The roles of ext1 and ext2 in heparan sulfate polymerization and hereditary multiple exostoses

A human cDNA library was screened for the ability to restore susceptibility to herpes simplex virus (HSV) infection in sog9 cells, a normally HSV-resistant murine L cell line. After extensive screening, a single cDNA was isolated that could completely restore the HSV-susceptibility of this cell line. The cDNA was sequenced and found to be identical to EXT1, a previously identified putative tumour suppressor gene involved in hereditary multiple exostoses (HME). Analysis of the protein product of the EXT1 cDNA revealed that EXT1 is a type II transmembrane glycoprotein that is decorated with N-linked oligosaccharides, localized predominantly to the endoplasmic reticulum. Examination of EXT1-expressing sog9 cells revealed that HSV-susceptibility is restored in these cells due to a restoration of the primary HSV-receptor, heparan sulfate (HS), to the cell surface. Taken together, these results suggested that EXT1 is likely involved in HS biosynthesis. HME is an autosomal dominant disorder caused by mutations in either EXT1 or EXT2. EXT2, like EXT1, is localized predominantly to the endoplasmic reticulum (ER), but EXT2 is unable to restore heparan sulfate to the surface of sog9 cells, suggesting that EXT1 and EXT2 do not share redundant functions in the cell. When these two proteins were co-transfected into the same cell, there was a striking re-location of EXT1 and EXT2 to the Golgi apparatus, and stable EXT 1/EXT2 complexes could be isolated. This EXT1/EXT2 complex had a greater cumulative HS polymerizing activity than either protein alone, suggesting that the complex is the more biologically active form of the enzyme. To determine whether the HS polymerizing activity of EXT1/EXT2 was relevant to HME, disease-causing EXT missense mutants were constructed and tested for complex formation, subcellular localization, and function. These missense mutants fall into two categories, those that form inactive hetero-oligomeric complexes that are retained in the ER, and those that can form Golgi-localized hetero-oligomeric complexes that are deficient in one of the two transferase activities necessary for the polymerization of HS. Moreover, none of the EXT mutants could restore HS to the surface of sog9 cells, as measured by the sensitive HSV-1 infection assay. Thus, defects in EXT1 or EXT2 inactivate the EXT1/EXT2 complex. These findings provide a rationale to explain why mutations in either of the two genes can cause HME. During the course of this study, the parental mouse L cell line was stably transfected with EXT1 to observe the effects of overexpression of one member of the EXT1/EXT2 complex. These L-EXT1 cells displayed an altered, more stellate cell morphology, and anion exchange chromatography of glycosaminoglycans revealed that their cell surface HS was less negatively charged and more heterogeneous than L cell HS. However, the most remarkable phenotype of L-EXT1 cells was a dramatic increase in the ability of progeny HSV to spread from cell-to-cell. These L-EXT1 cells also displayed a decreased affinity for certain extracellular matrices. Taken together, these observations suggest that disrupting the balance of the members of the EXT1/EXT2 complex leads to changes in the architecture of the cell surface that allow for more efficient cell-to-cell spread of HSV, possibly through altered cell-cell contacts.Science, Faculty ofMicrobiology and Immunology, Department ofGraduat