Metabolism of GPIs in mammalian cells

In Trypanosoma brucei, glycosylphosphatidylinositol (GPI) anchors of proteins and free GPIs with identical structures have been characterized. This identity provides strong presumptive evidence that the free GPIs are in fact precursors of the GPI anchors on proteins. In mammalian tissues, however, rather consistent differences in the structures of free GPIs and GPI anchors are observed. The terminal GPIs produced by the mammalian biosynthetic pathway differ from GPI anchors in being almost exclusively fatty acid acylated on the inositol residue, having a greater number of phosphoethanolamine residues, and perhaps in containing a greater percentage of diacylglycerol components. While in principle these differences could be reconciled by remodeling reactions before or after attachment of GPI anchors, it is possible that some of the mammalian free GPIs play cellular roles other than as anchor precursors. We have approached this question by studying the lifetimes of the last three GPIs on the biosynthetic pathway, denoted H6, H7 and H8, in K562 cells and in a K562 mutant designated class K that is devoid of GPI-anchored proteins. Pulse-chase metabolic labeling with [3H]-mannose indicated that H6 was a precursor of H7 and H8 and that the H8 lifetime was more than one hour in the parental cells and even longer in the mutant. Preliminary data indicated that the majority of each of the three GPIs was localized in the plasma membrane fraction rather than the endoplasmic reticulum. These observations argue that mammalian GPIs are not utilized exclusively as GPI anchor precursors.