HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 36, 27733-27740, September 8, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/36/27733    most recent M002990200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Wei, G. Articles by Esko, J. D. Search for Related Content PubMed PubMed Citation Articles by Wei, G. Articles by Esko, J. D. Social Bookmarking                      What's this?

Location of the Glucuronosyltransferase Domain in the Heparan Sulfate Copolymerase EXT1 by Analysis of Chinese Hamster Ovary Cell Mutants^*

Ge Wei^§, Xiaomei Bai, Mary M. G. Gabb, Karen J. Bame^¶, Thomas I. Koshy^**, Patricia G. Spear, and Jeffrey D. Esko^§§

From the  Department of Cellular and Molecular Medicine, Glycobiology Research and Training Program, University of California, San Diego, La Jolla, California, 92093-0687, ^¶ Division of Molecular Biology and Biochemistry, School of Biological Sciences, University of Missouri-Kansas City, Kansas City, Missouri 64110, and the  Department of Microbiology-Immunology, Northwestern University Medical School, Chicago, Illinois 60611

Received for publication, April 9, 2000, and in revised form, June 19, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Heparan sulfate formation occurs by the copolymerization of glucuronic acid (GlcA) and N-acetylglucosamine (GlcNAc) residues. Recent studies have shown that these reactions are catalyzed by a copolymerase encoded by EXT1 and EXT2, members of the exostosin family of putative tumor suppressors linked to hereditary multiple exostoses. Previously, we identified a collection of Chinese hamster ovary cell mutants (pgsD) that failed to make heparan sulfate (Lidholt, K., Weinke, J. L., Kiser, C. S., Lugemwa, F. N., Bame, K. J., Cheifetz, S., Massagué, J., Lindahl, U., and Esko, J. D. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 2267-2271). Here, we show that pgsD mutants contain mutations that either alter GlcA transferase activity selectively or that affect both GlcNAc and GlcA transferase activities. Expression of EXT1 corrects the deficiencies in the mutants, whereas EXT2 and the related EXT-like cDNAs do not. Analysis of the EXT1 mutant alleles revealed clustered missense mutations in a domain that included a (D/E)X(D/E) motif thought to bind the nucleotide sugar from studies of other transferases. These findings provide insight into the location of the GlcA transferase subdomain of the enzyme and indicate that loss of the GlcA transferase domain may be sufficient to cause hereditary multiple exostoses.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Heparan sulfate functions in a wide variety of events during cell differentiation and tissue morphogenesis by binding and activating a number of growth factors, adhesion molecules, and enzymes (1, 2). Its assembly initiates by the formation of an oligosaccharide primer consisting of -GlcNAc1-3GlcA1-3Gal1-3Gal1-4Xyl-O- attached to specific serine residues in a proteoglycan core protein (3). This primer serves as a substrate for the copolymerase that adds alternating 1-4-linked GlcA¹ and 1-4-linked GlcNAc residues, resulting in chains of ~100 or more sugar units in length. Over the years, several groups have studied these reactions in vitro with the aim of defining the specificity and overall characteristics of the transferase(s) (early literature reviewed in Ref. 4; also see Refs. 5-7). Several years ago, a set of heparan sulfate-deficient mutants of Chinese hamster ovary cells were isolated, and through indirect genetic analysis, it became clear that the two transferases were in fact encoded by a single genetic locus, designated pgsD (8). Subsequently, the copolymerase was purified to near homogeneity and the isolated 70-kDa protein was shown to have both GlcNAc and GlcA transferase activities (9). These genetic and biochemical findings indicated that the copolymerase appeared to be a bifunctional protein with two active sites.

Hereditary multiple exostosis (HME) is a dominant genetic disorder (10) resulting in the formation of generally benign cartilage-capped outgrowths from various bones (11-13). Positional cloning experiments have identified three relevant genetic loci, designated EXT1, EXT2, and EXT3 located at 8q24.1, 11p11-13, and 19p, respectively (14-16). By primary DNA sequence, EXT1 and EXT2 exhibit about 35% homology and are predicted to encode type II transmembrane proteins. In addition, three other loci designated EXTL1/R3, EXTL2, and EXTL3/R1 have been identified by hybridization and sequence homology searches (17-20), but no clinical cases of HME have been mapped to these loci. Mutations at EXT1 and EXT2 coupled with the loss of heterozygosity correlates with chondrosarcoma in some cases, suggesting that the wild-type alleles may normally behave like tumor suppressors (21, 22). The EXT gene family appears to be widely distributed, with orthologs in Caenorhabditis elegans (rib-1/rib-2) and Drosophila melanogaster (tout-velu).

The connection between the heparan sulfate copolymerase and EXT was established by McCormick et al. (23), who cloned a cDNA with homology to EXT1 based on its ability to restore viral susceptibility to a cell line selected for resistance to herpes simplex virus and found to be defective for GAG biosynthesis (24, 25). This finding was followed by a report that recombinant forms of EXT1 and EXT2 have GlcNAc and GlcA transferase activities (26). Recent studies suggest that EXT1 and EXT2 may actually collaborate to form a functional oligomeric complex in the Golgi, the site where heparan sulfate polymerization is thought to occur (27). Mutations in the EXT1 locus in Drosophila (tout-velu), cultured cells, and mice result in the loss of heparan sulfate in vivo, providing strong evidence that EXT1 is indeed the copolymerase responsible for heparan sulfate formation (8, 23, 28, 29).

Here we report the expression cloning of hamster EXT1 by complementation of pgsD strains. Ectopic expression of EXT1 fully corrects the enzymatic and heparan sulfate deficiencies in the mutants, whereas introduction of EXT2 and the related EXT-like cDNAs fail to complement the defect in pgsD cells. We also show that pgsD mutants contain mutations in EXT1 and that two classes of alleles exist based on differential activity of the GlcNAc and GlcA transferases in vitro. Sequencing the mutant alleles indicates that the GlcA transferase domain most likely resides in the N-terminal portion of the protein.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture-- CHO cells (CHO-K1) were obtained from the American Type Culture Collection (CCL-61). One class of pgsD mutants (strains 623, 677, 108, 115, 6, 154, 625p, and 803) were isolated by replica plating and colony autoradiography, using ³⁵SO₄ incorporation to identify clones that produced less proteoglycans (8, 30). H661 was discovered by in situ enzymatic assay of GlcN N-sulfotransferase activity in mutagenized clones replica-plated to polyester cloth (30). This strain failed to produce endogenous substrate for the enzyme (31). HVR101 and HVR104 were isolated from mutagenized CHO-K1 cells by selecting for cells that survived exposure to herpes simplex virus type 2, which is cytotoxic to wild-type cells (32), followed by screening for colonies that incorporated ³⁵SO₄ but were resistant to virus infection. The purity of each strain was ensured by its isolation from cultures containing only mutant colonies. Cells were maintained in Ham's F12 medium (33) (Hyclone Laboratories) supplemented with 7.5% (v/v) fetal bovine serum (Hyclone Laboratories), 100 µg/ml streptomycin sulfate, and 100 units/ml penicillin G. Cells were switched to sulfate-deficient medium (34) with 7.5% dialyzed fetal bovine serum and 100 unit/ml penicillin G for radiolabeling studies.

Purification of GAG Chains-- Cells were labeled for 24 h with 10 µCi/ml H₂³⁵SO₄ (1325 Ci/mmol, NEN Life Science Products) in sulfate-free medium. Radiolabeled GAG chains were isolated by anion-exchange chromatography on DEAE-Sephacel as described previously (31) and separated by anion-exchange HPLC on a 7.5-mm inner diameter × 7.5 cm column of DEAE-3SW (TosoHaas, Montgomeryville, PA). The column was equilibrated in 10 mM KH₂PO₄ buffer (pH 6.0) containing 0.2% (w/v) Zwittergent 3-12 and 0.2 M NaCl. GAGs were eluted with a linear gradient of NaCl (0.2-1 M) in the same buffer using a flow rate of 1 ml/min and by increasing the NaCl concentration by 10 mM/min. The effluent from the column was monitored for radioactivity with an in-line radioactivity detector (Radiomatic Flo One/, Packard Instruments) with sampling rates every 6 s. The data were averaged over 1-min intervals.

Enzyme Assays-- GlcNAc transferase (UDP-GlcNAc:N-acetylheparosan 1-4N-acetyl-D-glucosaminyltransferase) and GlcA transferase (UDP-GlcA:N-acetylheparosan 1-4-D-glucuronosyltransferase) activities were assayed using oligosaccharide acceptors prepared from the capsular polysaccharide of Escherichia coli K5 (35). Briefly, the polysaccharide was partially N-deacetylated with hydrazine and subjected to deaminative cleavage with nitrous acid at pH 3.9 (36) The resulting mixture of oligosaccharides averaged 16 residues in length with GlcA at their nonreducing termini (average mass, 3200 Da) (8, 37). This mixture was used as a substrate for GlcNAc transferase assays. Small portions (0.6-1.0 mg) were digested as needed with -D-glucuronidase from bovine liver (Sigma, G-0501, 0.5 mg) in 0.1 M sodium acetate buffer, pH 5.0, at 37 °C overnight. The chains were separated from the released GlcA by gel exclusion chromatography on a PD-10 column (Amersham Pharmacia Biotech), lyophilized, and used as substrate for GlcA transferase.

Confluent monolayers of CHO cells were scraped with a rubber policeman in homogenization buffer (0.25 M sucrose, 20 mM Tris-Cl, pH 7.4, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml each leupeptin and pepstatin A) and sonicated by two 1-s pulses at output setting 3 on a 550 Sonic Dismembrator (Fisher Scientific) equipped with a microtip. Each 25-µl reaction for GlcNAc transferase contained 25 mM MOPS, pH 6.5, 20 mM MnCl₂, 0.3% Triton X-100 (w/v), 1 mM UDP-[6-³H]GlcNAc (220 Ci/mole), 25 µg of cleaved N-acetylheparosan, and 10-20 µg of cell homogenate protein. To assay GlcA transferase, each 25-µl reaction contained 25 mM MES, pH 5.5, 20 mM MnCl₂, 0.03% Triton X-100 (w/v), 2 mM UDP-[1-³H]GlcA (250 Ci/mole), 70 µg of -glucuronidase-treated N-acetylheparosan and 10-20 µg of cell homogenate protein. Product formation was proportional to time for over 4 h at 37 °C, but reactions were routinely terminated after 3 h by boiling or by adding dilute acid. Products were isolated by DEAE-Sephacel chromatography (Amersham Pharmacia Biotech), mixed with chondroitin sulfate carrier (1 mg), precipitated in 80% ethanol, and quantified by liquid scintillation spectrometry. UDP-[1-³H]GlcA was synthesized from D-[1-³H]glucose (10 Ci/mmol; NEN Life Science Products) as described (38, 39).

To determine the apparent K_m values for UDP-GlcNAc and UDP-GlcA, the nucleotide sugars were added in increasing concentrations to a fixed amount of N-acetylheparosan oligosaccharides (0.8 mg/ml and 2 mg/ml, respectively). The signal obtained at each concentration of UDP sugar in the absence of acceptor was <10% of that obtained in its presence, and these values were subtracted. Increasing concentrations of cleaved N-acetylheparosan were added to determine the K_m for the acceptors while holding the nucleotide sugar concentration at 1 and 2 mM for the GlcNAc and GlcA transferase reactions, respectively. The data were fitted by a nonlinear least-squares algorithm (SigmaPlot 4.11, Jandel Scientific) to the Michaelis-Menten rate equation (40). The reactions were assumed to be pseudo first-order, because the second substrate was kept at a constant value. Optimal concentrations of detergent and divalent cations were determined for each reaction. The pH was varied by using a mixture of MOPS and MES buffers (25 mM each).

Cloning the pgsD Locus-- Approximately 1 × 10⁷ pgsD-H661 cells were transfected with a mixture of 500 µg of a wild-type CHO-K1 cDNA library prepared in pcDNA1 (Invitrogen) and 1.8 mg of pMAMneo (CLONTECH) using Lipofectin (Life Technologies, Inc.) under the conditions recommended by the manufacturer. Stable transfectants were selected by adding 0.4 mg/ml geneticin sulfate (G-418) (Life Technologies, Inc.) to the medium, and positive clones were screened by replica plating and ¹²⁵I-FGF-2 binding (41). Two clones that bound ¹²⁵I-FGF were identified in this way (461 and 477-1) and purified by two rounds of subcloning. Genomic DNA was then isolated from clone 477-1 and used to transfect ~1.2 × 10⁷ pgsD-H661 cells (200 µg of genomic DNA and 6.8 µg of pSV2neo [CLONTECH]). Secondary transfectants were selected by resistance to G-418 and screened by replica plating and binding to ¹²⁵I-FGF-2, yielding one positive clone (5607). The genomic DNA was also partially digested with Sau3AI (Life Technologies, Inc.) and fractionated by centrifugation through a 10-40% (w/vol) sucrose gradient. Fragments of 9-23 kilobases were pooled, and a genomic library was constructed using Lambda FIX II/XhoI partial Fill-in Vector Kit and Gigapack III Gold Cloning Kit (Stratagene). ~2 × 10⁶ plaques were screened using a ³²P-labeled probe prepared from the cytomegalovirus promoter region of pcDNA1 (BamHI-NheI fragment) as a probe. Two positive phages 13-1-2 and 26-5-1 were identified, and 13-1-2 was further characterized. A 7-kilobase NotI fragment in 13-1-2 was subcloned into pcDNA3 and sequenced by BWG-Biotech, Inc. (High Point, North Carolina). This clone contained a full-length copy of EXT1 (GenBank^TM accession number AF252858).

Transfection of CHO Cells-- Plasmid or lambda phage DNA was introduced into mutant cells using LipofectAMINE (Life Technologies, Inc.) according to the manufacturer's instructions. Two days after transfection, cells were labeled with 10 µCi/ml ³⁵SO₄ for 24 h in sulfate-deficient medium, and the GAGs were isolated and analyzed by anion-exchange HPLC. To isolate stable transfectants expressing murine mEXT1 and mEXT2, cells were transfected with pCDNA3-mEXT1-myc/His and pCDNA3-mEXT2-myc/His (gifts from Frank Tufaro, University of British Columbia, Canada) as above. After two days, cells were replated in 100-mm dishes and treated with 0.4 mg/ml G-418 for 5 days. Cells then were overlaid with Whatman filter paper no. 50 and glass beads to obtain discreet colonies (30). The medium was replenished once, and after 5 days, single colonies were isolated with cloning rings. Positive clones were screened by flow cytometry analysis using biotinylated FGF-2 (38) and confirmed by Western blot analysis using anti-Myc antibodies (Invitrogen).

Full-length forms of mutant EXT1 cDNAs were isolated by reverse transcriptase-PCR as described below. The PCR products were digested by EcoRI and BamHI and inserted into the homologous sites in pFLAG-CMV2 (Sigma) to produce FLAG-tagged EXT1. Wild-type and mutant pgsD cells were transfected with the cDNAs, and the cells were analyzed by flow cytometry using biotin-FGF-2 as probe (38).

Expression of EXT-like cDNA in pgsD Mutants-- Human EXTL1 cDNA was PCR-cloned from a human smooth muscle first strand cDNA library (CLONTECH) using the following two sets of primers: hEXTL1-803-5, 5'-CCTGCTCTTCCTGCTTGCTG-3' and hEXTL1-2297-3, 5'-CTTCCTGTGCCCATCAATGAC-3', and hEXTL1-2164-5, 5'-TGAAGCTCATCCAGGCGGT-3' and hEXTL1-3074-3, 5'-AGGAGAAGGTAAGGCCACGC-3'. PCR conditions were: 95 °C for 3 min, 35 cycles of 95 °C for 45 s, 50 °C for 45 s, and 72 °C for 2.5 min, and 72 °C for 10 min using pfu polymerase (Stratagene). Both PCR fragments were cloned into pPCR-Script Amp SK(+) plasmid (Stratagene), cut with BglII and NotI, and cloned into the NotI site of pcDNA3.1() to obtain a full-length cDNA. The correct orientation of the clone was determined by EcoRI digestion.

To clone human EXTL2 cDNA, the PCR primers were 5'-GGCTGGCCCTACTGCAATC-3' and 5'-cgcatcctcgagaaGCTACTCAAATGCCAAGC-3' (the lowercase letters stand for noncoding sequence and the underlined sequence refers to an XhoI site). The PCR conditions were similar to those described above. The PCR product was cloned into pPCR-Script Amp SK(+) plasmid (Stratagene). The insert with the correct orientation was cloned into the NotI and BamHI sites of pcDNA3.1().

The original cDNA clone of human EXTL3 in pBluescript II SK(+) (KIAA0519) was a gift from Dr. Takahiro Ngase (Kazusa DNA Research Institute, Kisarazu, Chiba 292, Japan). The full-length cDNA was cloned into the XhoI and NotI sites of pcDNA3.1().

pcDNA3.1-hEXTL1, pcDNA3.1-hEXTL2, and pcDNA3.1-hEXTL3 were transfected into pgsD-677, pgsD-H661, and pgsD-623. At least eight stable transfectants from each mutant for each cDNA were picked. Expression of the human cDNAs in the stable transfectants was confirmed by Northern blot analysis. Expression of heparan sulfate was examined by flow cytometry using biotin-FGF-2 (38).

Northern Blot Analysis-- mRNA was isolated from cells using QuickPrep Micro mRNA Purification Kit (Amersham Pharmacia Biotech). The mRNA was denatured at 65 °C in a solution of 50% formamide (v/v), 6% formaldehyde (v/v), 20 mM MOPS (pH 7.0), electrophoresed in 1.0% agarose containing 6% formaldehyde (v/v), and transferred passively for 18 h to a Nytran plus nylon membrane (Schleicher & Schuell). The blotted RNA was cross-linked by UV irradiation and then prehybridized for 2 h at 42 °C in a solution containing 50% formamide (v/v), 20 mM sodium phosphate (pH 6.8), 5× SSC, 1× Denhardt's solution, 1% SDS (w/v), 5% (w/v) dextran sulfate, and 100 µg/ml denatured salmon sperm DNA. Double-stranded DNA probes were labeled with [³²P]dCTP by random oligonucleotide primers (Prime IT II labeling kit, Stratagene) using full-length cDNAs as templates. The probe was purified by Elute-tip (Schleicher & Schuell) before hybridization. Hybridization was carried out at 42 °C overnight in prehybridization buffer containing ~1 × 10⁶ cpm/ml of the ³²P-labeled probe. The membrane was washed two times with 2× SSC, 0.1% SDS for 30 min at 42 °C and then two times with 0.2× SSC, 0.1% SDS for 30 min at 65 °C. Hybridization was visualized by phosphoimaging (Storm 860, Molecular Dynamics) after overnight exposure.

PCR Analysis of EXT1 Mutations-- mRNA from mutant and wild-type CHO cells was purified as described above. First strand cDNA was produced using random primers in the Superscript^TM Preamplification System according to the manufacturer's protocol (Life Technologies, Inc.). To determine the mutations in the mutants two PCR fragments of ~1 kilobase each, which together covered the entire coding sequence, were isolated from each cell line. PCR primers were: 5'-CCTCTTGACCCAGGC-AGGACACATG-5' and 5'-GCGTGAGCGGATCTGCATTGGGAAG-3', 5'-TGCCACTTTCTGTCTGGTTCC-3' and 5'-ATGACGGCAGCTTGGTTCC-3'. PCR was carried out with pfu polymerase (CLONTECH; 25 cycles at 94 °C for 1 min, 55 °C for 1 min, and 72 °C for 2 min, followed by a final incubation at 72 °C for 7 min). The PCR products were cloned into pCR-Script Amp SK(+) (Stratagene). At least four clones from each PCR product were sequenced on both strands by the dideoxy chain termination method using Taq polymerase (dye terminator cycle sequencing, Perkin-Elmer) with a DNA automatic sequencer (ABI PRISM genetic analyzer). For other pgsD mutants, full-length EXT1 cDNAs were isolated by reverse transcriptase-PCR using the following primers: 5'-cgcgaattcaCAGGCCAAAAAACGCGAGGGCATC-3' and 5'-gtggatcccTCAAAGCCGTTCAATGTCTC-3' (lowercase letters stand for non-cDNA sequences and the underlined sequences are EcoRI and BamHI sites, respectively). The sequences were determined directly from these PCR products.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Two Classes of pgsD Mutants Exist-- Several heparan sulfate-deficient CHO cell mutants were previously isolated and characterized (8). In addition, new mutants have been identified based on the their inability to produce substrate for GlcN N-sulfotransferase (31), failure to bind to FGF-2 (41), or by their resistance to herpes simplex virus (24). The latter assay is based on the observation that herpes simplex virus utilizes heparan sulfate as a co-receptor for attachment and fusion of the viral envelope with the plasma membrane (32, 42-44). Based on cell hybridization studies, all of these mutants belong to a single complementation group that was originally designated pgsD (8).

PgsD mutants have mutations that affect the copolymerase responsible for the alternating addition of 1-4-linked GlcNAc and 1-4-linked GlcA units to nascent heparan sulfate chains. These reactions are catalyzed by a bifunctional protein that contains both GlcNAc transferase and GlcA transferase activities. Previous studies showed that the mutation in strain 677 affected both enzyme activities, which provided the first evidence that both catalytic sites resided in a single protein (8). To characterize the other mutants in the collection, GlcNAc and GlcA transferase assays in CHO cell homogenates were optimized with respect to buffer, pH, detergent, oligosaccharide acceptors, and nucleotide sugar donors. In general the enzyme activity was dependent on detergent but did not vary significantly in the range of 0.03-1%. Both reactions also were dependent on added divalent cations. The GlcNAc transferase showed an absolute preference for Mn²⁺ over Ca²⁺ or Mg²⁺ (Fig. 1), as previously reported for enzyme solubilized from murine mastocytoma (6). The GlcA transferase reaction showed less selectivity for divalent cations. The pH dependence also varied for the two activities. The GlcNAc transferase reaction proceeded across a broad pH range (pH 5-8), whereas the GlcA transferase reaction showed an optimum at pH 5.5-6.5 (Fig. 1). Although the shape of the curves differ from previously published data, the overall conclusions are the same (6).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Divalent cation dependence and pH profiles for GlcNAc and GlcA transferase activities. Cell extracts prepared from wild-type CHO cells were assayed for GlcNAc and GlcA transferase activities using N-acetylheparosan as substrate and either UDP-GlcNAc or UDP-GlcA as donor ("Experimental Procedures"). The concentration of divalent cations and pH was varied to find the preferred concentration.

The enzyme activities differed significantly in their catalytic efficiency, as measured by the estimated value of V_max/K_m (Fig. 2). The apparent K_m values for UDP-GlcNAc and N-acetylheparosan oligosaccharides for the GlcNAc transferase reaction were 230 ± 30 µM and 0.3 ± 0.1 mg/ml (~95 µM), respectively. The extrapolated values of V_max were 24 ± 1 and 21 ± 2 pmol/min/mg, respectively. For the GlcA transferase, the apparent K_m values for UDP-GlcA and N-acetylheparosan oligosaccharides were 2.0 ± 0.7 mM and 2.2 ± 0.5 mg/ml (~0.7 mM), respectively, and the values of V_max were 24 ± 5 and 38 ± 5 pmol/min/mg, respectively. These rates adequately account for the overall synthesis of heparan sulfate in CHO cells (~5 nmol/mg cell protein/generation).² Given these values, the apparent catalytic efficiency (V_max/K_m) for the GlcNAc transferase was 10-fold greater than for GlcA transferase. However, under conditions of saturating amounts of donor and acceptor, the maximal velocities were comparable. Previous studies also showed that the GlcNAc transferase had a lower K_m for the acceptor compared with the GlcA transferase (6). No attempt was made to examine how activity depended on the size of the oligosaccharide acceptor.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Substrate saturation curves for GlcNAc and GlcA transferase activities. Cell extracts prepared from wild-type CHO cells were assayed for GlcNAc and GlcA transferase activities using N-acetylheparosan as substrate and either UDP-GlcNAc or UDP-GlcA as donor ("Experimental Procedures"). The individual data sets were analyzed for best fit to the Michaelis-Menten equation. Saturation of the GlcA transferase was difficult to measure because of the high background in the assay when the donor concentration was raised above 2 mM. The average length of the oligosaccharide acceptor was 16 sugar units (average mass of 3200 Da).

Using the optimized assay conditions, the activities of GlcA and GlcNAc transferases were measured in 20 different pgsD mutants, and the behavior of representative strains is shown in Table I. Based on differences in activity, the mutants fell into two subgroups. Strains in group 1D have normal or elevated GlcNAc transferase activity but reduced GlcA transferase activity (e.g. strains 108, 115, 154, 623, HVR 101, and HVR104). In contrast, strains in group 2D have reduced levels of both enzymes (677, H661, 6, 803). One of the isolates also exhibited partial reduction in both activities (625p). These results suggested that the strains bore different mutant alleles and that the catalytic site for GlcA transferase may be located in a discrete region of the protein that was susceptible to mutagenesis. Alternatively, the selective loss of GlcA transferase activity could reflect an ancillary role for pgsD in activating GlcA transferase. In either case, it became desirable to clone the pgsD gene to study this domain and the overall organization of the encoded protein.

Initial attempts to clone the pgsD locus based on transient expression of cDNA libraries and correction of the heparan sulfate deficiency were unsuccessful (45). However, we were able to isolate a stable transfectant that regained the capacity to produce heparan sulfate. A genomic clone containing the correcting activity was eventually isolated from a lambda library ("Experimental Procedures") and reintroduction of the clone into the mutant corrected the heparan sulfate deficiency (Fig. 3B). Direct sequencing showed that it contained a full-length copy of the hamster homolog of EXT1, which had been cloned from the mouse as this work was in progress (23). Hamster EXT1 is 98% identical to human and mouse EXT1, with most of the variations occurring in the putative stem region separating the catalytic domains from the transmembrane segment (Fig. 4). Two possible N-glycosylation sites (marked by circles) as well as the cytoplasmic and transmembrane domains are conserved in all three species. The high degree of conservation of EXT1 suggests that it plays an important essential function; a conclusion supported by recent reports showing that EXT1 null mutants in mice exhibit embryonic lethality (29) and corresponding mutations in the Drosophila homolog (tout-velu) cause developmental defects (46, 47).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Anion-exchange HPLC of GAGs from CHO cells. The various cell lines were labeled with 35SO4 (10 µCi/ml) for 24 h, and the radiolabeled GAGs in the cells and the media were isolated and analyzed by anion-exchange HPLC ("Experimental Procedures"). The amount of radioactivity in each fraction was normalized to the equivalent amount of cell protein that had been analyzed. A, wild type, mutant pgsD-623 (group 1D) and mutant pgsD-H661 (group 2D). B, mutant H661 transiently transfected with lambda clone 13-1-2, CHO EXT1 cDNA, or control lambda; C, clones of H661 stably transfected with mouse cDNAs of EXT1 or EXT2. The elution positions for heparan sulfate and chondroitin sulfate were established by selective enzymatic digestion of the chains and are indicated by the bars in C.

View larger version (102K): [in this window] [in a new window]   Fig. 4.   Primary sequence of hamster, human, and mouse EXT1. The shaded segment at the N terminus is the putative transmembrane domain. Two potential attachment sites for Asn-linked glycans are circled. Individual amino acids altered in the mutants are shaded and described in greater detail in Table II. The underlined tripeptides indicate the positions of DXD and a EXE motif possibly involved in nucleotide sugar binding.

Only EXT1 Corrects the pgsD Deficiency-- EXT1 is a member of a family of proteins, whose members include EXT2, EXTL1/R3, EXTL2, and EXTL3/R1. Lind et al. (26) have reported that both EXT1 and EXT2 contain enzyme activities corresponding to heparan sulfate GlcNAc transferase and GlcA transferase. However, inspection of the published data indicates that neither enzyme reaction was optimized with respect to donor or acceptor and that the increase in activity in mutant cells transfected with EXT2 was low. Furthermore, recent work indicates that EXT1 and EXT2 proteins may form oligomers, which would make analysis of recombinant proteins produced in animal cells somewhat problematic (27). To gain further insight into this problem, murine EXT1 and EXT2 and human EXTL1/R3, EXTL2, and EXTL3/R1 were introduced into H661 cells, and stable transfectants were isolated. As shown in Fig. 3C, only EXT1 corrected the heparan sulfate deficiency in the mutant. This was confirmed by examining numerous transfectants of different pgsD strains with the various cDNAs (16 of 30 drug-resistant clones from five different transfections showed correction by EXT1, whereas 0 of 30 clones transfected with EXT2 and 0 of 8 clones each for the EXT-like sequences regained heparan sulfate synthesis). Northern blots confirmed that many of the stably transfected lines expressed the integrated cDNAs (data not shown). Enzymatic assay of extracts prepared from EXT1 transfectants showed that both enzyme activities were recovered (38 ± 6 pmol/min/mg for GlcA transferase and 19 ± 0.2 pmol/min/mg for GlcNAc transferase). Transiently expressing both mEXT1 and mEXT2 simultaneously in H661 did not restore heparan sulfate synthesis to a greater extent than EXT1 alone (data not shown). Thus, only EXT1 appears to have the ability to correct the pgsD deficiency.

PgsD Strains Contain Point Mutations in EXT1-- The correction of pgsD mutants by EXT1 cDNA suggested that they contained mutations in the EXT1 locus. Northern blot analysis of poly(A⁺) mRNA from the various mutants showed that most of the strains in group 2D had greatly diminished levels of EXT1 message, consistent with their missing both GlcNAc and GlcA transferase activities (Fig. 5). A partial mutant designated 625p expresses ~3-fold less heparan sulfate than wild-type cells, a partial reduction of GlcNAc and GlcA transferase activities, and mRNA for EXT1. In contrast, group 1D mutants, which have reduced GlcA transferase activity and normal GlcNAc transferase activity, expressed EXT1 mRNA at levels comparable to wild-type CHO cells (Fig. 5). These findings suggested that group 1D mutants probably contained point mutations that affected only the GlcA transferase domain.

View larger version (61K): [in this window] [in a new window]   Fig. 5.   Northern blot analysis of EXT1 and EXT2. Approximately 5 µg of poly(A+) RNA from each strain was separated by electrophoresis, blotted to nylon membranes, and probed with radiolabeled EXT1 or EXT2 cDNA probes ("Experimental Procedures"). MST refers to a mastocytoma-derived cell line that produces heparin (63).

To test this possibility, full-length PCR products of EXT1 were prepared and sequenced from mutant and wild-type cells. As shown in Table II, strains in group 1D contain missense mutations that resulted in substitutions at different residues, indicating the independent origin of the various mutants. Many of the mutations were nonconservative, but two of them changed Arg residues to Lys, suggesting that these two positions are very important for activity. All of the group 1D mutations cluster toward the N-terminal side of the central region of the protein, surrounding a DXD motif at residue 313-315 (or EYE at residues 318-320), which is conserved in the active site of many glycosyltransferases (48-50). Three group 2D mutants yielded products by reverse transcriptase-PCR analysis, and sequencing revealed Gln-398Stop in H661. Mutations have not been found in the other two strains, suggesting that they may have mutations in promoter or regulatory sequences.

To confirm that the mutations in EXT1 inactivated the catalytic activity, cDNAs containing each mutation were generated and introduced into mutant cells. The mutant cDNAs did not correct the heparan sulfate deficiency nor did they restore enzyme activity in vitro. These constructs also did not suppress heparan sulfate formation when introduced into wild-type cells, indicating that they did not exhibit dominant-negative effects (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Recently, the GlcNAc and GlcA transferases involved in heparan sulfate biosynthesis were shown to be encoded by EXT1 and possibly EXT2 (23, 26). McCormick et al. (23) discovered EXT1 while seeking genes that could restore susceptibility to herpes simplex virus-1 infection in mutant sog9 cells, which fail to produce GAGs (24). These investigators also showed that EXT1 constructs containing the missense mutations present in the mutants did not restore susceptibility to herpes simplex virus-1 infection, demonstrating that the defects most likely caused a loss of functional enzyme. Subsequently, Lind et al. (26) reported the cloning of bovine heparan sulfate copolymerase based on sequencing the purified enzyme and found that it was homologous to EXT2. EXT1 and EXT2 apparently can form a hetero-oligomeric complexes based on co-immunoprecipitation and colocalization studies in transfected cells (27). Other glycosyltransferases have been reported to consist of multiple subunits (51, 52), but in these cases the function of the subunits have been established by genetic as well as biochemical criteria. Thus, mutants are needed in EXT2 to rigorously establish its role in heparan sulfate biosynthesis. In this regard, it is interesting that none of the CHO or L cell mutants altered in heparan sulfate formation have defects in EXT2, suggesting that it may play nonessential role in the assembly process, perhaps by acting as a chaperone for EXT1. Evidence suggesting that the localization of EXT1 in the Golgi may depend on expression of EXT2 is consistent with this idea (27).

Six members of the EXT gene family have now been discovered. EXT1-3 have been assigned to human chromosome 8q24.1, 11p11-13, and 19p, respectively (14, 21, 53-57), whereas EXTL1/R3, EXTL2, and EXTL3/R1 were discovered by homology screens (17-20). Recently, Kitagawa et al. (58) suggested that EXTL2 has GlcNAc transferase activity responsible for the addition of the first -GlcNAc residue in heparan sulfate biosynthesis, -GlcNAc1-3GlcA1-3,Gal1-3Gal1-4Xyl-O-, but more detailed kinetic studies of the protein are needed to confirm this idea (37, 59). Northern blots of mRNAs from CHO cells do not detect EXTL2 transcripts, suggesting that perhaps one of the other EXT loci may fulfill this role (data not shown). The function of the other EXT proteins is unknown, but their homology to EXT1 and EXT2 suggests that they might have related activities. However, none of them appear to function as a copolymerase, because introduction into the CHO cell mutants does not restore heparan sulfate biosynthesis.

Our studies of the group 1D pgsD mutants show that most of strains have point mutations clustered in the central part of EXT1 in close proximity to each other. McCormick et al. (27) have shown that two of the human mutant alleles of EXT1 in this region (G339D and R340C) lack the ability to correct the EXT1 defect in sog9 cells, and when immunoprecipitated with EXT2, the complex lacked GlcA transferase activity. Taken together, the data suggest that this central region defines the active site of the GlcA transferase domain. The location of the GlcNAc transferase domain is unknown. Although it is tempting to speculate that subdomains of EXT1 may fold independently and give rise to functional transferase subsites on the copolymerase, more mutants are needed to define these domains. We have not yet found any pgsD alleles that selectively inhibit GlcNAc transferase activity without affecting GlcA transferase, suggesting that mutations in this region may destabilize the overall structure of the protein. We also cannot exclude the possibility that the GlcA transferase activity may depend on the association of EXT1 with another component, for example another EXT. The apparent activation of transferase activity seen when EXT1 and EXT2 are co-expressed in mouse L cells is consistent with this idea (27).

Mutations in EXT1 and EXT2 are associated with HME (10, 12, 13, 60, 61). However, the relationship between loss of heparan sulfate synthesis seen in the mutants and HME in humans is unclear. HME is characterized by cartilage-capped bony outgrowths at the growth plates of bones. The trait is inherited in an autosomal dominant fashion, suggesting that loss of a single copy is sufficient to observe exostosis. Chondrosarcoma in HME patients often correlates with loss of heterozygosity at the locus, suggesting that the wild-type allele may normally act as a tumor suppressor. Recently, a Drosophila homologue of EXT1 (tout-velu) was found to be required for the diffusion of Hedgehog, and tout velu mutants lack heparan sulfate (28, 46). Because Hedgehog homologs (Indian Hedgehog) as well as heparin-binding growth factors (e.g. FGFs) are involved in vertebrate cartilage and bone development (62), it is intriguing to speculate that a decrease of heparan sulfate in HME interferes with normal signaling events mediated by these factors. Many mutations in human EXT1 from HME patients occur in the region defined by the mutant CHO alleles (13), suggesting that loss of just the GlcA transferase domain is sufficient to cause disease.

Further studies are needed to establish that exostosis is actually caused by diminished heparan sulfate assembly. Mice containing a null allele at the EXT1 locus show severe developmental anomalies very early in development, around E6.5 (29). Embryonic stem cells derived from homozygous null embryos completely lack heparan sulfate, indicating that EXT1 is the primary route for generating heparan sulfate in both vertebrates and in flies (28, 29). Interestingly, embryonic stem cells derived from heterozygous null animals contain reduced amounts of heparan sulfate and about 30-50% of the normal levels of GlcNAc and GlcA transferase activities (29). However, heterozygous null animals have no obvious growth defects, and cartilage-capped outgrowths reminiscent of HME are not present. Although this observation may reflect differences in bone growth between species, it could also point to other functions for the EXT gene products. Perhaps, the phenotype may be related to gain of function changes induced by the missense or splicing mutations known to occur at the EXT1 locus in HME patients. The availability of null mice creates the possibility of producing transgenic animals containing the mutant alleles described here and in HME patients.

	ACKNOWLEDGEMENTS

We thank the Kazusa DNA Research Institute for providing EXTL3 cDNA and Dr. Carol Wise (University of Texas, Southwestern Medical Center) for providing EXTL1 subclones.

	FOOTNOTES

^* This work was supported by Grants GM33063 (to J. D. E.) and AI36293 (to P. S.) from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ On leave from the Dept. of Biochemistry and Molecular Genetics, University of Alabama, Birmingham, AL 35294.

^** Present address: Abbott Diagnostics Div., Abbott Laboratories, 100 Abbott Park Rd., Abbott Park, IL 60064-6016.

^§§ To whom correspondence should be addressed: Dept. of Cellular & Molecular Medicine, CMM-East Rm. 1054, Glycobiology Research and Training Program, University of California-San Diego, 9500 Gilman Dr., La Jolla, CA 92093-0687. Tel.: 858-822-1100; Fax: 858-534-5611; E-mail: jesko@ucsd.edu.

Published, JBC Papers in Press, June 22, 2000, DOI 10.1074/jbc.M002990200

² CHO cells make ~1 µg of heparan sulfate/mg of cell protein. The calculation assumes a doubling time of ~12 h, a turnover rate of 300%/generation, and a molecular mass of 50,000 Da for the heparan sulfate chains.

	ABBREVIATIONS

The abbreviations used are: GlcA, glucuronic acid; HME, hereditary multiple exostosis; CHO, Chinese hamster ovary; GAG, glycosaminoglycan; HPLC, high pressure liquid chromatography; MOPS, 3-[N-morpholino]propanesulfonic acid; MES, 2-[N-morpholino]ethanesulfonic acid; FGF, fibroblast growth factor; PCR, polymerase chain reaction; GlcA transferase, UDP-glucuronic acid:GlcNAc 1-4 glucuronosyltransferase; GlcNAc transferase, UDP-N-acetylglucosamine:GlcA 1-4-glucosaminyltransferase.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. Busse, A. Feta, J. Presto, M. Wilen, M. Gronning, L. Kjellen, and M. Kusche-Gullberg Contribution of EXT1, EXT2, and EXTL3 to Heparan Sulfate Chain Elongation J. Biol. Chem., November 9, 2007; 282(45): 32802 - 32810. [Abstract] [Full Text] [PDF]

				A. Clamp, F. H. Blackhall, A. Henrioud, G. C. Jayson, K. Javaherian, J. Esko, J. T. Gallagher, and C. L. R. Merry The Morphogenic Properties of Oligomeric Endostatin Are Dependent on Cell Surface Heparan Sulfate J. Biol. Chem., May 26, 2006; 281(21): 14813 - 14822. [Abstract] [Full Text] [PDF]

				R. Zhong, M. J. Pena, G.-K. Zhou, C. J. Nairn, A. Wood-Jones, E. A. Richardson, W. H. Morrison III, A. G. Darvill, W. S. York, and Z.-H. Ye Arabidopsis Fragile Fiber8, Which Encodes a Putative Glucuronyltransferase, Is Essential for Normal Secondary Wall Synthesis PLANT CELL, December 1, 2005; 17(12): 3390 - 3408. [Abstract] [Full Text] [PDF]

				D. Stickens, B. M. Zak, N. Rougier, J. D. Esko, and Z. Werb Mice deficient in Ext2 lack heparan sulfate and develop exostoses Development, November 15, 2005; 132(22): 5055 - 5068. [Abstract] [Full Text] [PDF]

				S. Ropero, F. Setien, J. Espada, M. F. Fraga, M. Herranz, J. Asp, M. S. Benassi, A. Franchi, A. Patino, L. S. Ward, et al. Epigenetic loss of the familial tumor-suppressor gene exostosin-1 (EXT1) disrupts heparan sulfate synthesis in cancer cells Hum. Mol. Genet., November 15, 2004; 13(22): 2753 - 2765. [Abstract] [Full Text] [PDF]

				S. Yamada, M. Busse, M. Ueno, O. G. Kelly, W. C. Skarnes, K. Sugahara, and M. Kusche-Gullberg Embryonic Fibroblasts with a Gene Trap Mutation in Ext1 Produce Short Heparan Sulfate Chains J. Biol. Chem., July 30, 2004; 279(31): 32134 - 32141. [Abstract] [Full Text] [PDF]

				D. J. Bornemann, J. E. Duncan, W. Staatz, S. Selleck, and R. Warrior Abrogation of heparan sulfate synthesis in Drosophila disrupts the Wingless, Hedgehog and Decapentaplegic signaling pathways Development, May 1, 2004; 131(9): 1927 - 1938. [Abstract] [Full Text] [PDF]