Glycophosphatidylinositol(GPI)-Anchored Proteins and Phosphatidylinositol Mannosides

1.   Structure and Occurrence

A novel phospholipase C was obtained from Bacillus cereus in 1976 with the specificity to act upon phosphatidylinositol to generate diacylglycerol and inositol phosphate. When this was tested with tissues a year or two later, it was found to release in addition a variety of proteins including 5’-nucleotidase and erythrocyte acetylcholinesterase. It was apparent that these and many other proteins were covalently attached to phosphatidylinositol located in the cellular membranes. By 1985, detailed evidence was obtained for various components linking phosphatidylinositol to cell surface proteins but especially in relation to acetylcholinesterase in various species and of surface glycoproteins in the parasitic protozoan Trypanosoma brucei, and by 1988 a complete structure of the last was obtained by M.A.J. Ferguson and colleagues. It soon became apparent that there was a basic general structure for what became known as the glycosylphosphatidylinositol(GPI)-anchored proteins. As the lipid component is much more complex than in other proteolipids, they are discussed separately here.

General formula for a GPI-anchored proteinPhosphatidylinositol in the external leaflet of the plasma membrane is now recognized as the lipid anchor that binds a variety of proteins via the C-terminus to a phosphoethanolamine unit at the end of a complex glycosyl bridge. These are ubiquitous in eukaryotes (fungi, protozoans, plants, insects and animals) and they have also been shown to be present in some of the Archaebacteria (but not Eubacteria). In animals, they are found in every type of cell and tissue.

A typical molecule is illustrated schematically. These complicated glycophospholipid-protein aggregates are abundant in nature, amounting to about 1% of all proteins and up to 20% of membrane proteins (at least 250 different). They were first studied in detail in parasitic protozoa such as T. brucei (African sleeping sickness) or Leishmania spp., where they were more readily accessible in sufficient quantity for structural analysis (a hundred times greater than in mammalian cells), but they have now been characterized from most classes of eukaryotic organisms.

The aliphatic residues are embedded in the membrane, and their chemical composition is dependent on the organism and the stage in its life cycle, but commonly position sn-1 is occupied by a long-chain (C18 or C24) ether-linked alkyl moiety and position sn-2 by a saturated fatty acid (12:0 to 26:0). However, forms with simple fatty acid compositions, such as two myristic acid residues (14:0), are also known. Some GPI anchors contain an additional fatty acid, often 16:0, attached to position 2 of the inositol ring; this has the important property of inhibiting the action of phospholipase C.

The Man(α1-4)GlcN(α1-6)-myo-inositol-1-HPO4 lipid part is highly conserved (from yeast to humans), indicating that all are part of a single family of complex molecules. Similarly, the core glycan Man(α1-2)Man(α1-6)Man(α1-4)GlcN(α1-6)-myo-inositol is conserved, although it can be substituted in a species-specific manner with side chains, such as ethanolamine phosphate, mannose, galactose or sialic acid. For example, the GPI anchor for acetylcholinesterase from human erythrocytes is illustrated. It has either an 18:0 or an 18:1 alkyl group attached to position sn-1 of the phosphatidylinositol moiety with a 22:4, 22:5 or 22:6 acyl group linked to position sn-2 and a 16:0 fatty acid linked to position 2 of inositol. There are two ethanolamine phosphate residues attached to the glycan core. These are the type-1 GPIs.

Formula of the GPI anchor for acetylcholinesterase from human erythrocytes

Certain protozoa and trypanosomatid parasites contain type-2 and hybrid GPIs, which differ at one of the hexose linkage points. They also have one fewer ethanolamine phosphate residue than the mammalian form. The phospholipid moiety is variable among protozoan species, and includes diacylglycerol, alkylacylglycerol and ceramide forms, with differing fatty acid constituents. They are related to the lipophosphoglycans and phosphatidylinositol mannosides discussed below.

Yeasts are distinctive in that they contain both GPI-anchored proteins, with a characteristic C26 fatty acid component, and ceramide phosphorylinositol-anchored proteins. With the latter, the ceramide moiety is incorporated by an exchange reaction that occurs after the addition of the GPI precursor to proteins.

It is noteworthy that free or non-protein-bound glycosyl phosphatidylinositols are present on the external surface of the plasma membrane of some cells both in animals and protozoa (but not in yeast). Normally, they are present at low levels, but the parasitic protozoan Babesia bovis contains substantial amounts.


2.   Biosynthesis and Function

Considerable progress has been made towards an understanding of the biosynthesis of GPI-protein complexes, and it is apparent that both the biosynthesis of GPI precursors and post-translational modification of proteins with GPI take place in the endoplasmic reticulum. The process starts on the cytoplasmic side of this membrane and is completed on the lumenal side, so the intermediate glycophospholipid must be flipped across the membrane. In mammalian cells, the lipid precursor is a conventional phosphatidylinositol molecule, which is first attached to an N-acetylglucosamine residue. This is deacetylated before a saturated fatty acid (usually palmitate) is attached to the inositol residue, and this is followed by a sequence of reactions in which further carbohydrate moieties are added. The palmitate attached to inositol can be removed after mannose-I is linked to the complex glycolipid, but this usually occurs later in the sequence. The final step before attachment of the protein is the addition of phosphoethanolamine to mannose-III.

The GPI proteins all contain a characteristic carboxyl-terminal signal peptide with a hydrophobic tail, which is split off before the protein with a new carboxyl-terminal is combined with the amino group of the ethanolamine residue of the GPI moiety. A GPI-transamidase complex catalyses the overall process of cleavage and GPI attachment. The palmitate attached to inositol may then be removed before the GPI-anchored proteins are transported to the Golgi. Here, the unsaturated fatty acid in position sn-2 of the glycerol moiety is removed by the action of phospholipase A2 to form a lyso-GPI-protein, and this is reacylated with a saturated acid (26:0 in yeast and mainly 18:0 in mammalian cells). Remodelling of the glycan side chain by removal of an ethanolamine phosphate residue and addition of an N-acetylgalactosamine can also occur. The remodelled GPI-anchored protein containing two saturated fatty acids is finally transferred to the outer leaflet of the plasma membrane.

Animals with defects in the biosynthesis of GPI anchors do not survive beyond the embryo stage. They are also essential for viability in yeast, for virulence and survival of parasitic protozoa in their host, and for many aspects of development in plants.

GPI-anchored proteins have a diverse range of functions, but many are hydrolytic enzymes (including peptidases) or serve as receptors, cell surface antigens or cell adhesion molecules. Most GPI-anchored proteins can be identified from DNA analysis by the presence of the characteristic N- and C-terminal signal peptides. While its complexity suggests that a variety of functions might be possible, it seems that the main purpose of the GPI anchor is to act as a stable anchoring device that resists the action of most extracellular proteases and lipases. It targets its protein/enzyme component to a specific membrane, where the latter is required for its specific function. However, some further movement is possible and transfer between membranes and even between cells can take place.

In addition, the nature of the hydrophobic moiety, resembling that of a ceramide, ensures that the GPI anchor is readily incorporated into those membrane regions enriched in sphingolipids and cholesterol and termed ‘rafts’, where the glycan core may aid lateral mobility. The two saturated fatty acids in the GPI anchor are essential for raft association to occur, but the anchored proteins may also affect microdomain formation by forming transient homodimers. As many important signalling proteins are found in these membrane domains, there are suggestions that the GPI-anchored proteins may be important in signal transduction. For example, the GPI-anchor may function as a sorting signal for transport of GPI-anchored proteins in the secretory and endocytic pathways, facilitated by the remodelling processes that occur in the Golgi.

GPI-linked proteins are not as tightly anchored to the membrane as transmembrane proteins and so can migrate from one cell to another enabling cell communication. Thus, GPI proteins, with and without the GPI anchors, are found in serum and other body fluids. They are released from the plasma membrane either via membrane vesicles or through the activity of phospholipases, such as phosphatidylinositol-specific phospholipases C and D, by a process termed ‘shedding’, which can remove parts of the GPI anchor possibly as part of a regulatory mechanism. Such processes are reversible in that GPI-linked proteins can be reinserted into cell membranes.

GPI-anchored proteins are involved in a number of diseases. For example, when associated with lipid rafts, they can be incorporated into the lipid envelopes of viruses, where they may promote viral replication. In T. brucei and related species, GPI-anchor proteins, especially a glycoprotein termed the ‘promastigote surface protease’, accompanied by lipophosphoglycans (see below) form a dense layer as a protective barrier around the organism. They stimulate the immune system in mammalian host tissues by activating macrophages and promoting the release of different proinflammatory cytokines and chemokines, such as tumor necrosis factor-alpha, interleukin-1 and nitric oxide. Synthetic GPIs are under investigation as potential vaccines against such intractable parasitic diseases as malaria. A further important example is the prion protein responsible for ‘mad cow’ disease where the GPI-anchor may have a role in the pathogenicity of the disease. Similarly, certain bacterial toxins bind to GPI-anchors to exert their pathological effects.


3.   Lipophosphoglycans and Phosphatidylinositol Mannosides

In addition to the GPI-anchor molecules, carbohydrates attached to phosphatidylinositols play a role in the surface antigenicity both of protozoal parasites and of prokaryotic organisms, especially those of actinomycetes or coryneform bacteria. In particular in the parasitic protozoal parasites, lipophosphoglycans are present on the external cell surface, where they are intimately involved in host-pathogen interactions. They are based on a type-2 GPI core, Manα1-3Manα1-4GlcNα1-6PI, as part of a conserved hexaglycosyl unit, which is attached to a long phosphodisaccharide-repeat domain that carries species-specific side-chain modifications and is completed by a neutral oligosaccharide. These are essential for successful invasion of the host animal. In addition, the galactofuranose unit (Galf), which does not occur in mammalian cells, is also believed to play a part in the pathogenicity. In Leishmania species, the lipid component is a monoalkyl-lysophosphatidylinositol with saturated C22 to C24 alkyl groups.

Structure of the lipophosphoglycans from Leishmania species

Analogous lipophosphoglycans with the lipid backbone consisting of a ceramide, i.e. ceramide phosphorylinositol, rather than a diacylglycerol, are also found in nature, especially in plants, yeasts and other fungi.

Further related lipids are the phosphatidylinositol mannosides, with the first mannose residue attached to the 2-hydroxyl group and the second to the 6-hydroxyl of myo-inositol, which are found uniquely in the cell walls of the bacterial suborder Corynebacterineae, including Mycobacteria and related species, many of which are important pathogens. These lipids range in structure from simple mono-mannosides in some Streptomyces and Mycobacterium species and in propionibacteria to molecules with 80 or more hexose units. In addition, one or two further fatty acyl groups can be linked to the inositol-mannose chain. They are present in both the inner and outer membranes of the cell envelope of the organisms.

Formula of a phosphatidylinositol dimannosideThe phosphatidylinositol dimannoside from Mycobacterium tuberculosis and M. phlei illustrated has been characterized as 1-phosphatidyl-L-myo-inositol 2,6-di-O-α-D-mannopyranoside. The main fatty acid constituents are palmitic and 10-methyl-stearic (tuberculostearic) acids. This is the basic structure from which additional phosphatidylinositol mannosides are produced with up to four further mannose units. Di- and hexamannosides are the main components, and they can have one to four fatty acyl groups in total, with the additional fatty acyl substituents linked to position 2 of the inositol moiety and/or position 6 of one of the inner mannose units. The biosynthesis of such complex lipids involves a number of reactions, and it is apparent that the first two mannosylation steps of the pathway occur on the cytoplasmic face of the plasma membrane, but that further mannosylations require integral membrane-bound glycosyltransferases on the periplasmic side of the membrane.

The lipomannans have a longer chain of mannose units, comprising a backbone of α1,6 mannose residues with α1,2 mannose side chains (mannans) attached to phosphatidylinositol, and these can be further modified with arabinan (arabinose polysaccharide) branch to produce the highly complex lipoarabinomannans. For example, in M. tuberculosis, the arabinan component contains a linear polymer of ~70 residues of D-arabinofuranose in α1,5 linkage, modified with α1,3 branch points. Such molecules are believed be important for the structural integrity of the cell walls of the organisms, a function similar to that of the lipoteichoic acids.

In infected animals, these lipopolysaccharides interact with different receptors and exert potent anti-inflammatory effects, which may assist in repressing the host innate immune system. It is hoped that knowledge of the biosynthetic enzymes may lead to improved drug therapies.


Recommended Reading

  • Fujita, M. and Kinoshita, T. GPI-anchor remodeling: Potential functions of GPI-anchors in intracellular trafficking and membrane dynamics. Biochim. Biophys. Acta, 1821, 1050-1058 (2012) (DOI: 10.1016/j.bbalip.2012.01.004).
  • Guerin, M.E., Korduláková, J., Alzari, P.M., Brennan, P.J. and Jackson, M. Molecular basis of phosphatidyl-myo-inositol mannoside biosynthesis and regulation in Mycobacteria. J. Biol. Chem., 285, 33577-33583 (2010) (DOI: 10.1074/jbc.R110.168328).
  • Kinoshita, T., Maeda, Y. and Fujita, M. Transport of glycosylphosphatidylinositol-anchored proteins from the endoplasmic reticulum. Biochim. Biophys. Acta, 1833, 2473-2478 (2013) (DOI: 10.1016/j.bbamcr.2013.01.027).
  • Levental, I., Grzybek, M. and Simons, K. Greasing their way: lipid modifications determine protein association with membrane rafts. Biochemistry, 4930, 6305-6316 (2010) (DOI: 10.1021/bi100882y).
  • Maeda, Y. and Kinoshita, T. Structural remodeling, trafficking and functions of glycosylphosphatidylinositol-anchored proteins. Prog. Lipid Res., 50, 411-424 (2011) (DOI: 10.1016/j.plipres.2011.05.002).
  • Mishra, A.K., Driessen, N.N., Appelmelk, BJ. and Besra, G.S. Lipoarabinomannan and related glycoconjugates: structure, biogenesis and role in Mycobacterium tuberculosis physiology and host-pathogen interaction. FEMS Microbiol. Rev., 35, 1126-1157 (2011) (DOI: 10.1111/j.1574-6976.2011.00276.x).
  • Morita, Y.S., Fukuda, T., Sena, C.B.C., Yamaryo-Botte, Y., McConville, M.J. and Kinoshita, T. Inositol lipid metabolism in mycobacteria: Biosynthesis and regulatory mechanisms. Biochim. Biophys. Acta, 1810, 630-641 (2011) (DOI: 10.1016/j.bbagen.2011.03.017).
  • Paulick, M.G. and Bertozzi, C.R. The glycosylphosphatidylinositol anchor: A complex membrane-anchoring structure for proteins. Biochemistry, 47, 6991-7000 (2008) (DOI: 10.1021/bi8006324).
  • Tsai, Y.H., Grube, M., Seeberger, P.H. and Silva, D.V. Glycosylphosphatidylinositols of protozoan parasites. Trends Glycosci. Glycotechnol., 24, 231-243 (2012) (DOI: 10.4052/tigg.24.231).
  • Yu, S., Guo, Z., Johnson, C., Gu, G. and Wu, Q. Recent progress in synthetic and biological studies of GPI anchors and GPI-anchored proteins. Curr. Opinion Chem. Biol., 17, 1006-1013 (2013) (DOI: 10.1016/j.cbpa.2013.09.016).


Updated February 24, 2014