Glycosylphosphatidylinositol anchor

The glycan glycosylphosphatidylinositol (GPI) occurs in all eukaryotes, from fungi to plants and animals, as a linker anchoring proteins to membranes. The structure of the GPI anchor slightly differs between lower and higher organisms but the respective biosynthesis pathways are strongly conserved. All GPI anchors include GlcN, Man and ethanolamine (EtN) residues. Several hundred GPI-anchored proteins have been identified in eukaryotes. Functionally, GPI-anchored proteins include enzymes, adhesion proteins, proteoglycans, signaling receptors, inhibitors, and proteins of unknown function like the prion protein. The biosynthesis of GPI is essential for the viability of fungi as well as of mammals. Parasites like the protozoan Trypanosoma and Plasmodium heavily rely on GPI-anchored proteins to proliferate and evade their host’s immune system. For example, the variant surface glycoproteins of Trypanosoma brucei are GPI-anchored.

FIG: GPI CORE STRUCTURE

The biosynthesis of the GPI anchor takes place at the ER membrane. Somewhat similar to the assembly of lipid-linked oligosaccharides in the N-glycosylation pathway, the GPI anchor is built up by stepwise additions and involves the flipping of an intermediate from the cytosolic to the luminal side of the membrane. Although the structure of the GPI anchor differs between lower and higher eukaryotes, the enzymes involved in GPI biosynthesis are conserved across eukaryotes. For example, the enzymatic complex catalyzing the transfer of GlcNAc to phosphatidylinositol contains in yeasts and humans the same six subunits, which share between 21 and 46% of sequence similarity. Dol-P-Man is the donor substrate used by the ER Man-transferases. This means that defects of Dol-P-Man biosynthesis, as found in DPM1-CDG, also decrease the cellular GPI pool, thereby reducing the presentation of GPI-anchored proteins.

The total loss of the GPI anchor is not compatible with life but a deficiency limited to hematopoietic cells causes a form of hemolytic anemia called paroxysmal nocturnal hemoglobinuria (PNH, OMIM 311770). This complex and somewhat intimidating name indicates that the patients notice dark colored urine in the morning after sleep. Red blood cells lacking GPI-anchored proteins are sensitive to complement-mediated lysis, thus explaining the hemolysis and the accumulation of hemoglobin in the urine. PNH is an acquired genetic disease mainly caused by mutation in the X-linked PIGA gene arising in hemopoietic stem cell clones. PNH patients are usually diagnosed between 30 and 40 years of age. The disease also features an increased incidence of leukemia and increased thrombotic episodes. These complications usually lead to the death of the patients within 10-20 years after diagnosis. Platelet transfusion and anticoagulant therapies are commonly used to improve the survival rate. In 2007 the drug Soliris has been approved by the FDA as a PNH therapy. Soliris is a monoclonal antibody binding the complement protein C5, thus inhibiting complement mediated intravascular hemolysis in PNH patients. Soliris has set the sad world record for the most expensive medicine with an annual cost exceeding US$ 450’000 per patient!

FIG: GPI BIOSYNTHESIS

After N-deacetylation of GlcNAc-phosphatidylinositol at the cytosolic side of the ER membrane, the intermediate GPI flips across the membrane. In spite of extensive mutagenesis screens performed in yeast and in mammalian cells, the flippase acitivity could not be assigned to a specific protein. It is likely that a loss of flippase activity is not viable even at the cellular level. Accordingly, it has been postulated that the same flippase translocates both GPI and phospholipids, which would explain the essential role of this activity for cell viability.

FIG: GPI FLIPPING IN ER

Once assembled, the GPI anchor is linked to the C-terminus of proteins by transamidation. The GPI transamidase complex includes four to five transmembraneous proteins. PIG-T is the central component interacting with GPI8/PIG-K, GAA1/GPAA1 and GPI17/PIG-S. The C-terminal signal peptide on substrate proteins comprises three amino acids symbolized ω, ω+1 and ω+2. The first amino acid ω is where the GPI attaches. The ω+1 position can be any amino acid except proline and tryptophan, whereas the ω+2 position is usually occupied by amino acids with short side chains. The ω segment is followed by a short   hydrophilic spacer of 5 to 10 amino acids and a hydrophobic segment of up to 20 amino acids.

FIG: GPI-PROTEIN TRANSAMIDATION

The GPI transamidase complex recognizes and cleaves the signal peptide of substrate proteins. Then, the catalytic component GPI8/PIG-K attacks the peptide bond between the ω and ω+1 sites with a sulfhydryl group of its active site cysteine to form a carbonyl intermediate. The amino group of the terminal EtN group of GPI then attacks the thioester in the intermediate to complete the reaction.

Although the overall structure of the GPI anchor is conserved among eukaryotes, some differences in its biosynthesis exist between protozoan and mammalian cells. The acylation of inositol precedes the addition of Man to the GPI in mammals, whereas acylation takes place during mannosylation in Trypanosoma. Subsequently, the removal of the acyl chain on inositol takes place after protein linkage in mammals, whereas it occurs before protein linkage in Trypanosoma and the same acyl chain is never removed in Plasmodium. These differences in GPI biosynthesis are further being studied by scientists to develop selective inhibitors of the protozoan enzymes, which may one day represent valuable drugs against malaria and sleeping sickness.

FIG: DIFFERENCE MAMMALIAN-TRYPANOSOME GPI SYNTHESIS

Several questions related to the intracellular trafficking of GPI-anchored proteins remain to be solved. For example, what are the signals directing the transit of GPI-anchored proteins from the ER to the Golgi and to the plasma membrane? The elegant work of Taroh Kinoshita on the study of GPI trafficking has provided the first answers to these questions. This research group has demonstrated that EtN group on the second Man residue acts as a signal mediating the departure of GPI-anchored proteins from the ER to the ERGIC. The cleavage of the second EtN group is controlled by the phosphoesterase enzyme PGAP5. Interestingly, there is no PGAP5 homolog in Trypanosoma, which generate a lot of GPI-anchored proteins. 

FIG: PGAP5 FUNCTION

In the ERGIC or Golgi compartments, the phosphatidylinositol moiety is remodeled by the PGAP3 and PGAP2 proteins, which respectively cleave the unsaturated fatty acid at the sn-2 position and then add a saturated fatty acid at the same position. This lipid remodeling of GPI-anchored proteins is critical for their association to lipid rafts in mammalian cells and in yeasts. Mutations in the PGAP3 and PGAP2 genes cause a severe mental retardation syndrome featuring hyperphosphatasia (OMIM 615716 and 614207) due to increased secretion of the normally GPI-anchored alkaline phosphatase enzyme in the blood. In addition to the PIGA deficiency described here above, defects of GPI biosynthesis also result from mutations in the PIGL, PIGW, PIGM, PIGN, PIGV, PIGO, and PIGT genes. These main clinical manifestations of such defects are psychomotor retardations and multiorgan dysfunctions.

FIG: GPI BIOSYNTHESIS DEFECTS