptor substrate specificity we have shown that the disaccharide GalNAcβ1,4GlcNAc contains sufficient information for recognition and transfer by the GalNAc-4-sulfotransferase, and thus can account for recognition of native oligosaccharide acceptors. We and others have previously shown that in addition to the pituitary glycoprotein hormones LH and thyroid-stimulating hormone, proopiomelanocortin synthesized in the pituitary also bears Asn-linked oligosaccharides terminating with GalNAc-4-SO4(
      • Skelton T.P.
      • Kumar S.
      • Smith P.L.
      • Beranek M.C.
      • Baenziger J.U.
      ,
      • Siciliano R.A.
      • Morris H.R.
      • McDowell R.A.
      • Azadi P.
      • Rogers M.E.
      • Bennett H.P.
      • Dell A.
      ). Other studies have also shown that the major O-glycan attached to Thr45 of the 16-kDa amino-terminal fragment of bovine proopiomelanocortin has the structure SO4-4-GalNAcβ1,4GlcNAcβ1,6(Galβ1,3)GalNAc(
      • Siciliano R.A.
      • Morris H.R.
      • Bennett H.P.J.
      • Dell A.
      ). Our results here show that both submaxillary and pituitary GalNAc-4-sulfotransferase can transfer sulfate to the structure GalNAcβ1,4GlcNAcβ1,6(Galβ1,3)GalNAc, demonstrating that this GalNAc-4-sulfotransferase can account for sulfation of both N-linked and O-linked glycans on proopiomelanocortin.
      Sulfate has been shown to occur in a number of different linkages and to different underlying sugars on N-linked oligosaccharides (
      • Freeze H.H.
      • Wolgast D.
      ,
      • Yamashita K.
      • Ueda I.
      • Kobata A.
      ,
      • Green E.D.
      • Boime I.
      • Baenziger J.U.
      ,
      • Spiro R.G.
      • Bhoyroo V.D.
      ,
      • Roux L.
      • Holojda S.
      • Sundblad G.
      • Freeze H.H.
      • Varki A.
      ). However, this is the first sulfotransferase to be isolated which is responsible for the sulfation of N-linked glycans. Obtaining the peptide sequence of this purified sulfotransferase will enable us to isolate its cDNA clone and compare its sequence to that of the N-heparan sulfotransferase, the only mammalian saccharide-specific sulfotransferase cloned to date(
      • Hashimoto Y.
      • Orellana A.
      • Gil G.
      • Hirschberg C.B.
      ). More importantly, cloning of the GalNAc-4-sulfotransferase will provide a crucial tool for investigating the regulation of sulfation of GalNAcβ1,4GlcNAc-bearing oligosaccharides on LH and other glycoproteins.

      REFERENCES

        • Lincoln D.W.
        • Fraser H.M.
        • Lincoln G.A.
        • Martin G.B.
        • McNeilly A.S.
        Recent Prog. Horm. Res. 1985; 41: 364-419
        • Crowley W.F.
        • Hofler J.G.
        Crowley W.F. Hofler J.G. The Episodic Secretion of Hormones. Wiley, New York1985: 121-235
        • Wang H.
        • Segaloff D.L.
        • Ascoli M.
        J. Biol. Chem. 1991; 266: 780-785
        • Veldhuis J.D.
        • Carlson M.L.
        • Johnson M.L.
        Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7686-7690
        • Evans W.S.
        • Sollenberger M.J.
        • Booth R.A.J.
        • Rogol A.D.
        • Urban R.J.
        • Carlsen E.C.
        • Johnson M.L.
        • Veldhuis J.D.
        Endocr. Rev. 1992; 13: 81-104
        • Green E.D.
        • Van Halbeek H.
        • Boime I.
        • Baenziger J.U.
        J. Biol. Chem. 1985; 260: 15623-15630
        • Green E.D.
        • Baenziger J.U.
        J. Biol. Chem. 1988; 263: 25-35
        • Green E.D.
        • Baenziger J.U.
        J. Biol. Chem. 1988; 263: 36-44
        • Smith P.L.
        • Bousfield G.S.
        • Kumar S.
        • Fiete D.
        • Baenziger J.U.
        J. Biol. Chem. 1993; 268: 795-802
        • Baenziger J.U.
        • Kumar S.
        • Brodbeck R.M.
        • Smith P.L.
        • Beranek M.C.
        Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 334-338
        • Fiete D.
        • Srivastava V.
        • Hindsgaul O.
        • Baenziger J.U.
        Cell. 1991; 67: 1103-1110
        • Smith P.L.
        • Baenziger J.U.
        Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 329-333
        • Mengeling B.J.
        • Manzella S.M.
        • Baenziger J.U.
        Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 502-506
        • Skelton T.P.
        • Hooper L.V.
        • Srivastava V.
        • Hindsgaul O.
        • Baenziger J.U.
        J. Biol. Chem. 1991; 266: 17142-17150
        • Dharmesh S.M.
        • Baenziger J.U.
        Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 11127-11131
        • Dharmesh S.M.
        • Skelton T.P.
        • Baenziger J.U.
        J. Biol. Chem. 1993; 268: 17096-17102
        • Hooper L.V.
        • Beranek M.C.
        • Manzella S.M.
        • Baenziger J.U.
        J. Biol. Chem. 1995; 270: 5985-5993
        • Skelton T.P.
        • Kumar S.
        • Smith P.L.
        • Beranek M.C.
        • Baenziger J.U.
        J. Biol. Chem. 1992; 267: 12998-13006
        • Vandana
        • Hindsgaul O.
        • Baenziger J.U.
        Can. J. Chem. 1987; 65: 1645-1652
        • Srivastava V.
        • Hindsgaul O.
        Carbohydr. Res. 1989; 185: 163-169
        • Schwyzer M.
        • Hill R.L.
        J. Biol. Chem. 1977; 252: 2338-2345
        • Palcic M.M.
        • Hindsgaul O.
        Glycobiology. 1991; 1: 205-209
        • Fernley R.T.
        • Coghlan J.P.
        • Wright R.D.
        Biochem. J. 1988; 249: 201-207
        • Feldstein J.B.
        • Silverman D.N.
        J. Biol. Chem. 1984; 259: 5447-5453
        • Murakami H.
        • Sly W.S.
        J. Biol. Chem. 1987; 262: 1382-1388
        • Otterness D.M.
        • Powers S.P.
        • Miller L.J.
        • Weinshilboum R.M.
        Mol. Pharmacol. 1991; 39: 34-41
        • Mandon E.C.
        • Milla M.E.
        • Kempner E.
        • Hirschberg C.B.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10707-10711
        • Siciliano R.A.
        • Morris H.R.
        • Bennett H.P.J.
        • Dell A.
        J. Biol. Chem. 1994; 269: 910-920
        • Brandan E.
        • Hirschberg C.B.
        J. Biol. Chem. 1988; 263: 2417-2422
        • Fernley R.T.
        • Darling P.
        • Aldred P.
        • Wright R.D.
        • Coghlan J.P.
        Biochem. J. 1989; 259: 91-96
        • Parkkila S.
        • Kaunisto K.
        • Rajaniemi L.
        • Kumpulainen T.
        • Jokinen K.
        • Rajaniemi H.
        J. Histochem. Cytochem. 1990; 38: 941-947
        • Ogawa Y.
        • Chang C.-K.
        • Kuwakara H.
        • Hong S.-S.
        • Toyosawa S.
        • Yagi T.
        J. Histochem. Cytochem. 1992; 40: 807-817
        • Siciliano R.A.
        • Morris H.R.
        • McDowell R.A.
        • Azadi P.
        • Rogers M.E.
        • Bennett H.P.
        • Dell A.
        Glycobiology. 1993; 3: 225-239
        • Freeze H.H.
        • Wolgast D.
        J. Biol. Chem. 1986; 261: 127-134
        • Yamashita K.
        • Ueda I.
        • Kobata A.
        J. Biol. Chem. 1983; 258: 14144-14147
        • Green E.D.
        • Boime I.
        • Baenziger J.U.
        Mol. Cell. Biochem. 1986; 72: 81-100
        • Spiro R.G.
        • Bhoyroo V.D.
        J. Biol. Chem. 1988; 263: 14351-14358
        • Roux L.
        • Holojda S.
        • Sundblad G.
        • Freeze H.H.
        • Varki A.
        J. Biol. Chem. 1988; 263: 8879-8889
        • Hashimoto Y.
        • Orellana A.
        • Gil G.
        • Hirschberg C.B.
        J. Biol. Chem. 1992; 267: 15744-15750