No abstract

Glycosylphosphatidylinositol (GPI) represents an important anchoring molecule for cell surface proteins. The first step in its synthesis is the transfer of N-acetylglucosamine (GlcNAc) from UDP to phosphatidylinositol (PI). The products of three mammalian genes, PIG-A, PIG-C and PIG-H, have previously been shown to be involved in the putative enzymic complex. Here we report the cloning of human and mouse cDNAs encoding a fourth participant in the GlcNAc transfer reaction which are homologues of the Saccharomyces cerevisiae and Schizosaccharomyces pombe Gpi1 proteins. To provide evidence for their function, these proteins were expressed in GPI1-disrupted yeast strains. In Sacch. cerevisiae, where GPI1 disruption results in a temperature-sensitive phenotype and abolishes in vitro GlcNAc-PI synthesis, restoration of growth could be demonstrated in a temperature-dependent manner. In addition, in vitro GlcNAc-PI synthetic activity was again detectable. In Schiz. pombe, gpi1+ disruption is lethal. Using random spore analysis, we were able to show that the mammalian GPI1 homologues can rescue haploids harbouring the lethal gpi1+::his7+ allele. Our data demonstrate that the genes identified are indeed involved in the first step of GPI biosynthesis, and allow conclusions about a specific function for Gpi1p in stabilizing the enzymic complex. The finding that, despite a low degree of identity, the mammalian Gpi1 proteins are able to participate in the yeast GlcNAc-PI synthetic machinery as heterologous components further demonstrates that GPI biosynthesis has been highly conserved throughout evolution.

Paroxysmal nocturnal hemoglobinuria (PNH) is an acquired clonal defect of hematopoietic stem cells in which affected cells are characterized by the lack of glycosylphosphatidylinositol (GPI)-anchored proteins. The lesion in PNH lies in the defective synthesis of N-acetyl-D- glucosaminyl-phosphatidylinositol (GlcNAc-Pl), the first intermediate in GPI biosynthesis. Reintroduction of the PIG-A gene into GPI(-) patient cells reportedly complements this defect. We have analyzed here PIG-A transcripts of six PNH patients. GPI+ and GPI- cell lines from each individual were used, ie, Epstein-Barr virus-transformed B- lymphoblastoid cell lines, T-cell lines, and natural killer cell clones. Reverse transcriptase polymerase chain reaction and sequencing showed three different PIG-A splicing variants in GPI+ cell lines, in which the largest transcript contained the wild-type PIG-A coding region sequence. GPI-deficient cell lines showed abnormal splicing variants. Sequencing of PIG-A complementary DNA and genomic DNA showed heterogeneous mutations ranging from different point mutations to small deletions. Two lymphocyte cell lines (T- and B-cell lines) of one patient presented with the same mutation. For another patient, two different mutations were detected in one natural killer cell line. Therefore, different cell lineages have somatic mutations in PIG-A that lead to PNH.

No abstract