Portal de Periódicos da CAPES

Dynamic O-Glycosylation of Nuclear and Cytosolic Proteins

2001; Elsevier BV; Volume: 276; Issue: 13 Linguagem: Inglês

10.1074/jbc.m010420200

1083-351X

Yuan Gao, Lance Wells, Frank I. Comer, Glendon J. Parker, Gerald W. Hart,

Lysosomal Storage Disorders Research

Dynamic modification of cytoplasmic and nuclear proteins by O-linked N-acetylglucosamine (O-GlcNAc) on Ser/Thr residues is ubiquitous in higher eukaryotes and is analogous to protein phosphorylation. The enzyme for the addition of this modification, O-GlcNAc transferase, has been cloned from several species. Here, we have cloned a human brain O-GlcNAcase that cleaves O-GlcNAc off proteins. The cloned cDNA encodes a polypeptide of 916 amino acids with a predicted molecular mass of 103 kDa and a pI value of 4.63, but the protein migrates as a 130-kDa band on SDS-polyacrylamide gel electrophoresis. The cloned O-GlcNAcase has a pH optimum of 5.5–7.0 and is inhibited by GlcNAc but not by GalNAc.p-Nitrophenyl (pNP)-β-GlcNAc, but notpNP-β-GalNAc or pNP-α-GlcNAc, is a substrate. The cloned enzyme cleaves GlcNAc, but not GalNAc, from glycopeptides. Cell fractionation suggests that the overexpressed protein is mostly localized in the cytoplasm. It therefore has all the expected characteristics of O-GlcNAcase and is distinct from lysosomal hexosaminidases. Northern blots show that the transcript is expressed in every human tissue examined but is the highest in the brain, placenta, and pancreas. An understanding of O-GlcNAc dynamics and O-GlcNAcase may be key to elucidating the relationships between O-phosphate and O-GlcNAc and to the understanding of the molecular mechanisms of diseases such as diabetes, cancer, and neurodegeneration. Dynamic modification of cytoplasmic and nuclear proteins by O-linked N-acetylglucosamine (O-GlcNAc) on Ser/Thr residues is ubiquitous in higher eukaryotes and is analogous to protein phosphorylation. The enzyme for the addition of this modification, O-GlcNAc transferase, has been cloned from several species. Here, we have cloned a human brain O-GlcNAcase that cleaves O-GlcNAc off proteins. The cloned cDNA encodes a polypeptide of 916 amino acids with a predicted molecular mass of 103 kDa and a pI value of 4.63, but the protein migrates as a 130-kDa band on SDS-polyacrylamide gel electrophoresis. The cloned O-GlcNAcase has a pH optimum of 5.5–7.0 and is inhibited by GlcNAc but not by GalNAc.p-Nitrophenyl (pNP)-β-GlcNAc, but notpNP-β-GalNAc or pNP-α-GlcNAc, is a substrate. The cloned enzyme cleaves GlcNAc, but not GalNAc, from glycopeptides. Cell fractionation suggests that the overexpressed protein is mostly localized in the cytoplasm. It therefore has all the expected characteristics of O-GlcNAcase and is distinct from lysosomal hexosaminidases. Northern blots show that the transcript is expressed in every human tissue examined but is the highest in the brain, placenta, and pancreas. An understanding of O-GlcNAc dynamics and O-GlcNAcase may be key to elucidating the relationships between O-phosphate and O-GlcNAc and to the understanding of the molecular mechanisms of diseases such as diabetes, cancer, and neurodegeneration. O-linked N-acetylglucosamine N-acetylgalactosamine N-acetylglucosamine O-GlcNAc transferase N-acetyl-β-d-glucosaminidase polyacrylamide gel electrophoresis polymerase chain reaction p-nitrophenyl O-(2-acetamido-2-deoxy-d-glucopyranosylidene)-amino-N-phenylcarbamate base pair(s) C-terminal domain kilobase(s) expressed sequence tag matrix-assisted laser desorption ionization-time of flight mass spectrometry concanavalin A phenylmethylsulfonyl fluroride Since the description of O-linkedN-acetylglucosamine (O-GlcNAc)1 as an abundant modification in murine lymphocytes (1Torres C.R. Hart G.W. J. Biol. Chem. 1984; 259: 3308-3317Abstract Full Text PDF PubMed Google Scholar), a myriad of cytoplasmic and nuclear proteins in all metazoans have been found to carry this modification. Such proteins cover a broad range, including many transcription factors, RNA polymerase II, oncogenes, nuclear pore proteins, viral proteins, and tumor repressors (for details, see Refs.2Hart G.W. Annu. Rev. Biochem. 1997; 66: 315-335Crossref PubMed Scopus (455) Google Scholar and 3Snow D.M. Hart G.W. Int. Rev. Cytol. 1998; 181: 43-74Crossref PubMed Google Scholar and citations within). Unlike classic O- orN-linked protein glycosylations, the O-GlcNAc modification involves only a single GlcNAc moiety linked to the hydroxyl group of Ser/Thr residues, generally is not elongated, and is found exclusively in the cytoplasm and nucleoplasm. Protein O-GlcNAcylation is highly dynamic, and the cycle of addition/removal of the sugar moiety is rapid, analogous to protein phosphorylation/dephosphorylation catalyzed by kinases and phosphatases (2Hart G.W. Annu. Rev. Biochem. 1997; 66: 315-335Crossref PubMed Scopus (455) Google Scholar). Indeed, existing evidence suggests that this modification has a "yin-yang" relationship with protein phosphorylation in some cases (4Comer F.I. Hart G.W. J. Biol. Chem. 2000; 275: 29179-29182Abstract Full Text Full Text PDF PubMed Scopus (328) Google Scholar). Many O-GlcNAcylation sites have been mapped to phosphorylation sites or adjacent sites (5Chou T.Y. Hart G.W. Dang C.V. J. Biol. Chem. 1995; 270: 18961-18965Abstract Full Text Full Text PDF PubMed Scopus (365) Google Scholar, 6Cole R.N. Hart G.W. J. Neurochem. 1999; 73: 418-428Crossref PubMed Scopus (95) Google Scholar, 7Cheng X. Cole R.N. Zaia J. Hart G.W. Biochemistry. 2000; 39: 11609-11620Crossref PubMed Scopus (152) Google Scholar). Such spatial localization indicates that O-GlcNAc may regulate the target protein by competing with protein kinases (4Comer F.I. Hart G.W. J. Biol. Chem. 2000; 275: 29179-29182Abstract Full Text Full Text PDF PubMed Scopus (328) Google Scholar). Recent studies using phosphatase and kinase inhibitors have provided direct evidence for a general reciprocal relationship between O-GlcNAcylation and phosphorylation on some proteins (8Griffith L.S. Schmitz B. Eur. J. Biochem. 1999; 262: 824-831Crossref PubMed Scopus (96) Google Scholar, 9Lefebvre T. Alonso C. Mahboub S. Dupire M.J. Zanetta J.P. Caillet-Boudin M.L. Michalski J.C. Biochim. Biophys. Acta. 1999; 1472: 71-81Crossref PubMed Scopus (61) Google Scholar). O-GlcNAcylation appears to be involved in gene transcription. Most transcription factors examined so far, including Sp1, AP1, AP2, AP4 (10Jackson S.P. Tjian R. Cell. 1988; 55: 125-133Abstract Full Text PDF PubMed Scopus (650) Google Scholar), serum response factor (11Reason A.J. Morris H.R. Panico M. Marais R. Treisman R.H. Haltiwanger R.S. Hart G.W. Kelly W.G. Dell A. J. Biol. Chem. 1992; 267: 16911-16921Abstract Full Text PDF PubMed Google Scholar), the estrogen receptor (7Cheng X. Cole R.N. Zaia J. Hart G.W. Biochemistry. 2000; 39: 11609-11620Crossref PubMed Scopus (152) Google Scholar, 12Jiang M.S. Hart G.W. J. Biol. Chem. 1997; 272: 2421-2428Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar), the insulin promoter factor-1, 2Y. Gao and G. W. Hart, manuscript in preparation. and peroxisome proliferator-activated receptor-γ, as well as RNA polymerase II (13Kelly W.G. Dahmus M.E. Hart G.W. J. Biol. Chem. 1993; 268: 10416-10424Abstract Full Text PDF PubMed Google Scholar) and chromatin (14Kelly W.G. Hart G.W. Cell. 1989; 57: 243-251Abstract Full Text PDF PubMed Scopus (128) Google Scholar) are O-GlcNAcylated.O-GlcNAcylation of Sp1 appears to enhance its activity in transcription, and, conversely, blocking the GlcNAc residues with lectin wheat germ agglutinin suppresses the transcriptional activity (10Jackson S.P. Tjian R. Cell. 1988; 55: 125-133Abstract Full Text PDF PubMed Scopus (650) Google Scholar). O-GlcNAcylation of Sp1 also controls its degradation by the proteasome (15Han I. Kudlow J.E. Mol. Cell. Biol. 1997; 17: 2550-2558Crossref PubMed Scopus (377) Google Scholar). Hyperglycemia-induced superoxide production increases Sp1 glycosylation resulting in the activation of genes that contribute to the pathogenesis of diabetes (16Du W.L. Edelstein D. Rossetti L. Fantus I.G. Goldberg H. Ziyadeh F. Wu J. Brownlee M. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 12222-12226Crossref PubMed Scopus (900) Google Scholar). O-GlcNAc transferase (OGT), which transfers GlcNAc from the donor substrate UDP-GlcNAc to target proteins, has been purified and cloned from several species including human, rat, andCaenorhabditis elegans (17Haltiwanger R.S. Blomberg M.A. Hart G.W. J. Biol. Chem. 1992; 267: 9005-9013Abstract Full Text PDF PubMed Google Scholar, 18Lubas W.A. Frank D.W. Krause M. Hanover J.A. J. Biol. Chem. 1997; 272: 9316-9324Abstract Full Text Full Text PDF PubMed Scopus (421) Google Scholar, 19Kreppel L.K. Blomberg M.A. Hart G.W. J. Biol. Chem. 1997; 272: 9308-9315Abstract Full Text Full Text PDF PubMed Scopus (612) Google Scholar). It does not share any significant homology with any other known proteins, including glycosyltransferases, and is highly conserved from C. elegans to human. Disruption of the ogt gene is lethal in mouse embryonic stem cells, further underscoring the importance ofO-GlcNAc modification in cellular functions (20Shafi R. Iyer S.P. Ellies L.G. O'Donnell N. Marek K.W. Chui D. Hart G.W. Marth J.D. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5735-5739Crossref PubMed Scopus (605) Google Scholar).O-GlcNAcase, the enzyme that removes O-GlcNAc from such proteins, was purified several years ago from rat spleen (21Dong D.L. Hart G.W. J. Biol. Chem. 1994; 269: 19321-19330Abstract Full Text PDF PubMed Google Scholar). It is a neutral cytosolic β-glucosaminidase or hexosaminidase C (EC 3.2.1.52). To further study the function of this modification, we have now extensively purified O-GlcNAcase from bovine brain, sequenced the protein by mass spectrometry, and cloned the cDNA. The O-GlcNAcase is evolutionarily conserved, distinct from lysosomal acidic hexosaminidases A and B. The recombinant protein has all the expected characteristics of O-GlcNAcase, including the ability to cleave O-GlcNAc from glycopeptides. All chromatographic materials were purchased from Amersham Pharmacia Biotech (Piscataway, NJ). Purification procedure is substantially modified from that of Dong and Hart (21Dong D.L. Hart G.W. J. Biol. Chem. 1994; 269: 19321-19330Abstract Full Text PDF PubMed Google Scholar). All steps were conducted at 4 °C or on ice. Three bovine brains (∼1 kg) were frozen in liquid N2 and shipped on dry ice from Pel Freez Biologicals (Rogers, AR) and stored at −80 °C until use. The brains were smashed into smaller pieces and homogenized in 5 volumes (v/w) of homogenization buffer (20 mm sodium phosphate, pH 7.5, 15 mm 2-mecaptoethanol, 10 mmMgCl2, 1 mm PMSF, 1 mm EDTA) in a Hamilton Beach blender with 5 × 20-s bursts. The homogenate was centrifuged at 18,000 × g for 30 min. The pellet was discarded, and the cytosolic supernatant was pooled in a 5-liter beaker. The cytosolic supernatant was subjected to 30–50% ammonium sulfate precipitation. The pellet was resuspended in 500 ml of buffer A (20 mmsodium phosphate, pH 7.5, 5 mm 2-mercaptoethanol) and centrifuged to clarify the solution. The solution was then thoroughly dialyzed against buffer A and centrifuged again to eliminate any insoluble materials that had resulted from dialysis. The dialyzed sample was loaded onto a DE52 column (900-ml bed vol) at a flow rate of 2 ml/min using a peristaltic pump. After washing the column with 3 liters of buffer A, bound proteins were eluted with a linear gradient of 0–1 m NaCl in 4 liters of buffer A at a flow rate of 4 ml/min. The protein profile was monitored by absorbance at 280 nm. Fractions (16 ml) enriched in O-GlcNAcase activity were pooled. MgCl2 was added to the pooled fractions at a final concentration of 1 mm. The preparation was then applied to a ConA column (60 ml) equilibrated in ConA buffer (20 mm sodium phosphate, pH 7.5, 5 mm2-mercapoethanol, 150 mm NaCl, 1 mmMgCl2). The column was washed with 200 ml of ConA buffer. The flow-through and the wash were combined. The enzyme solution from Step 4 was concentrated by 60% ammonium sulfate precipitation, dialyzed, and applied three times to a Blue A-Sepharose column (25 ml) equilibrated in buffer A. Again the activity was present in the flow-through fraction. The protein was pooled and clarified by centrifugation. The sample from Step 5 was injected to the DE52 cellulose column (same size as above), and protein was eluted with a linear gradient of 50–350 mmNaCl in 4 liters of buffer A. Activity was recovered as in Step 3 and precipitated with 60% ammonium sulfate. The pellet was resuspended in 20 ml of Mono-Q buffer (20 mm Tris, pH 7.5, 5 mm 2-mercaptoethanol, 10% glycerol, 1 mm EDTA plus protease inhibitor mixture (22Roquemore E.P. Chou T.Y. Hart G.W. Methods Enzymol. 1994; 230: 443-460Crossref PubMed Scopus (130) Google Scholar) and 1 mm PMSF), dialyzed, and clarified by centrifugation. Native PAGE was performed using a preparative Prepcell apparatus (Bio-Rad, Hercules, CA). The sample from Step 6 was divided into three equal volumes (45 mg of protein each) and loaded batch-wise onto a 6% native polyacrylamide gel (5-cm-long separating gel). The gel was run for 24 h at 12 watts of constant power. Protein was eluted in Mono-Q buffer at a flow rate of 0.75 ml/min. 5-min fractions were collected and assayed for protein content and enzyme activity. TheO-GlcNAcase-containing fractions from each native PAGE run was resolved on a Mono-Q column (HR10/10) with a linear gradient of 0–500 mm NaCl in 450 ml of Mono-Q buffer at a flow rate of 3 ml/min. Fractions (4.5 ml) rich inO-GlcNAcase activity were pooled and then separated for a second time on the Mono-Q column. The final preparation was concentrated using Millipore concentrators to a final volume of 0.4 ml. Glycerol (40% final) and 1 mm PMSF and protease inhibitor mixture were added to the preparation. The enzyme was stored at −20 °C. The final preparation from Step 8 was separated by 10% SDS-PAGE and stained with Coomassie Blue G-250 or with silver. The desired protein bands were excised individually, reduced, alkylated, and digested in-gel with modified trypsin (Worthington, Freehold, NJ) as described previously (23Gharahdaghi F. Weinberg C.R. Meagher D.A. Imai B.S. Mische S.M. Electrophoresis. 1999; 20: 601-605Crossref PubMed Scopus (842) Google Scholar). The tryptic peptides from each protein were analyzed by capillary reversed phase high pressure liquid chromatography with in-line tandem MS/MS on a Finnigan LCQ ion-trap mass spectrometer. Proteins were identified by the SEQUEST algorithm with sequencing of at least seven tryptic peptides for each protein (24Ducret, A., Van, O.ostveen, I., Eng, J. K., Yates, J. R., and Aebersold, R. (1998) Protein Sci. 7, 706–719.Google Scholar). A putative O-GlcNAcase with a theoretical length of 916 amino acids in human was identified by the above proteomic approach. A cDNA fragment, KIAA0679 (GenBank™ accession number AB014579), which contains the coding sequence for 767 amino acids of the C terminus of the humanO-GlcNAcase and a 2.0-kb 3′-untranslated region, was obtained from the Kazusa DNA Research Institute, Japan, in the vector pBluescript. The coding sequence of this fragment was subsequently transferred to pcDNA3.1His A using XhoI andXbaI, which were located within the polycloning cloning site of pBluescript and in the 3′-untranslated region of the cDNA, respectively. The missing 5′-end fragment of the full-length coding cDNA (447 bp) was amplified by PCR from a human brain Marathon cDNA library (CLONTECH, Palo Alto, CA) using the forward primer GGATGGTGCAGAAGGAGAGTCAAGCGAC and the reverse primer TAGAAACCTCTTCGATGGACTCTACTGG. The forward primer sequence was based on published data (25Heckel D. Comtesse N. Brass N. Blin N. Zang K.D. Meese E. Hum. Mol. Genet. 1998; 7: 1859-1872Crossref PubMed Scopus (106) Google Scholar), and the reverse primer was located in the KIAA0679 clone. PCR conditions were 94 °C for 30 s, 63 °C for 30 s, and 72 °C for 3 min for 30–35 cycles. A second round of PCR using the first PCR product as template and a forward primer incorporating a NotI site (CCGGGCGGCCGCGGATGGTGCAGAAGGAGAG) and the same reverse primer was performed. The product was digested with NotI andHindIII (unique site in the PCR product) and ligated in-frame into the pcDNA3.1His A-KIAA0679 construct. This gave rise to a full-length cDNA in the vector pcDNA3.1His A. The final construct was sequenced. Unless stated otherwise,O-GlcNAcase activity was assayed as described in 50 mm sodium cacodylate, pH 6.5, 0.3% bovine serum albumin, 2 mm pNP-β-GlcNAc, 50 mm GalNAc (21Dong D.L. Hart G.W. J. Biol. Chem. 1994; 269: 19321-19330Abstract Full Text PDF PubMed Google Scholar). Purified bovine kidney lysosomal β-hexosaminidase (Roche Molecular Biochemicals, Indianapolis, IN) activity was assayed in citrate phosphate buffer, pH 4.5, 0.3% bovine serum albumin, 2 mm pNP-β-GlcNAc. To test the ability of recombinant O-GlcNAcase to cleave O-GlcNAc from glycopeptides, two glycopeptides, CTD-GlcNAc (N-YSPTS(GlcNAc)PSK-C) or CTD-GalNAc (N-YSPTS(GalNAc)PSK-C), were synthesized by standard Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry. The peptides were purified on a C18 column under reversed phase high pressure liquid chromatography conditions and used as a substrate for cloned O-GlcNAcase. The reaction products were analyzed by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF). A plasmid of pcDNA3.1His A containing the full-lengthO-GlcNAcase cDNA was prepared using a Qiagen kit (Qiagen, Valencia, CA). Transfection was mediated by LipofectAMINE Plus (Life Technologies Inc., Gaithersburg, MD) using 50–90% confluent Cos-7 cells. Cells were harvested 2 days post-transfection and sonicated for 2 × 12 s in 20 mm Tris (pH 7.5), 10% glycerol, 150 mm NaCl, 1 mmdithiothreitol, 0.1 mm EDTA, 1 mm PMSF and protease inhibitor mixture, and clarified by centrifugation. For characterizations, the recombinant protein was purified over a nickel affinity column. After transfection withO- GlcNAcase, Cos-7 cells were separated into cytoplasmic and nuclear fractions as described (26Schreiber E. Matthias P. Muller M.M. Schaffner W. Nucleic Acids Res. 1989; 17: 6419Crossref PubMed Scopus (3918) Google Scholar) with one modification: The 25- to 50-μl nuclear pellet was carefully washed in 500 μl of hypotonic buffer A to minimize cross-contamination. Immunoblot analysis was performed using antibodies recognizing the nuclear protein retinoblastoma (Rb) (Santa Cruz Biotechnology, Santa Cruz, CA), cytoplasmic protein α-tubulin (Sigma Chemical Co., St. Louis, MO), or with anti-Xpress antibody, which is specific for the sequence DLYDDDDK located at the N terminus of the overexpressed O-GlcNAcase fusion protein (Invitrogen, Carlsbad, CA). Northern blot analyses were performed on a human multiple tissue Northern blot (CLONTECH) using the manufacturer's protocol. To prepare anO-GlcNAcase-specific probe, the full-length coding sequence (2.75 kb) was amplified by PCR and labeled by random primer using [α-32P]dCTP (Stratagene, La Jolla, CA). After stripping in 0.5% SDS at 100 °C for 10 min, the blot was reprobed for β-actin. Historically,O- GlcNAcase, or neutral hexosaminidase C, has been difficult to purify. For example, an early report described a purification of only 25- to 40-fold from bovine brain despite the extensive use of chromatographic steps (27Overdijk B. van der Kroef W.M.J. van Steijn G.J. Lisman J.W. Biochim. Biophys. Acta. 1981; 659: 255-266Crossref PubMed Scopus (29) Google Scholar). However, in rat brain, the enzyme has been purified over 2000-fold to a major band (28Izumi T. Suzuki K. J. Biol. Chem. 1983; 258: 6991-6999Abstract Full Text PDF PubMed Google Scholar). More recently, renewed effort has gone into its purification from rat spleen and bovine brain in the pursuit to cloning the cDNA (21Dong D.L. Hart G.W. J. Biol. Chem. 1994; 269: 19321-19330Abstract Full Text PDF PubMed Google Scholar, 29Hanover J.A. Lai Z. Lee G. Lubas W.A. Sato S.M. Arch. Biochem. Biophys. 1999; 362: 38-45Crossref PubMed Scopus (114) Google Scholar). We have taken an approach, partly based on published literature (21Dong D.L. Hart G.W. J. Biol. Chem. 1994; 269: 19321-19330Abstract Full Text PDF PubMed Google Scholar) but yet incorporating some novel steps in the purification ofO-GlcNAcase from bovine brain. Notably, we have discovered that the enzymatic activity survives the harsh conditions of native PAGE (high pH and high ionic strength) and migrates more slowly (RF = 0.28 in 6% native gel) than most other proteins in the gel (data not shown). This property allows the effective separation of O-GlcNAcase from other proteins with higher RF values on a preparative scale native gel (Fig. 1 b). This step, in conjunction with other chromatographic steps outlined in the protocol, purified the protein ∼1500-fold, with a specific activity of 1840 nmol/min/mg of protein. The final preparation still shows seven well defined bands on SDS-PAGE following silver staining, even after extensive purification (Fig.2). We do not understand the basis for this difficulty, but we have observed that the peaks forO-GlcNAcase activity are very broad throughout the purification procedure (Fig. 1). One example of this is illustrated in the Mono-Q step, where the general protein peaks are sharp but yet the activity peak spreads over 50 ml (Fig. 1 c). We have also tried ion exchange on Superose Q and hydroxylapatite columns or hydrophobic interaction chromatography on a phenyl-Sepharose column. They, too, give poor separations or the enzyme binds very tightly to phenyl-Sepharose resulting in >50% activity loss (data not shown). The seven bands on the silver-stained SDS-PAGE gel were excised individually, digested with trypsin, and sequenced by electrospray MS/MS. The fragmentation data were used to search protein and DNA data bases. This approach identified six proteins with known functions in six of the bands (Fig. 2). Another protein, which runs as a 130-kDa band on SDS-PAGE, is a hypothetical protein without any clearly defined functions (GenBank, KIAA0679). BLAST searches indicated that the hypothetical protein shared significant homology with a protein from C. elegans called "similar to hyaluronoglucosaminidase" (GenBank, AAA68333.1). Because hyaluro- noglucosaminidase degrades hyaluronic acid, which is a GlcNAcβ1–4GlcUA polymer, it was possible that a hyaluronoglucosaminidase may share some homology withO-GlcNAcase. Furthermore, careful comparisons ofO-GlcNAcase activity and protein patterns on the SDS gels of different pools during the purification procedure indicated that this protein was one of only two bands that corresponded with activity (the other band was Protein 1 in Fig. 2, data not shown). We therefore hypothesized that this may be the O-GlcNAcase, and cloned the cDNA. Further characterization of the expression product of the cDNA confirmed that band 2 on Fig. 2 was, indeed,O-GlcNAcase (see below). BLAST searches of data bases reveal that O-GlcNAcase is conserved in higher eukaryotic species, and the homologue is absent in yeast or prokaryotes. The sequences and alignment of O-GlcNAcase from human, C. elegans, and Drosophila are shown in Fig. 3. In a pairwise alignment, the human sequence shares 55 and 43% homology with that ofDrosophila and C. elegans, respectively, whereasDrosophila and C. elegans are 43% similar. Close inspection of the sequences indicates that the N-terminal ∼400 and the C-terminal ∼350 amino acids in the human sequence are conserved to a higher degree. These two domains are separated by a highly variable region of ∼150 amino acids. Another feature is that most of the aromatic residues are conserved among the species. For example, out of the 13 Trp residues found in the human sequence, 9 are invariant in Drosophila and C. elegans, two are conservative (substituted by Tyr or Phe), and only two are variable. The O-GlcNAcase sequence is conserved at a strikingly higher level in mammals. Four overlapping EST sequences from cow, which cover 46% of the human protein, show that these two species are 100% identical in these regions (BE481597, BE588694, BF043559, andAW463869). Five EST entries for mouse, most of which are overlapping, show that human and mouse are 97.8% identical (AW907793, AW324047,AI530529, AW762257, and AA240394). In the case of zebrafish, two overlapping EST sequences covering 33.8% of the human protein indicate that zebrafish and human are 85% identical and 92% similar (AI882982and AI722710). Apart from the above-described homologues, O-GlcNAcase does not show significant homology with any other proteins, including known glycosidases. Short stretches of ∼200 amino acids of the polypeptide do show loose homology to a number of proteins such as hyaluronidase (AAA23259.1), a putative acetyltransferase (AL158057), eukaryotic translation elongation factor-1γ (Z11531, S26649), and the 11-1 polypeptide (X07453, S00485). Sequence analyses by a computer program PSORT II (available on the Web) show thatO-GlcNAcase does not possess any known signal peptides, domains, or motifs. The analyses do, however, suggest that the endogenous protein is localized in the cytoplasm (p = 0.522) and the nucleus (p = 0.391). To ascertain that the cloned cDNA indeed encoded O-GlcNAcase, we subcloned the entire coding region in-frame into the mammalian expression vector pcDNA3.1His and overexpressed for activity in Cos-7 cells. Transient transfection resulted in a 6-fold increase inO-GlcNAcase activity over endogenous activity in the cells (Fig. 4 a). After nickel affinity purification, the activity from theO-GlcNAcase-transfected cells was 230 nmol/min/mg of protein but was not detectable from control transfected cells (Fig.4 b). These data show that the activity is due to overexpression from the plasmid. Fig. 4 c shows that a distinct band of the correct molecular mass (135 kDa) was isolated after nickel purification from transfected cells. This band was immunoreactive with the Xpress antibody, which was specific for a peptide sequence in the overexpressed protein. We further characterized the properties of the cloned O-GlcNAcase and compared them with those of lysosomal β-hexosaminidase purified from bovine kidney. As expected, the lysosomal β-hexosaminidase had an acidic pH optimum (pH 3.5–5.5) with little activity at pH 7.0 or above (Fig.5 a). On the other hand, the cloned O-GlcNAcase had a pH optimum of 5.7–7.0 and retained significant activity (∼30%) at pH 7–8. This pH profile is consistent with the expected localization of O-GlcNAcase in the cytoplasm and the nucleus. The two enzymes also responded differently to inhibitors. GalNAc, a widely used inhibitor of acidic β-hexosaminidase, inhibited the lysosomal enzyme 50% at 5.0 mm and 88% at 50 mm. The cloned O-GlcNAcase was not inhibited at all by GalNAc up to 50 mm (Fig. 5 b). GlcNAc and its synthetic analogue PUGNAc inhibited both enzymes but were more potent with the O-GlcNAcase (Fig. 5, c andd). O-GlcNAcase also differed from lysosomal β-hexosaminidase in substrate requirements. In the in vitro assays, purified recombinant O-GlcNAcase cleaved only pNP-β-GlcNAc, but not pNP-β-GalNAc orpNP-α-GlcNAc (Fig. 6). The activity using the latter two compounds as substrates was not detectable. This substrate specificity was in contrast to the lysosomal β-hexosaminidase, which also cleaved pNP-β-GalNAc, albeit with slightly lower efficiency compared withpNP-β-GlcNAc. A trueO-GlcNAcase should cleave O-GlcNAc attached to proteins or peptides. We synthesized two glycopeptides containing one repeat of the (CTD) C- terminaldomain of RNA polymerase II linked to β-GlcNAc or α-GalNAc through the hydroxyl group of a serine residue (CTD-GlcNAc or CTD-GalNAc). The design of these glycopeptides is based on earlier information that this serine residue is glycosylated in vivo(13Kelly W.G. Dahmus M.E. Hart G.W. J. Biol. Chem. 1993; 268: 10416-10424Abstract Full Text PDF PubMed Google Scholar). These peptides were tested as substrates for the purified recombinant O-GlcNAcase, and the product peptides were analyzed by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF). Cleavage of GlcNAc or GalNAc from the peptides should result in a downshift of 203 in the molecular weight (the weight of GlcNAc or GalNAc minus 18). O-GlcNAcase did not cleave GalNAc from the peptide (Fig. 7, a andb) but did successfully cleave GlcNAc from the peptide, as judged by the expected shift in molecular weight (1069.8 to 866.6, 1091.8 to 888.6) (Fig. 7, c and d). As stated earlier, sequence analyses suggest that O-GlcNAcase is localized in the cytoplasm and nucleus. This is consistent with the localization of the O-GlcNAc modification. To obtain direct evidence on its localization, we performed cellular fractionation and assayed O-GlcNAcase activity in the cytoplasm and the nucleus. The data show that, in nontransfected cells,O-GlcNAcase activity was distributed in both the cytoplasm and the nucleus. However, when O-GlcNAcase was overexpressed, it was predominantly found in the cytoplasm (Fig.8 a). We also probed for its localization by Western blots (Fig. 8 b). Retinoblastoma protein (Rb) and α-tubulin, which were exclusively localized in the nucleus and cytoplasm, respectively, were used as markers. In agreement with activity assays, overexpressed O-GlcNAcase protein was only detected in the cytoplasm of overexpressed cells. To estimate the number of transcripts of O-GlcNAcase and their levels of expression, we performed Northern analyses using a human multiple tissue blot consisting of RNA from eight different tissues. Labeled PCR product of the entire coding region of the O-GlcNAcase cDNA was used as a probe. The Northern blot analysis showed only one transcript of ∼5.5 kb for O-GlcNAcase (Fig.9). Exposing the film for extended time (32 h) did not reveal any additional bands (data not shown). The gene was expressed in every tissue on the blot but was the highest in the brain, followed by placenta, and pancreas. Lung and liver had the lowest expression (Fig. 9). This pattern of expr