GLYCOSYL PHOSPHATIDYLINOSITOL ANCHORS

Abstract

The earliest studies suggesting proteins were associated with membranes solely through a covalently-linked glycosyl phosphatidylinositol (GPI) took advantage of Pl-specific phospholipases (PIPLCs) [reviewed by Low (1)]. The demonstration in the mid-1970s that alkaline phosphatase could be released from tissues by PIPLCs purified from B. cereus (2) and S. aureus (3) led eventually to the suggestion (4) of a direct protein-PI linkage. Another line of evidence, which suggested that the hydrophobic properties of a protein could be due solely to the covalent, C-terminal attachment of a lipid moiety containing ethanolamine and glucosamine, came from the protein sequence and compositional analyses of Thy-1 published in 1981 (5). It was not until 1988, however, that the complete primary structure of the GPI anchor of Thy-1 was published (6). Publication in the same year (7) of the complete structure of the Trypanosoma brucei variant surface glycoprotein (VSG) anchor culminated a long series of investigations that also began in the early 1980s (see ref. 8) Another long-standing investigation, that of human acetylcholinesterase (AChE), has led to the recent publication of a third protein-linked GPI structure (9,10). In fact, the last five years have witnessed an explosion in our knowledge of GPls, including not only their Structural features, but also a proposed biosynthetic pathway and information from gene fusion experiments about the signals directing their addition to proteins and their functional importance.