Identification of the Palmitoylation Site in Rat Myelin P0 Glycoprotein

Abstract: P₀ glycoprotein, the major protein of PNS myelin, contains approximately 1 mol of covalently bound long-chain fatty acids. To determine the chemical nature of the fatty acid-protein linkage, P₀ was labeled in rat sciatic nerve slices with [³H]palmitic acid and subsequently treated with various reagents. The protein-bound palmi-tate was released by incubation with the reducing agents dithiothreitol and 2-mercaptoethanol, and with 1 M hydrox-ylamine at pH 7.5. In addition, P₀ was deacylated by treatment with 10 m M NaBH₄ with the concomitant production of [³H]hexadecanol, indicating that the fatty acid is bound in a thioester linkage. This conclusion was supported further by the fact that deacylation with hydroxylamine generated free thiol groups, which were titrated with [¹⁴C]-iodoacetamide. To identify the cysteine residue involved in the thioester linkage, [¹⁴C]carboxyamidomethylated P₀was digested with trypsin and the resulting peptides analyzed by reversed-phase HPLC. Identification of the radioactive protein fragments by amino acid analysis and amino-terminal peptide sequencing revealed that Cys¹⁵³ in rat P₀ glycoprotein is the acylation site. The acylated cysteine is located at the junction of the putative transmem-brane and cytoplasmic domains. This residue is also present in the P₀ glycoprotein of other species, including human, bovine, mice, and chicken.     

A Phorbol Ester-Sensitive Kinase Catalyzes the Phosphorylation of P0 Glycoprotein in Myelin

The proposed structural protein of peripheral nerve myelin, P0, has been shown to have several covalent modifications. In addition to being glycosylated, sulfated, and acylated, P0 is phosphorylated, with the intracellular site of this latter addition being in question. By employing nerve injury models that exhibit different levels of P0 biosynthesis in the absence and presence of myelin assembly, we have examined the cellular location of P0 phosphorylation. It is demonstrated that there is comparable P0 phosphorylation in both normal and crush-injured adult rat sciatic nerves, although the level of biosynthesis of P0 differs between these myelin maintaining and actively myelinating nerve models, respectively. The glycoprotein does not appear to be phosphorylated readily in the transected adult sciatic nerve, a preparation in which P0 biosynthesis is observed but that lacks myelin membrane. These observations suggest that the modification is not associated with the biosynthesis or maturation of P0 in the endoplasmic reticulum or Golgi, but that it instead occurs after myelin assembly. That P0 phosphorylation occurs in the normal nerve even when translation is inhibited by cycloheximide treatment lends further support to this conclusion. P0 is shown to be phosphorylated on one or more serine residues, with all or most of the phosphate group(s) being labile as evidenced by pulse-chase analysis. Addition of a biologically active phorbol ester, 12-O-tetradecanoylphorbol-13-acetate or 4 beta-phorbol 12,13-dibutyrate, substantially increases the extent of [32P]orthophosphate incorporation into the glycoprotein of normal and crushed nerve but not transected nerve. Biologically inactive 4 alpha-phorbol 12,13-didecanoate has no effect on P0 phosphorylation. Similarly, the addition of the cyclic AMP analog 8-bromo-cyclic AMP causes no appreciable changes in P0 labeling. These findings indicate that the phorbol ester-sensitive enzyme, protein kinase C, may be responsible for the phosphorylation of P0 within the myelin membrane.     

Phosphorylation of P0 Glycoprotein in Peripheral Nerve Myelin

The P0 protein in mammalian PNS myelin is known to undergo several posttranslational modifications, such as glycosylation, acylation, sulfation, and phosphorylation. Phosphorylation of purified P0 protein in vitro was studied comparatively using three enzymes, i.e., calcium/phospholipid-dependent protein kinase (protein kinase C), calcium/calmodulin-dependent protein kinase II (CaM kinase II), and the catalytic subunit of cyclic AMP-dependent protein kinase (A kinase). The phosphorylation of P0 protein by CaM kinase II was the greatest, followed by that by protein kinase C; phosphorylation by A kinase, however, was much lower. In order to identify phosphorylation sites, P0 protein was phosphorylated with [32P]ATP and each kinase and then digested with lysylendopeptidase. The resulting phosphopeptides were isolated by HPLC. Subsequent amino acid sequence analysis and comparison with the known sequence of P0 protein revealed that Ser181 and Ser204 were strongly phosphorylated by both protein kinase C and CaM kinase II. In addition, Ser214 was also phosphorylated by protein kinase C, but not by CaM kinase II. Because all of these sites are located in the cytoplasmic domain of P0 protein, phosphorylation may be important for maintenance of the major dense line of PNS myelin.     

Formation of a Disulfide Bond in the Immunoglobulin Domain of the Myelin P0 Protein Is Essential for Its Adhesion

Abstract: It is widely accepted, although never demonstrated, that the formation of a disulfide bond in the majority of immunoglobulin (Ig)-like domains stabilizes their final conformation and thus is essential to their functioning as adhesion/recognition molecules. The myelin P₀ protein, which has been shown directly to behave as a homophilic adhesion molecule, contains a single Ig-like domain, stabilized by a putative Cys²¹-Cys⁹⁸ disulfide bond. To test if this bond is indeed necessary to the adhesive function of P₀, the nucleotides in the P₀ cDNA coding for Cys²¹ were altered to code for an alanine. The mutated P₀ cDNA was transfected into Chinese hamster ovary cells, expression of the mutated P₀ protein was characterized, and the adhesiveness of Cys²¹-mutated P₀-expressing cells and that of cells expressing equivalent surface amounts of the unmutated protein were compared. It was found, as we previously reported, that incubation of a single cell suspension of the unmutated P₀-expressing cells resulted in the rapid formation of large aggregates. In contrast, after a similar incubation the cells expressing the Cys²¹-mutated P₀ were still mostly single cells, a result indistinguishable from that observed with the control transfected cells. This suggests that the P₀ protein, when mutated at Cys²¹, does not behave as a homophilic adhesion molecule, which in turn implies that the formation of an Ig domain disulfide bond is essential to the functioning of this molecule.     

Myelin P0 Glycoprotein and a Synthetic Peptide Containing the Palmitoylation Site Are Both Autoacylated

Abstract: P₀, the major protein of the PNS myelin, is palmitoylated at the cytoplasmic Cys¹⁵³. To gain insights into the mechanism of P₀ acylation, the in vitro palmitoylation of both P₀ and a synthetic Cys¹⁵³-containing octapeptide was studied. Incubation of PNS myelin membranes or isolated P₀ with [³H]palmitoyl-CoA resulted in specific labeling of this protein, suggesting that the reaction is nonenzymatic. Incorporation of the labeled fatty acid into P₀ was not affected by boiling the isolated P₀ for 15 min before incubation or by adding sciatic nerve homogenate to the reaction mixture, which confirms the nonezymatic nature of the reaction. After chemical deacylation, P₀ was palmitoylated at a higher rate, suggesting that the original site was reacylated. Furthermore, tryptic digestion and peptide mapping showed that the same sites are acylated in vitro as in nerve slices indicating that the reaction has physiological significance. On incubation with [¹⁴C]palmitoyl-CoA, the synthetic peptide encompassing the natural P₀ acylation site (I¹⁵⁰RYCWLRR¹⁵⁷) was also spontaneously acylated at the cysteine residue. Thus, the integrity of the protein is not required for the nonenzymatic transacylation reaction. At pH 7.4 and 37°C, peptide palmitoylation followed a second-order reaction (k₂ = 246 ± 6 M^?1 min^?1) and is likely a bimolecular nucleophilic substitution with the peptide thiolate attacking the highly reactive thioester bond in palmitoyl-CoA. The activation energy calculated from the Arrhenius plot is ～2 kcal/mol and much lower than that of enzyme-catalyzed transacylations. Finally, two other P₀ peptides (V¹²¹PTRYG¹²⁶ and K¹⁰⁹TSQVTL¹¹⁵) as well as various unrelated thiol-containing compounds, including cysteine, glutathione, pressinoic acid (CYFQNC), and crustacean cardioactive peptide (PFCNAFTGC), were not autoacylated. These results indicate that the IRYCWLRR peptide represents a particular structural motif and/or has some chemical features that allow the reaction to occur spontaneously.     

Age-Dependent Changes in the Oligosaccharide Structure of the Major Myelin Glycoprotein, P0

The single oligosaccharide moiety of the major myelin glycoprotein, P0, resides in an immunoglobulin-like domain that appears to participate in homophilic binding. The studies presented here indicate that the structure of the P0 oligosaccharide from rat nerve changes as a function of Schwann cell age. Examination of 5-day-old nerve revealed that P0 contained predominantly endo-beta-N-acetylglucosaminidase H (endo H)-resistant, complex-type oligosaccharide. In contrast, P0 from adult rats had mostly endo H-sensitive carbohydrate, indicating the presence of appreciable high-mannose and/or hybrid-type oligosaccharide on the glycoprotein. The endo H-sensitive and -resistant P0 of adult nerve could be readily phosphorylated by protein kinase C, as could the complex-type P0 from 5-day-old nerve. This suggests that the glycoprotein progresses to the plasma membrane and myelin regardless of the type of oligosaccharide chain. Analysis of 35SO4(2-)-labeled P0 showed that the sulfate group was found on both endo H-sensitive and -resistant oligosaccharide. The endo H-sensitive P0 carbohydrate from adult nerve appears to be primarily of the hybrid type, as evidenced by (a) the elution profile of [3H]mannose-labeled P0 glycopeptides from adult nerve during concanavalin A chromatography and (b) the inability of P0 from adult nerve to interact with Galanthus nivalis agglutinin. The observed age-dependent changes of P0 oligosaccharide may modify the binding properties of this myelin glycoprotein.     

Immunological Nonidentity of 19K Protein and TP0 in Peripheral Nervous System Myelin

Peripheral nervous system myelin contains as the major structural protein a glycoprotein known as P0. Another glycoprotein present in smaller amounts, known as the 19K or X protein, has been previously identified as derived from P0 and identical with the main tryptic degradation product of P0 (TP0). Although both P0 and 19K protein incorporated fucose in vitro and stained on polyacrylamide gels with the periodic acid-Schiff stain for carbohydrate, only the P0 blotted to nitrocellulose paper showed immunoreactivity to an antibody to P0, whereas the 19K protein did not. Furthermore, when P0 was hydrolyzed with trypsin or elastase, the main degradation products reacted with P0 on immunoblots, whereas the 19K protein showed no immunoreactivity. From these studies and those of others, it may be concluded that the 19K protein shows some similarities to TP0, but probably has a different structure. P0 and 19K protein do not appear to be related as shown by lack of cross-immunoreactivity.     

Myelin P0 Glycoprotein: Identification of the Site Phosphorylated In Vitro and In Vivo by Endogenous Protein Kinases

Abstract: Myelin membrane prepared from mouse sciatic nerve possesses both kinase and substrates to incorporate [³²P]PO₄³⁻ from [γ-³²P]ATP into protein constituents. Among these, P₀ glycoprotein is the major phosphorylated species. To identify the phosphorylated sites, P₀ protein was in vitro phosphorylated, purified, and cleaved by CNBr. Two ³²P-phosphopeptides were isolated by HPLC. The exact localization of the sequences around the phosphorylated sites was determined. The comparison with rat P₀ sequence revealed, besides a Lys¹⁷² to Arg substitution, that in the first peptide, two serine residues (Ser¹⁷⁶ and Ser¹⁸¹) were phosphorylated, Ser¹⁷⁶ appearing to be modified subsequently to Ser¹⁸¹. In the second peptide, Ser¹⁹⁷, Ser¹⁹⁹, and Ser²⁰⁴ were phosphorylated. All these serines are clustered in the C-terminal region of P₀ protein. This in vitro study served as the basis for the identification of the in vivo phosphorylation sites of the C terminal region of P₀. We found that, in vivo, Ser¹⁸¹ and Ser¹⁷⁶ are not phosphorylated, whereas Ser¹⁹⁷, Ser¹⁹⁹, Ser²⁰⁴, Ser²⁰⁸, and Ser²¹⁴ are modified to various extents. Our results strongly suggest that the phosphorylation of these serine residues alters the secondary structure of this domain. Such a structural perturbation could play an important role in myelin compaction at the dense line level.     

Regulation of Myelin P0 Glycoprotein Synthesis in Cultured Rat Schwann Cells and Continuous Rat PNS Cell Lines

We studied the effects of agents that raise intracellular cyclic AMP on synthesis of myelin components by cultured neonatal rat sciatic nerve Schwann cells and by continuous PNS cell lines derived from the fusion of neonatal rat sciatic nerve Schwann cells with rat RN22 Schwannoma. Treatment with N6,2'-O-dibutyryl cyclic AMP (dibutyryl cyclic AMP) caused a fourfold increase in Schwann cell incorporation of 35SO4 into sulfogalactosylceramide (sulfatide), and elicited a 10- to 20-fold increase in such incorporation by the continuous PNS cell lines; a similar effect on PNS cell line sulfatide radiolabelling was obtained with forskolin. Cultured Schwann cells expressed barely detectable levels of myelin P0 glycoprotein (P0) mRNA and myelin basic protein (MBP) mRNA. Treatment of the Schwann cells with axolemmal fragments or with dibutyryl cyclic AMP did not elicit a detectable increase in the levels of these mRNAs. The PNS cell lines constitutively expressed much higher levels of P0 mRNA than did the Schwann cells, and synthesized immunochemically demonstrable P0 glycoprotein, but did not express MBP. Treatment of the PNS cell lines with dibutyryl cyclic AMP markedly reduced expression of P0 mRNA and also diminished immunoreactive P0 glycoprotein. These PNS cell lines should prove useful for further studies of the control of Schwann cell differentiation.     

Comparison of the N-Linked Oligosaccharide Structures of the Two Major Human Myelin Glycoproteins MAG and P0: Assessment of the Structures Bearing the Epitope for HNK-1 and Human Monoclonal Immunoglobulin M Found in Demyelinating Neuropathy

The epitope for HNK-1 and patient's monoclonal autoantibodies in demyelinating polyneuropathy associated with immunoglobulin M gammopathy is borne by different types of N-linked oligosaccharide structures in human P0 and myelin-associated glycoprotein (MAG). Fourteen glycopeptide fractions bearing different oligosaccharide structures were obtained from either MAG or P0 glycopeptides by serial lectin affinity chromatography on concanavalin A-Sepharose, Phaseolus vulgaris erythrophytohemagglutinin-agarose, Pisum sativum agglutinin-agarose, and Phaseolus vulgaris leucophytohemagglutinin-agarose. As shown by dot-TLC plate immunostaining, the same MAG and P0 glycopeptide fractions were recognized by HNK-1 and patient's immunoglobulin M, confirming that these antibodies display similar specificities. The antigenic carbohydrate was present in glycopeptide fractions that either interact with Pisum sativum agglutinin-agarose or were bound by Aleuria aurantia agglutinin-digoxigenin, indicating that these structures contained alpha(1-6)fucose residues. This study demonstrates that the L2/HNK-1 epitope is borne mainly or even exclusively by N-linked oligosaccharide structures alpha(1-6)fucosylated in the core.     

Production and Characterization of Monoclonal Antibodies to the Major Peripheral Myelin Glycoprotein P0

Several monoclonal antibodies were generated against the major glycoprotein P0 of human peripheral nervous system myelin. Antibodies were selected for their reactivity with P0 in Western blots. The antibodies were of the immunoglobulin G subclass and reacted with the glycopeptidase F-treated P0, indicating that the reactive epitope resides in the protein backbone. In fresh frozen and paraffin-embedded sections of central and peripheral nervous system of rat and human, P0 antibody 592 reacted with myelin sheaths of peripheral, but not central, nervous system.     

Comparison of the N-Linked Oligosaccharide Structures of the Two Major Human Myelin Glycoproteins MAG and P0: Assessment and Relative Occurrence of Oligosaccharide Structures by Serial Lectin Affinity Chromatography of 14C-Glycopeptides

The N-linked oligosaccharide structures of human myelin-associated glycoprotein (MAG) and P0 have been characterized by serial lectin affinity chromatography (SLAC) of 14C-glycopeptides. 14C-Glycopeptides were prepared from purified MAG derivative and P0 by extensive proteolytic digestion and N-14C-acetylation. Assuming that all the 14C-glycopeptides were radiolabelled to the same specific radioactivity, the relative occurrence of the oligosaccharide structures was correlated to the amount of incorporated radioactivity. Sixteen and 15 fractions were generated by SLAC of MAG and P0 14C-glycopeptides, respectively. Despite this tremendous structural heterogeneity, the oligosaccharide "fingerprints" of MAG and P0 obtained by SLAC displayed similarities: (a) of the three types of N-linked oligosaccharides, the complex type accounted for 80.4% and 94.9% of MAG and P0 radioactivity, respectively; (b) biantennary complex oligosaccharides were the major structures present on MAG and P0; (c) approximately 60% of MAG and P0 oligosaccharides possessed a bisecting N-acetylglucosamine residue; and (d) large amounts of oligosaccharides with an alpha(1-6)fucose residue were found in both MAG and P0 and, noticeably, approximately 25% of the tri- and/or tetraantennary and approximately 90% of the bisected biantennary oligosaccharides of both glycoproteins contained alpha(1-6)fucose residues in the core. This study demonstrates that MAG and P0, both belonging to the immunoglobulin superfamily, display structural similarities in their N-linked oligosaccharide contents.     

Partial Structure and Mapping of the Human Myelin P2 Protein Gene

Abstract: The myelin P₂ protein, a 14,800-Da cytosolic protein found primarily in peripheral nerves, belongs to a family of fatty acid binding proteins. Although it is similar in amino acid sequence and tertiary structure to fatty acid binding proteins found in the liver, adipocytes, and intestine, its expression is limited to the nervous system. It is detected only in myelin-producing cells of the central and peripheral nervous systems, i.e., the oligodendrocytes and Schwann cells, respectively. As part of a program to understand the regulation of expression of this gene, to determine its function in myelin-producing cells, and to study its role in peripheral nerve disease, we have isolated and characterized overlapping human genomic clones encoding the P₂ protein. We report here on the partial structure of this gene, and on its localization within the genome. By using a panel of human-hamster somatic cell hybrids and by in situ hybridization, we have mapped the human P₂ gene to segment q21 on the long arm of chromosome 8. This result identifies the myelin P₂ gene as a candidate gene for autosomal recessive Charcot-Marie-Tooth disease type 4A.     

Degradation of the P0, P1, and Pr, Proteins in Peripheral Nervous System Myelin by Plasmin: Implications Regarding the Role of Macrophages in Demyelinating Diseases

Activated macrophages secrete a variety of neutral proteinases, including plasminogen activator. Since macrophages are implicated in primary demyelination in the peripheral nervous system (PNS) in Guillain-Barré syndrome and experimental allergic neuritis, we have investigated the ability of plasmin and of conditioned media from cultured macrophages, in the presence of plasminogen, to degrade the proteins in bovine and rat PNS myelin. The results indicate that (a) the major glycoprotein P0 and the basic P1 and Pr proteins in PNS myelin are extremely sensitive to plasmin, perhaps more so than is the basic protein in CNS myelin; (b) the initial product of degradation of P0 by plasmin has a molecular weight higher than that of the "X" protein; (c) large degradation products of P0 are relatively insensitive to further degradation; and (d) the neuritogenic P2 protein in PNS myelin is quite resistant to the action of plasmin. Results similar to those with plasmin were obtained with conditioned media from macrophages and macrophage-like cell lines together with plasminogen activator, and the degradation of the PNS myelin proteins, Po and P1, under these conditions was inhibited by p-nitrophenylguanidinobenzoate, an inhibitor of plasmin and plasminogen activator. The results suggest that the macrophage plasminogen activator could participate in inflammatory demyelination in the PNS.     

Incorporation of P0 Protein into Liposomes: Demonstration of a Two-Domain Structure by Immunochemical and PAGE Analysis

The amphiphilic nature of P0, the major glycoprotein of peripheral nerve myelin, has been suggested previously. In the present study, purified P0 from human peripheral nerve myelin was incorporated into an artificial lipid bilayer consisting of dimyristoyl lecithin and cholesterol. The liposomes were fractionated on a sucrose gradient. The continued expression of P0 antigenicity by the liposomes was shown by specific complement consumption with a multivalent antiserum against P0 or with an IgM monoclonal antibody. Both antibodies recognized P0 expressed on the surface of peripheral nerve myelin and the P0 liposomes. P0 liposomes and peripheral nerve myelin treated with trypsin lost the surface determinant that reacted with the monoclonal antibody. Analysis of the trypsin-treated liposomes and peripheral nerve myelin by polyacrylamide gel electrophoresis revealed molecular weights for this protein of 19,500 and 20,500, respectively. Similar treatment of the P0 in the fluid phase resulted in many smaller fragments. These results indicate that P0 consists of two domains, a hydrophilic domain accessible to trypsin digestion and a hydrophobic domain, which is potentially trypsin-sensitive, but shielded by the lipid bilayer. Binding studies with an anti-P0 monoclonal antibody and polyacrylamide gel analysis of the lipid-shielded P0 fragment in liposomes and peripheral nerve myelin suggest that the orientation of the protein in the liposome is similar to that in peripheral nerve myelin.     

Myelin Structure and Composition in Zebrafish

To establish a standard for genotype/phenotype studies on the myelin of zebrafish (Danio rerio), an organism increasingly popular as a model system for vertebrates, we have initiated a detailed characterization of the structure and biochemical composition of its myelinated central and peripheral nervous system (CNS; PNS) tissues. Myelin periods, determined by X-ray diffraction from whole, unfixed optic and lateral line nerves, were approximately 153 and approximately 162 Angstrom, respectively. In contrast with the lability of PNS myelin in higher vertebrates, zebrafish lateral line nerve myelin exhibited structural stability when exposed to substantial changes in pH and ionic strength. Neither optic nor lateral line nerves showed swelling at the cytoplasmic apposition in CaCl(2)-containing Ringer's solution, in contrast with nerves from other teleost and elasmobranch fishes. Zebrafish optic nerve showed greater stability against changes in NaCl and CaCl(2) than lateral line nerve. The nerves from zebrafish having mutations in the gene for myelin basic protein (mbpAla2Thr and mbpAsp25Val) showed similar myelin periods as the wildtype (WT), but gave approximately 20% less compact myelin. Analysis of proteins by SDS-PAGE and Western blotting identified in both CNS and PNS of WT zebrafish two orthologues of myelin P0 glycoprotein that have been characterized extensively in trout--intermediate protein 1 (24 kDa) and intermediate protein 2 (28 kDa). Treatment with endoglycosidase-F demonstrated a carbohydrate moiety of approximately 7 kDa, which is nearly threefold larger than for higher vertebrates. Thin-layer chromatography for lipids revealed a similar composition as for other teleosts. Taken together, these data will serve as a baseline for detecting changes in the structure and/or amount of myelin resulting from mutations in myelin-related genes or from exogenous, potentially cytotoxic compounds that could affect myelin formation or stability.     

Kinetics of Entry of P0 Protein into Peripheral Nerve Myelin

Abstract: Sciatic nerves from 9-day-old rat pups were removed, sliced into 0.4-mm sections, and incubated with [³H]fucose or [¹⁴C]glycine precursors. The nerve slice system gave nearly linear incorporation of [³H]fucose as a function of time for 3 h, after an initial lag of ˜30 min for homogenate and ˜60 min for myelin. Incorporation of [³H]fucose at constant specific radioactivity was directly proportional to exogenous fucose levels over the range 3.0 × 10⁻⁸ m to 1.5 × 10⁻⁶ m . Analysis of labeled proteins by sodium dodecyl sulfate polyacrylamide gel electrophoresis showed that greater than 50% of labeled glycoprotein was P₀, with no other major constituents. This system was used in fucose-chase experiments to determine that a period of ˜20 min elapses between fucosylation and assembly of P₀ into myelin. Cycloheximide inhibition of protein synthesis was used to determine that a period of ˜33 min elapses between protein synthesis and appearance of P₀ myelin.