Structural characterization of the glycoinositol phospholipid membrane anchor of human erythrocyte acetylcholinesterase by fast atom bombardment mass spectrometry

The glycoinositol phospholipid membrane anchor of human erythrocyte acetylcholinesterase (EC 3.1.1.7) is composed of a glycan linked through a glucosamine residue to an inositol phospholipid that is resistant to the action of phosphatidylinositol-specific phospholipase C. Deamination cleavage of the glucosamine with nitrous acid released the inositol phospholipid which was purified by high performance liquid chromatography. Analysis by fast atom bombardment mass spectrometry with negative ion monitoring and by the complementary technique of collision-induced dissociation revealed molecular and daughter ions that indicated a plasmanylinositol with a palmitoyl group on an inositol hydroxyl. The intact membrane anchor was released from reductively methylated human erythrocyte acetylcholinesterase by proteolysis with papain or Pronase, deacylated by base hydrolysis, and purified by high performance liquid chromatography. Positive and negative ion fast atom bombardment mass spectrometry of the major products isolated by high performance liquid chromatography indicated the following structure for the complete glycoinositol phospholipid anchor. (formula; see text) Methylation of free amino groups by reduction with deuterium instead of hydrogen permitted determination of the number of free amino groups in individual fragment ions as further confirmation of structural assignments. The structure of the glycan portion of the human erythrocyte acetylcholinesterase membrane anchor appears to be similar to that described for Trypanosome brucei variant surface glycoprotein MITat 1.4 (variant 117) (Ferguson, M.A.J., Homans, S.W., Dwek, R.A., and Rademacher, T.W. (1988) Science 239, 753-759) except for the absence of a galactose antenna and the presence of a phosphorylethanolamine on the hexose adjacent to glucosamine.     

Lipid analysis of the glycoinositol phospholipid membrane anchor of human erythrocyte acetylcholinesterase. Palmitoylation of inositol results in resistance to phosphatidylinositol-specific phospholipase C

The glycoinositol phospholipid membrane anchor of human erythrocyte acetylcholinesterase (EC 3.1.1.7) contains a novel inositol phospholipid which in this and the accompanying paper (Roberts, W.L., Santikarn, S., Reinhold, V.N., and Rosenberry, T.L. (1988) J. Biol. Chem 263, 18776-18784) is shown to be a plasmanylinositol that is palmitoylated on the inositol ring. The inositol phospholipid was radiolabeled with the photoactivated reagent 3-(trifluoromethyl)-3-(m-[125I] iodophenyl)diazirine and characterized by various chemical and enzymatic cleavage procedures whose products were analyzed by thin layer chromatography and autoradiography or gas chromatography. Acidic methanolysis of human erythrocyte acetylcholinesterase (Ehu AChE) revealed 18:0 and 18:1 alkylglycerols (0.55 and 0.20 mol/mol AChE, respectively). Acetolysis was shown by TLC to release alkylacylglycerol acetates from Ehu AChE. Analysis by gas chromatography revealed that 83% of the alkylacylglycerol acetates contained an 18:0 or 18:1 1-alkyl group and a 22:4 (n - 6), 22:5 (n - 3), or 22:6 (n - 3) 2-acyl group. The inositol phospholipid is linked to the anchor by a glucosamine in glycosidic linkage, and deamination with nitrous acid cleaved the glycosidic linkage and released the phospholipid. The deamination and acetolysis products from Ehu AChE were purified by high performance liquid chromatography, and fatty acid analysis following acidic methanolysis of the purified products revealed that 2 fatty acid residues were associated with the deamination product and only one with the alkylacylglycerol acetolysis product. The other fatty acid residue was primarily palmitate and was indicated to be in ester linkage to an inositol hydroxyl(s). This linkage was shown to be responsible for the resistance of the inositol phospholipid to cleavage by Staphylococcus aureus phosphatidylinositol-specific phospholipase. Deacylation of the inositol phospholipid deamination product by treatment with base removed this palmitoyl group and facilitated release of alkyl- and alkylacylglycerol species by phosphatidylinositol-specific phospholipase C with concomitant formation of inositol 1-phosphate. In contrast, digestion of Ehu AChE with a recently reported anchor-specific phospholipase D resulted in release of plasmanic acids from the intact palmitoylated plasmanylinositol.     

Drosophila acetylcholinesterase: demonstration of a glycoinositol phospholipid anchor and an endogenous proteolytic cleavage

The presence of a glycoinositol phospholipid anchor in Drosophila acetylcholinesterase (AChE) was shown by several criteria. Chemical analysis of highly purified Drosophila AChE demonstrated approximately one residue of inositol per enzyme subunit. Selective cleavage by Staphylococcus aureus phosphatidylinositol-specific phospholipase C (PI-PLC) was tested with Drosophila AChE radiolabeled by the photoactivatable affinity probe 3-(trifluoromethyl)-3-(m-[125I]iodophenyl)diazirine [( 125I]TID), a reagent that specifically labels the lipid moiety of glycoinositol phospholipid-anchored proteins. Digestion with PI-PLC released 75% of this radiolabel from the protein. Gel electrophoresis of Drosophila AChE in sodium dodecyl sulfate indicated prominent 55- and 16-kDa bands and a faint 70-kDa band. The [125I]TID label was localized on the 55-kDa fragment, suggesting that this fragment is the C-terminal portion of the protein. In support of this conclusion, a sensitive microsequencing procedure that involved manual Edman degradation combined with radiomethylation was used to determine residues 2-5 of the 16-kDa fragment. Comparison with the Drosophila AChE cDNA sequence [Hall, L.M.C., & Spierer, P. (1986) EMBO J. 5, 2949-2954] confirmed that the 16-kDa fragment includes the N-terminus of AChE. Furthermore, the position of the N-terminal amino acid of the mature Drosophila AChE is closely homologous to that of Torpedo AChE. The presence of radiomethylatable ethanolamine in both 16- and 55-kDa fragments was also confirmed. Thus, Drosophila AChE may include a second posttranslational modification involving ethanolamine.     

Folate binding protein from kidney brush border membranes contains components characteristic of a glycoinositol phospholipid anchor.

A number of cell surface proteins have been shown to be anchored to the plasma membrane by a covalently attached glycoinositol phospholipid (GPL) in amide linkage to the C-terminus of the mature protein. We applied several criteria to establish that folate binding protein (FBP) in brush border membranes of rat kidney contains a GPL anchor. Brush border membranes were isolated and labeled with [3H]folate, and the complex of FBP and [3H]folate was shown to be released to the supernatant by incubation with purified bacterial phosphatidylinositol-specific phospholipase C (PIPLC) but not by incubation with a purified bacterial phosphatidylcholine-specific phospholipase C. The FBP-[3H]folate complex both in crude extracts and after FBP purification by ligand-directed affinity chromatography interacted with Triton X-114 micelles, and prior incubation with PIPLC prevented this detergent interaction. Individual residues characteristic of GPL anchors were found to be covalently associated with FBP following polyacrylamide gel electrophoresis in sodium dodecyl sulfate. These included glucosamine and ethanolamine, which were radiolabeled by reductive methylation and identified by chromatography on an amino acid analyzer, and inositol phosphate, which was inferred by Western blotting with an anti-CRD antisera. This antisera gave positive immunostaining only after FBP had been cleaved by PIPLC, a reliable diagnostic of a GPL anchor. The relationship between GPL-anchored FBP in biological membranes and soluble FBP in biological fluids also is discussed.     

Characterization of putative glycoinositol phospholipid anchor precursors in mammalian cells. Localization of phosphoethanolamine.

A number of mammalian cell surface proteins are anchored by glycoinositol phospholipid (GPI) structures that are preassembled and transferred to them in the endoplasmic reticulum. The GPIs in these proteins contain linear ethanolamine (EthN)-phosphate (P)-6ManManManGlcN core glycan sequences bearing an additional EthN-P attached to the Man residue (Man 1) proximal to GlcN. The biochemical precursors of mammalian GPI anchor structures are incompletely characterized. In this study, putative [3H]Man-labeled GPI precursors were obtained by in vitro GDP-[3H] Man labeling of HeLa cell microsomes and by in vivo [3H]Man labeling of class B and F Thy-1 negative murine lymphoma mutants known to accumulate incomplete GPIs. The high performance liquid chromatography-purified in vitro and accumulated in vivo GPI products were structurally analyzed by nitrous acid deamination, hydrofluoric acid, trifluoroacetic acid hydrolysis, biosynthetic labeling, and exoglycosidase treatment. The data were consistent with a biosynthetic scheme in which Man and EthN-P are added stepwise to the developing glycan. Several additional points were demonstrated: 1) putative mammalian GPI precursors contain incomplete core glycans corresponding to those in previously characterized trypanosome GPI precursors. 2) The proximal EthN-P found in mature mammalian GPI anchor structures is added to Man 1 prior to incorporation of Man 2 and Man 3. 3) Glycans in the incomplete GPIs that accumulate in classes B and F lymphoma mutants consist of Man2- and Man3GlcN in which EthN-P is linked to Man 1. 4) Distal EthN-P linked to the 6-position of Man, characteristic of the complete GPI core, is found both in a subsequent GPI species with the glycan sequence EthN-P-6ManMan(EthN-P----)ManGlcN and in a more polar GPI product.     

Novel allosteric sites on human erythrocyte acetylcholinesterase identified by two monoclonal antibodies.

Monoclonal antibodies against human erythrocyte acetylcholinesterase (acetylcholine acetylhydrolase EC 3.1.1.7) have been examined for inhibition of enzyme activity. Of sixteen antibodies analyzed, only one (C1B7) inhibited enzyme activity, indicating selection of an unusual susceptible site. The inhibitory activity of C1B7 was characterized and compared to another inhibitory antibody, AE-2, previously described by Fambrough et al. (Proc. Natl. Acad. Sci. USA 79, 1078, 1982). Maximal demonstrated inhibition was 84% for C1B7 and 72% for AE-2 and antibody inhibition of enzyme activity was equivalent for the reduced and alkylated acetylcholinesterase monomer and the intact dimer. The Ki (stoichiometry of the enzyme-antibody reaction estimated from enzyme kinetics) was 1.0 for C1B7 and 4.8 molecules of antibody per monomer of acetylcholinesterase for AE-2. The antibodies did not compete with one another for binding to acetylcholinesterase, indicating that they have different target epitopes on the enzyme. Antibody binding to the enzyme was not specifically affected by any of the anticholinesterase agents tested: (a) the irreversible esteratic site-directed inhibitor diisopropylfluorophosphate; (b) the reversible active site-directed inhibitors edrophonium, neostigmine, BW284c51, and carbachol; and (c) allosteric site-directed compounds propidium and gallamine. Kinetic analysis of their effects provide evidence that both antibodies decrease the catalytic rate of enzyme activity and have little or no effect on substrate binding.     

Replacement of the glycoinositol phospholipid anchor of Drosophila acetylcholinesterase with a transmembrane domain does not alter sorting in neurons and epithelia but results in behavioral defects.

Drosophila has a single glycoinositol phospholipid (GPI)-anchored form of acetylcholinesterase (AChE) encoded by the Ace locus. To assess the role that GPI plays in the physiology, of AChE, we have replaced the wild-type GPI-AChE with a chimeric transmembrane form (TM-AChE) in the nervous system of the fly. Ace null alleles provided a genetic background completely lacking in endogenous GPI-AChE, and Ace minigene P transposon constructs were used to express both GPI- and TM-AChE forms in the tissues where AChE is normally expressed. Control experiments with the GPI-AChE minigene demonstrated a threshold between 9 and 12% of normal AChE activity for adult viability. Ace mutant flies were rescued by GPI-AChE minigene lines that expressed 12-40% of normal activity and were essentially unchanged from wild-type flies in behavior. TM-AChE minigene lines were able to rescue Ace null alleles, although with a slightly higher threshold than that for GPI-AChE. Although rescued flies expressing GPI-AChE at a level of 12% of normal activity were viable, flies expressing 13-16% of normal activity from the TM-AChE transgene died shortly after eclosion. Flies expressing TM-AChE at about 30% of normal levels were essentially unchanged from wild-type flies in gross behavior but had a reduced lifespan secondary to subtle coordination defects. These flies also showed reduced locomotor activity and performed poorly in a grooming assay. However, light level and electron microscopic immunocytochemistry showed no differences in the localization of GPI- and TM-AChE. Furthermore, endogenous and ectopic-induced expression of both AChEs in epithelial tissues of the adult and embryo, respectively, showed that they were sorted identically. Most epithelial cells sorted GPI- and TM-AChE to the apical surface, but cuticle-secreting epithelia sorted both proteins basolaterally. Our data suggest that rather than having a primary role in protein sorting, the GPI anchor or AChE plays some other more subtle cellular role in neuronal physiology.     

The anion-transport protein of the human erythrocyte membrane. Studies on fragments produced by pepsin digestion.

We have studied the fragmentation by pepsin in 1 M-acetic acid of the erythrocyte anion-transport protein in erythrocyte membranes. The location of the fragments obtained was determined by radioiodinating the protein with the use of lactoperoxidase, and identifying the labelled peptides obtained in peptide "maps" of thermolysin digests of the fragments. Three of the fragments were found to be related overlapping products, and shared a common C-terminus. The major site of pepsin cleavage leading to the C-termini of these fragments was shown to be close to the major site of extracellular cleavage of the protein by proteinases active at a neutral pH. Another two fragments were isolated and shown to be derived from the C-terminal portion of the protein. No well-defined large radioactive fragments of the protein were solubilized from the membrane by pepsin in 1 M-acetic acid, the bulk of the radioactivity attributable to the anion transport protein being recovered in very small fragments that could not be resolved by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. Our results suggest that the polypeptide chain of the anion-transport protein emerges at the extracellular face of the membrane 8000-13000 daltons on the N-terminal side of the major site of extracellular cleavage of the protein by proteinases that are active at a neutral pH.     

Release of the cellular prion protein from cultured cells after loss of its glycoinositol phospholipid anchor

Secreted forms of the sialoglycoprotein designated cellularprion protein (PrP^C) have been identified that cannot be explainedby alternative splicing. We report that secreted forms of PrP^Cderive from precursors that are bound to the plasma membraneby glycoinositol phospholipid (GPI) anchors. Secreted PrP^C slowlyappeared in the culture medium of metabolically radiolabelledcells after incubations of 8—24 h. Digestion of nascentPrP^C with phosphatidylinositol-specific phospholipase C (PIPLC)prevented the appearance of secreted PrP^C. Secreted PrP^c partitionedinto the aqueous phase of Triton X–114 like PIPLC-releasedPrP^C. While the M_r of PIPLC-released PrP^C was reduced 2–4kDa after treatment with aqueous hydroflouric acid, which removesthe entire GPI anchor modification, the M_r of secreted PrP^Cwas unchanged. Both PIPLC-released and secreted PrP^c were recognizedby antiserum raised against a synthetic C-terminal peptide correspondingto residues 220–233 (amino acid 231 is the site of GPIattachment). We conclude that GPI-anchored PrP^C is post-translationallyprocessed to remove most, if not all, of the GPI modificationand then shed into culture medium. Whether PrP^C is shed afterproteolysis near the C-terminus remains to be established. Aminority of PrP^C in normal Syrian hamster brain partitionedinto the aqueous phase of Triton X–114 like shed PrP^Csuggesting physiological significance. post-translational prion protein secretion sialoglycoprotein     

Structure of the anion-transport protein of the human erythrocyte membrane. Further studies on the fragments produced by proteolytic digestion.

The topology of the human erythrocyte membrane anion-transport protein (band 3) has been investigated by isolation and peptide 'mapping' of the major and minor fragments derived from proteolytic cleavage of the lactoperoxidase 125I-labelled protein in erythrocytes and erythrocyte membranes. The content, in each fragment, of lactoperoxidase 125I-labelled sites (which have a known location in the extracellular or cytoplasmic domain of the protein), together with the location of the sites of proteolytic cleavage yielding the fragments, has allowed us to determine the alignment of the fragments on the linear amino acid sequence and to infer the topology of the polypeptide in the membrane. The results suggest that a region in the C-terminal portion of the polypeptide forms part of the cytoplasmic domain of the protein in addition to a large N-terminal segment. The membrane-bound regions of the protein are located in the C-terminal two-thirds of the molecule. In this region the polypeptide chain traverses the membrane at least four times and an additional loop of polypeptide is either embedded in the membrane or also penetrates through it to the other surface. The location of the lectin receptors on the protein and the site of binding of an anion-transport inhibitor have also been studied.     

Identification of amine components in a glycolipid membrane-binding domain at the C-terminus of human erythrocyte acetylcholinesterase

Purified human erythrocyte acetylcholinesterase was labeled by reductive radiomethylation with saturating amounts of [14C]formaldehyde and sodium cyanoborohydride. Acid hydrolysis and automated amino acid analysis permitted both identification of radiomethylated components by their coelution with radiomethylated standards and quantitation of these components. The methylated N-terminal amino acids glutamate and arginine were observed at levels of 0.66 and 0.34 residues, respectively, per 70-kilodalton subunit, and lysine residues were methylated on their epsilon-amino groups to a level of 7.40 residues per subunit [Haas, R., & Rosenberry, T.L. (1985) Anal. Biochem. 148, 154-162]. In addition, each subunit contained 1.35 residues of methylated ethanolamine and 0.98 residue of methylated glucosamine. Papain digestion cleaved the intact enzyme into two fragments, an enzymatically active hydrophilic fragment and a small hydrophobic fragment that represented the membrane-binding domain. The radiomethylated amino acids were quantitatively retained in the hydrophilic fragment, while the methylated ethanolamine and glucosamine were confined exclusively to the hydrophobic domain fragment. This fragment included the C-terminal dipeptide of the subunit. Peptide sequencing by manual Edman methods was combined with radiomethylation to demonstrate the sequence His-Gly-ethanolamine-Z for the hydrophobic domain fragment. The ethanolamine residue in this sequence is in amide linkage to the C-terminal Gly and is clearly distinct from the ethanolamine residues in Z which are susceptible to radiomethylation in the intact enzyme. Since Z also includes glucosamine and 2 mol of fatty acids [Roberts, W.L. & Rosenberry, T.L. (1985) Biochem. Biophys. Res. Commun. 133, 621-627], we conclude that the membrane-binding domain of human erythrocyte acetylcholinesterase is a covalently linked glycolipid at the C-termini of the subunits.(ABSTRACT TRUNCATED AT 250 WORDS)     

Multiple molecular forms of purified human erythrocyte acetylcholinesterase.

1. Human erythrocyte acetylcholinesterase was solubilized by Triton X-100 and purified by affinity chromatography to a specific activity of 3800 IU/mg of protein. The yield of the purified enzyme was 25--45%. 2. Gel filtration on Sepharose 4-B in the presence of Triton X-100 revealed one peak of enzyme activity with a Stokes' radius of 8.7 nm. Density gradient centrifugation in 0.1% Triton X-100 showed one peak of enzyme activity with an S4 value of 6.3S. 3. Isoelectric focusing in Triton X-100 resolved the enzyme into five molecular forms with isoelectric points of 4.55, 4.68, 4.81, 4.98 and 5.18. Upon incubation with neuraminidase the enzyme activity in the first four forms was decreased with a concommitant increase in activity in the form with the higher isoelectric point. 4. After removal of excess Triton X-100 on Bio-Gel HTP, polyacrylamide gel electrophoresis showed seven bands of protein and corresponding bands of enzyme activity. Density gradient centrifugation of the detergent-depleted enzyme at high ionic strength revealed five multiple molecular forms with S4 values of 6.3 S, 10.2 S, 12.2 S, 14.2 S and 16.3 S. At low ionic strength, higher aggregates were observed in addition to the other forms. Dodecylsulfate-polyacrylamide gel electrophoresis gave one subunit only with an apparent molecular weight of 80 000. 5. These results suggest that human erythrocyte acetylcholinesterase, solubilized by Triton X-100, exists in various forms differing in net charge but of apparently similar molecular dimensions. After removal of the detergent, forms with different molecular sizes are observed.     

Molecular forms of purified human erythrocyte membrane acetylcholinesterase investigated by crosslinking with diimidates.

Several molecular forms of human erythrocyte membrane acetylcholinesterase have been studied after crosslinking with bifunctional diimidates. The crosslinked products were analysed by centrifugation on linear sucrose density gradients containing Triton X-100. Molecular weights of covalently linked oligomers were estimated by sodium dodecylsulfate gel electrophoresis. It was shown that acetylcholinesterase crosslinked in absence of Triton X-100 consists of molecular forms built up by dimeric protomers. These dimers were identical with the enzymatically active species sedimenting with 6.5S in linear sucrose density gradients.     

Permethylation and tandem mass spectrometry of oligosaccharides having free hexosamine: analysis of the glycoinositol phospholipid anchor glycan from the scrapie prion protein

Permethylation of the glycan isolated from the glycoinositol phospholipid (GPI) anchor of the scrapie prion protein (PrPSc) trimethylates a free hexosamine to form a quarternary ammonium salt, substantially increasing the sensitivity for analysis by mass spectrometry. This derivatization induces specific fragmentation reactions in collision-induced dissociation spectra obtained on a four-sector tandem mass spectrometer, identifying the branching pattern of the PrPSc GPI glycan.     

Analysis of human erythrocyte membrane vesicles produced by shearing

Shearing of ghosts in a French pressure cell produces three classes of microvesicles that differ from endocytic vacuoles, exocytic vacuoles, and inside-out vesicles. It was thought that an analysis of these vesicles might provide some clues about the assembly of proteins within the human erythrocyte membrane. The microvesicles were separated into three visible bands, labeled top, middle, and bottom, and assayed for activity of Mg⁺⁺-ATPase, Na⁺, K⁺-ATPase, acetylcholinesterase, glyceraldehyde-phosphate dehydrogense, and NADH oxidoreductase. Their proteins were also characterized by polyacrylamide gel electrophoresis with both Coomassie blue staining, to assess total protein content and distribution, and PAS-staining, to characterize sialoglycopeptides. In order to minimize problems inherent in ghost preparation, Dodge or hypotonic ghosts and glycol or isotonic ghosts were used in all studies. Middle membrane vesicles most resembled intact ghosts. Top vesicles had reduced levels of NADH oxidoreductase and more PAS-2 at the expense of PAS-1. The bottom vesicle class was very much enriched with PAS-1 at the expense of PAS-2, and PAS-3 was completely absent. In addition bottom vesicles had highest NADH oxidoreductase activity but lowest activity of all the other enzymes measured. These vesicle classes could not have been produced by tangential shearing through the membrane, nor could radial shearing through a membrane in which all proteins were free to move laterally have accounted for the three discrete vesicle classes or for their different patterns of enzymes and proteins. The analysis of the microvesicles produced by shearing is most consistent with radial shearing through membranes where there may be fixed domains superimposed on the basic fluid-mosaic structure.     

A study of the mechanism by which some amphiphilic drugs affect human erythrocyte acetylcholinesterase activity.

The effects of the local anaesthetics procaine, tetracaine and lidocaine and of the antidepressant imipramine on human erythrocyte acetylcholinesterase were investigated. All four amphiphilic drugs inhibited enzymic activity, the IC50 (the concentration causing 50% inhibition) being about 0.40 mM for procaine, 0.05 mM for tetracaine, 0.70 mM for imipramine and 7.0 mM for lidocaine. Procaine and tetracaine inhibited acetylcholinesterase activity competitively at concentrations at which they did not perturb the physical state of the membrane lipid environment, as assessed by steady-state fluorescence polarization, whereas lidocaine and imipramine displayed mixed inhibition kinetics at concentrations at which they induced an enhancement of membrane fluidity. The question was addressed as to whether membrane integrity is a prerequisite for imipramine and lidocaine action. Membrane solubilization by 1% Triton X-100 and a decrease, by dilution, in the detergent concentration to 0.05% [which is above the Triton X-100 critical micelle concentration (c.m.c.)] did not substantially affect the inhibitory potency of the two amphiphilic drugs at their IC50; in the presence of increasing detergent concentrations the inhibitory potency of imipramine was gradually decreased, but not abolished, whereas the inhibitory effect of lidocaine was only slightly diminished, even at 1% Triton X-100. It is suggested that neither competitive nor mixed inhibition kinetics arise from conformational changes of the protein driven by a modification of the physical state of the lipid environment, but from a direct interaction of the amphiphilic drugs with acetylcholinesterase. In particular, the partial loss of the inhibitory potency of imipramine and lidocaine that is observed upon increasing Triton X-100 concentration well above its c.m.c. could be explained in terms of amphiphile partition in detergent micelles and, in turn, of a decreased effective concentration of the two inhibitors in the aqueous phase.