While studying the cellular localization and activity of enzymes involved in heparan sulfate biosynthesis, we discovered that the published sequence for the glucuronic acid C5-epimerase responsible for the interconversion of D-glucuronic acid and L-iduronic acid residues encodes a truncated protein. Genome analysis and 5-rapid amplification of cDNA ends was used to clone the full-length cDNA from a mouse mastocytoma cell line. The extended cDNA encodes for an additional 174 amino acids at the amino terminus of the protein. The murine sequence is 95% identical to the human epimerase identified from genomic sequences and fits with the general size and structure of the gene from Drosophila melanogaster and Caenorhabditis elegans. Full-length epimerase is predicted to have a type II transmembrane topology with a 17-amino acid transmembrane domain and an 11-amino acid cytoplasmic tail. An assay with increased sensitivity was devised that detects enzyme activity in extracts prepared from cultured cells and in recombinant proteins. Unlike other enzymes involved in glycosaminoglycan biosynthesis, the addition of a c-myc tag or green fluorescent protein to the highly conserved COOH-terminal portion of the protein inhibits its activity. The amino-terminally truncated epimerase does not localize to any cellular compartment, whereas the fulllength enzyme is in the Golgi, where heparan sulfate synthesis is thought to occur.Heparan sulfate proteoglycans are located on the cell surface and in the extracellular matrix, where they play important roles in cell adhesion, differentiation, and growth in vitro and in vivo (1-3). To a large extent, these biological activities depend on the heparan sulfate chains attached to the core protein. Heparan sulfate, a type of glycosaminoglycan, initially assembles by the copolymerization of N-acetyl-D-glucosamine (GlcNAc) and D-glucuronic acid (GlcA). The backbone then undergoes extensive modification initiated by the N-deacetylation and N-sulfation of subsets of GlcNAc residues. Subsequently, D-GlcA residues adjacent to the N-sulfated sugars are converted to L-IdoUA 1 by a C5-epimerase and are sulfated at C-2 by a specific sulfotransferase. The glucosamine units also can be sulfated at C-6 and to a lesser extent at C-3. The blocklike arrangement of the modified residues confers specific binding properties to the chains for protein ligands, which in turn facilitate various biological activities. Many of the enzymes involved in heparan sulfate and heparin formation seem to be members of multienzyme gene families. Two exceptions are the C5-epimerase that interconverts D-GlcA and L-IdoUA and the 2-O-sulfotransferase that adds sulfate to C-2 of IdoUA residues and to a lesser extent GlcA residues. The C5-epimerase has been partially purified from mouse mastocytoma (4) and purified to homogeneity from bovine liver (5). A bovine cDNA for the epimerase has been cloned as well (6). Kinetic studies have clarified the substrate specificity of the epimerase, showing that the enzyme will react with both D...