CYTOPLASMIC DOMAINS OF THE MYELIN-ASSOCIATED GLYCOPROTEIN

PETRIKURSULA

Department of Pathology

OULU 2000

OULUN YLIOPISTO, OULU 2000

CYTOPLASMIC DOMAINS OF THE MYELIN-ASSOCIATED GLYCOPROTEIN

PETRI KURSULA

Academic Dissertation to be presented with the assent of the Faculty of Medicine, University of Oulu, for public discussion in the Auditorium of the Department of Pharmacology and Toxicology, on June 21st, 2000, at 12 noon.

Copyright © 2000Oulu University Library, 2000

OULU UNIVERSITY LIBRARYOULU 2000

ALSO AVAILABLE IN PRINTED FORMAT

Manuscript received 12 May 2000Accepted 23 May 2000

Communicated by Docent Matti Poutanen Professor Heikki Rauvala

ISBN 951-42-5669-7

ISBN 951-42-5668-9ISSN 0355-3221 (URL: http://herkules.oulu.fi/issn03553221/)

For Inari, Ville, Lotta, and Jussi

Imagination is more important than knowledge.Albert Einstein

Kursula, Petri, Cytoplasmic domains of the myelin-associated glycoproteinDepartment of Pathology, University of Oulu, FIN-90220 Oulu, Finland2000Oulu, Finland(Manuscript received 12 May 2000)

Abstract

The function of the vertebrate nervous system is based on the rapid and accurate transmission of electrical impulses. The myelin sheath is a lipid-rich membrane that envelops the axon, preventing the leakage of the nervous impulse to the environment. Myelin is formed when the plasma membrane of a myelinating glial cell differentiates and wraps around an axon. The compaction of myelin leads to the extrusion of most of the glial cell cytoplasm from the structure. Both the compact and noncompact regions of myelin carry distinct subsets of proteins.

The myelin-associated glycoprotein (MAG) is present in noncompact myelin. It is a cell adhesion molecule expressed only by myelinating glial cells. Two isoforms of MAG, S- and L-MAG, exist, and these forms differ from each other only by their cytoplasmic domains. Until now, little information has been available on the differences between the MAG isoforms. This study was carried out in order to gain information on the cytoplasmic domains of S- and L-MAG.

Significant differences were observed in the properties of the MAG cytoplasmic domains. An interaction between the L-MAG cytoplasmic domain and the S100β protein was characterised, and a role for this interaction was found in the regulation of L-MAG phosphorylation. Evidence was also obtained for the dimerisation of the L-MAG cytoplasmic domain. The S-MAG cytoplasmic domain bound zinc, which induced a change in the surface properties of the protein. The S-MAG cytoplasmic domain was also found to interact directly with tubulin, the core component of microtubules.

In conclusion, this study has brought information on the functions of the MAG cytoplasmic domains. The results are complementary with ealier hypotheses on the roles of the MAG isoforms in myelinating glia. While the properties of L-MAG suggest a role as a signaling molecule, a dynamic structural role for S-MAG during myelin formation and maintenance can be envisaged.

Keywords: nervous system, Schwann cell, protein interaction, cytoskeleton

Acknowlegdements

This work was carried out at the Department of Pathology, University of Oulu, during the years 1995-1999.

I wish to express my gratitude to Docent Anthony Heape for guidance, criticism, language corrections, friendship, and support during the good and bad times of this study. His faith in the project has always been strong, even at times when I have been ready to give up. Professor Veli-Pekka Lehto has provided many useful suggestions, comments, and thoughts, and Professor Frej Stenbäck has often helped in practical matters.

I would like to thank Professor Heikki Rauvala and Docent Matti Poutanen for their critical review of this manuscript and helpful suggestions.

I owe my thanks to Professor Karl Tryggvason for originally introducing me to the world of science. I thank all my co-authors for providing crucial materials for this study. The entire staff of the Pathology Department is acknowledged for their role in creating and maintaining unique working conditions. I am also grateful to the structural biologists at the Deparment of Biochemistry, especially Professors Kalervo Hiltunen and Rik Wierenga and Dr. Daan van Aalten, for sharing their expertise on protein structural analysis and helping in my efforts to start learning protein crystallography.

It is a shame how little time one spends with friends when involved in a research project. I wish to thank all my friends that have shared their time with me both before and during this study. Special thanks are directed to the nine godparents of our children.

This study would have never been finished without my family. My parents have always allowed me to develop my own independent critical way of thinking, and my parents-in-law are acknowledged for all possible kinds of support. Most of all I would like to thank my wife Inari for her love, encouragement, and patience and, during the late stages of this study, collaboration and guidance. Our three children, Ville, Lotta, and Jussi, are given a kiss and a hug for their open-mindedness and everyday life.

This work has been supported by grants from the Maud Kuistila Memorial Foundation, the Emil Aaltonen Foundation, the Finnish Multiple Sclerosis Foundation, and the Federation of European Biochemical Societies. I am especially grateful to the friendly people of the Finnish MS Foundation.

Oulu, May 2000 Petri Kursula

Abbreviations

3D three-dimensionalATP adenosine 5'-triphosphatebp base pairBSA bovine serum albumincAMP cyclic 3',5'-adenosine monophosphateCAM cell adhesion moleculeCD circular dichroismcDNA complementary DNACK2 casein kinase 2CMT Charcot-Marie-Tooth diseaseCNPase 2',3'-cyclic nucleotide 3'-phosphodiesteraseCNS central nervous systemCx32 connexin32DIG detergent-insoluble glycosphingolipid-enriched complexdMAG degradation product of MAG, containing the extracellular domainDMEM Dulbecco's modified Eagle's mediumDNA deoxyribonucleic acidDRG dorsal root ganglionDSS disuccinimidyl suberateE. coli Escherichia coliEDTA ethylene diamine tetraacetic acidGalC galactocerebrosideGAPDH glyceraldehyde-3-phosphate dehydrogenaseGDP guanosine 5'-diphosphateGPI glycosylphosphatidylinositolGST glutathione S-transferaseGTP guanosine 5'-triphosphateHBS Hepes-buffered salineHepes N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid)Ig immunoglobulinkDa kilodaltonL-MAG large myelin-associated glycoprotein

mag the gene encoding MAGMAG myelin-associated glycoproteinMAG -/- MAG-deficientMAGct MAG cytoplasmic domainMAP microtubule-associated proteinMBP myelin basic proteinMOBP myelin oligodendrocytic basic proteinMOG myelin/oligodendrocyte glycoproteinmRNA messenger RNAMS multiple sclerosisN-CAM neural cell adhesion moleculeOMGP oligodendrocyte-myelin glycoproteinP0 myelin protein zeroP2 myelin protein 2PBS phosphate-buffered salinePKA protein kinase APKC protein kinase CPLCγ phospholipase CγPLP proteolipid proteinPMP-22 peripheral myelin protein 22PNS peripheral nervous systemqv quiveringRFLP restriction fragment length polymorphismRGD arginine-glycine-aspartic acidRNA ribonucleic acidRT-PCR reverse transcription polymerase chain reactionS-MAG small myelin-associated glycoproteinSDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresisSMP Schwann cell myelin proteinTBS Tris-buffered salineTris tris(hydroxymethyl)aminomethaneV variable

List of original articles

This thesis is based on the following articles, referred to in the text by their Roman numerals. In addition, unpublished results are presented.

I Kursula P, Lehto V-P, Garbay B, Cassagne C & Heape AM (1998) Expression of the amino acid dimorphism in the small myelin-associated glycoprotein cytoplasmic domain in rat sciatic nerves during postnatal development (Published erratum appears in Mol Brain Res 62: 110). Mol Brain Res 54: 252-261.

II Heape AM, Lehto V-P & Kursula P (1999) The expression of recombinant large myelin-associated glycoprotein cytoplasmic domain and the purification of native myelin-associated glycoprotein from rat brain and peripheral nerve. Protein Expr Purif 15: 349-361.

III Kursula P, Tikkanen G, Lehto V-P, Nishikimi M & Heape AM (1999) Calcium-dependent interaction between the large myelin-associated glycoprotein and S100β. J Neurochem 73: 1724-1732.

IV Kursula P, Meriläinen G, Lehto V-P & Heape AM (1999) The small myelin-associated glycoprotein is a zinc-binding protein. J Neurochem 73: 2110-2118.

V Kursula P, Tanner S, Quarles RH, Lehto V-P & Heape AM (2000) The small myelin-associated glycoprotein is a microtubule-binding protein: A link between the axon and the myelinating glial cell cytoskeleton? (Submitted for publication)

VI Kursula P, Lehto V-P & Heape AM (2000) S100β inhibits the phosphorylation of the L-MAG cytoplasmic domain by PKA. Mol Brain Res 76: 407-410.

VII Kursula P, Lehto V-P & Heape AM (2000) Estimation of total ribonucleic acid quantity from dilute samples by non-denaturing electrophoresis and silver staining. Electrophoresis 21: 545-547.

Contents

AbstractAcknowledgementsAbbreviationsList of original articles1. Introduction...................................................................................................................172. Review of the literature................................................................................................ 18

2.1. The myelin sheath.................................................................................................182.1.1. Mature peripheral nerve myelin....................................................................182.1.2. Myelinogenesis in the peripheral nervous system........................................ 20

2.1.2.1. The Schwann cell.................................................................................. 202.1.2.2. Axoglial contact and the initiation of myelinogenesis.......................... 212.1.2.3. Envelopment and spiralling...................................................................232.1.2.4. Compaction........................................................................................... 23

2.1.3. Some differences between central and peripheral nervous system myelin...242.2. Biochemistry of the peripheral nerve myelin sheath............................................ 25

2.2.1. Proteins of compact myelin...........................................................................252.2.1.1. Protein zero........................................................................................... 252.2.1.2. Myelin basic protein..............................................................................272.2.1.3. Peripheral myelin protein-22.................................................................272.2.1.4. Myelin protein 2.................................................................................... 28

2.2.2. Proteins of noncompact myelin.....................................................................282.2.3. Protein kinases implicated in myelination.................................................... 302.2.4. The myelinating glial cell cytoskeleton and myelination............................. 312.2.5. Lipid composition of the myelin membrane................................................. 322.2.6. Inorganic components of myelin...................................................................33

2.3. The myelin-associated glycoprotein..................................................................... 352.3.1. The mag gene................................................................................................ 352.3.2. The expression and subcellular localisation of MAG...................................362.3.3. The MAG protein..........................................................................................37

2.3.3.1. The extracellular domain.......................................................................372.3.3.2. The cytoplasmic domain....................................................................... 38

2.3.4. Posttranslational modifications..................................................................... 39

2.3.4.1. Phosphorylation.....................................................................................392.3.4.2. Glycosylation........................................................................................ 402.3.4.3. Acylation............................................................................................... 402.3.4.4. Proteolysis............................................................................................. 41

2.3.5. The proposed functions of MAG.................................................................. 412.3.5.1. Initiation of myelination........................................................................412.3.5.2. Membrane spacing and intercellular adhesion...................................... 422.3.5.3. Membrane motility................................................................................422.3.5.4. Endocytosis........................................................................................... 432.3.5.5. Signal transduction................................................................................432.3.5.6. Effects on axonal properties..................................................................44

2.2.6. Pathological conditions related to MAG.......................................................442.2.6.1. Animal mutants..................................................................................... 442.2.6.2. Human peripheral neuropathies............................................................ 452.2.6.3. Multiple sclerosis.................................................................................. 462.2.6.4. Gliomas................................................................................................. 462.2.6.5. The MAG knockout mice......................................................................47

3. Aims of the present study............................................................................................. 494. Materials and methods..................................................................................................50

4.1. Nucleic acid methods............................................................................................504.1.1. Isolation and analysis of RNA...................................................................... 504.1.2. Cloning of cDNA.......................................................................................... 51

4.2. Protein analysis methods.......................................................................................514.2.1. Proteins, antibodies, and peptides................................................................. 514.2.2. Protein interactions and function.................................................................. 52

4.2.2.1. Zinc binding assays............................................................................... 524.2.2.2. Ligand blotting and immunoblotting.....................................................534.2.2.3. Affinity chromatography and covalent crosslinking.............................534.2.2.4. Polymerisation of microtubules in vitro................................................544.2.2.5. L-MAG dimerisation assays..................................................................544.2.2.6. Protein phosphorylation in vitro............................................................55

4.2.3. Structural analyses........................................................................................ 564.2.3.1. Primary sequence analyses....................................................................564.2.3.2. Circular dichroism spectroscopy...........................................................564.2.3.3. X-ray crystallography............................................................................57

4.3. Cell culture and immunocytochemistry................................................................575. Results ....................................................................................................................... 59

5.1. Analyses of primary structure...............................................................................595.2. Production of recombinant proteins and antibodies..............................................615.3. The amino acid dimorphism of the rat S-MAG.................................................... 615.4. Characterisation of S-MAG as a zinc-binding protein......................................... 625.5. S-MAG is a microtubule-associated protein.........................................................63

5.5.1. S-MAG binds directly to tubulin.................................................................. 635.5.2. Cell culture studies........................................................................................63

5.6. Crystallisation of the S-MAG cytoplasmic domain..............................................655.7. The interaction between S100β and L-MAG........................................................67

5.8. Phosphorylation of recombinant MAG in vitro.................................................... 685.9. Dimerisation of the L-MAG cytoplasmic domain................................................ 69

6. Discussion.....................................................................................................................716.1. The cytoplasmic domain of S-MAG.....................................................................71

6.1.1. The dimorphism of amino acid 579.............................................................. 726.1.2. The binding of zinc....................................................................................... 726.1.3. Association with tubulin............................................................................... 73

6.2. The cytoplasmic domain of L-MAG.....................................................................756.3. Concluding remarks..............................................................................................78

7. References.....................................................................................................................80Original articles

1. Introduction

Although a general property of all cells, excitability and its transmission are especially developed in nerve and muscle cells. The nervous system is a complex tissue, allowing the coordinated function of all organs of the animal body by the transmission of excitatory and inhibitory signals.

In higher animals, the nervous system is divided into two main parts, the central nervous system (CNS), consisting of the brain and spinal cord, and the peripheral nervous system (PNS). Nerve fibers within both CNS and PNS may be either myelinated or unmyelinated. While unmyelinated nerve fibers are associated with somatic and visceral functions, myelinated nerve fibers assure the rapid, focused signal transmission in cases such as the voluntary and reflex stimulation of locomotory muscles and the perception of external stimuli. The rapid impulse conduction by myelinated nerve fibers is made possible by the presence of the myelin sheath, a lipid-rich multilamellar membrane enveloping the axons of nerve cells. While the general structure and function of myelin is the same in both the CNS and PNS, important morphological and biochemical differences between CNS and PNS myelin exist.

The correct formation and maintenance of myelin are key prerequisites for the normal functioning of the vertebrate nervous system. Myelin is formed by the differentiation of the plasma membrane of specialised glial cells - oligodendrocytes in the CNS and Schwann cells in the PNS - which wrap around the axon. Different structures within the myelin sheath carry certain subsets of proteins, many of which are only expressed by myelinating glial cells. The myelin-associated glycoprotein (MAG) is a transmembrane cell adhesion molecule (CAM) present in both CNS and PNS myelin.

Despite extensive studies during the last 30 years, the functions of MAG are still largely enigmatic. This study was carried out in order to shed light on the functions of MAG, and on the mechanisms by which it could fulfill these functions during myelin formation and maintenance.

2. Review of the literature

2.1. The myelin sheath

The myelin sheath, first described almost 150 years ago (Virchow 1854), is a multilamellar membrane structure that envelops large diameter axons. It has the same functions in both the central and peripheral nervous systems, acting as an insulator around the axon, guiding the localisation of axonal ion channels (Vabnick et al. 1996, Kaplan et al. 1997), and providing mechanical support. The confinement of ion channels to specific regularly spaced regions of the axon (Salzer 1997) is the key prerequisite for the very fast saltatory conduction (Huxley & Stämpfli 1949, Ritchie 1984) of nervous impulses. When damaged by disease or lesions, or when the sheath fails to form normally due to defects in the genetic program, serious neurological conditions, including motor and sensory deficits, may ensue. In addition to vertebrates, myelin-like sheaths have recently been found in some species of tiny marine crustaceans called copepods (Davis et al. 1999).

2.1.1. Mature peripheral nerve myelin

Although there is a continuous membrane surface from the perikaryon of the myelin-forming cell to compact myelin, there are many specialisations within the membrane which exhibit variability in terms of structure, function, and biochemical composition (Fig. 1). In addition to the compact myelin, the specialised domains include the surface membranes of the cell body and processes, surface and periaxonal membranes of the myelin sheath, paranodal loops, and incisures (reviewed by Morell et al. 1994, Quarles et al. 1997). It is very important to distinguish the constituents of compact myelin from those of the other myelin-related membranes, also termed non-compact myelin.

The multilamellar structure of compact myelin is characterised by the small volume that is occupied by non-membranous components. The cytoplasmic leaflets of apposing membranes are practically fused, forming the so-called major dense line seen in electron micrographs. The less dense double band at the extracellular apposition is called the

intraperiod line. The extracellular space between apposing membranes is only 2 nm (Peters et al. 1991). The periodic structure of compact myelin has enabled scientists to determine its properties using X-ray diffraction techniques (Schmitt et al. 1941). From these studies, it has been concluded that each layer of compact myelin is approximately 15-18 nm thick. The calculated spacing represents the thickness of two membranes per repeat, and while it is unique for CNS and PNS myelin within a given species, it differs somewhat across evolutionary lines (Kirschner & Blaurock 1992).

Important features of the membranes of so-called noncompact myelin is that they face another membrane at an extracellular distance of 12-14 nm, and that their intracellular side is in contact with cytoplasm. Of these membranes, the periaxonal Schwann cell membrane, the innermost membrane of myelin, makes direct contact with the axon. Thus, it is very likely that molecules present in this membrane are largely responsible for the generation and maintenance of axoglial contact, and for signal transduction between the two cells.

Another important structure only found in the PNS myelin sheath is the Schmidt-Lanterman incisure (Ghabriel & Allt 1981), a cytoplasm-containing channel connecting the adaxonal (innermost) and abaxonal (outermost) cytoplasmic compartments of the sheath. Although an important role has been suggested for these structures for years, it is not until recently that the mechanism of their function has been partially unravelled. The incisures contain gap junctions (Sandri et al. 1982, Scherer 1996), carrying the gap junction protein connexin 32 (Cx32) (Bergoffen et al. 1993, Scherer et al. 1995a). While such junctions are usually described between adjacent cells in various epithelia, in the PNS myelin sheath they link apposed membranes of the same cell (Scherer 1996, Scherer et al. 1997). If these gap junctions formed a radial pathway directly across the myelin sheath, this would provide a significant short-cut for ions and small molecules: for the

19

Ax

R

PnCoSL

Ex In

Sc

Fig. 1. Schematic representation of a peripheral nerve myelin sheath. Compact myelin (Co) has formed around an axon (Ax). The cytoplasm of a myelinating glial cell is confined to narrow canals that run spirally through the sheath (SL, Schmidt-Lanterman incisures), to the internal (In) and external (Ex) mesaxons of myelin, to the lateral edges of each myelin segment (Pn, paranodal loops) and to the Schwann cell body (Sc). The region of each myelin segment located between two nodes of Ranvier (R) is called an internode.

largest myelin sheaths, this radial pathway is approximately 1000 times shorter than the potential circumferential pathway through the cytoplasmic channels of the myelin sheath itself (Scherer et al. 1995a, 1997). Indeed, such a radial pathway for small molecules is apparently present in the myelin sheath (Balice-Gordon et al. 1998), and it is most likely formed by gap junctions containing Cx32 (Scherer et al. 1997). The rapid passive diffusion of small molecules directly through the sheath could be of major importance in, for example, cytoplasmic signaling events related to myelin formation and maintenance.

The myelin sheath is interrupted at regular intervals along its length. These interruptions are the nodes of Ranvier (Ranvier 1878), and each myelin segment located between two nodes is called an internode (Johnston & Peters 1973). The myelin sheath has an important role in guiding the localisation of ion channels within the axonal plasma membrane at the nodes of Ranvier (Waxman & Ritchie 1993). Each myelinated fibre in the PNS is surrounded along its entire length by a continuous basal lamina secreted by the myelinating Schwann cells (Bunge et al. 1986). At the nodes of Ranvier, compact myelin ends, and the Schwann cell plasma membrane forms cytoplasm-containing membrane whorls, the paranodal loops, that are involved in the generation of tight adhesion zones, the paranodal junctions, between the axon and the glial cell (Bargmann & Lindner 1964, Andres 1965). The molecular mechanisms by which these junctions are formed have not been fully elucidated, but they seem to involve the axolemmal protein Caspr and unidentified glial counterparts (reviewed by Arroyo & Scherer 2000).

2.1.2. Myelinogenesis in the peripheral nervous system

The formation of myelin involves the morphological and biochemical differentiation of the plasma membrane of myelinating glial cells - oligodendrocytes in the CNS and Schwann cells in the PNS. While many aspects of myelin formation are similar in the CNS and PNS, there are also numerous differences which make it impractical to treat both systems here. Consequently, for the sake of clarity, this chapter will focus on myelinogenesis in the PNS.

2.1.2.1. The Schwann cell

Classically, glial cells have been viewed as providers of physical and trophic support for neurons (Virchow 1846, Kandel 1991). Schwann cells are the principal glial cells of the vertebrate PNS (Schwann 1839, 1847), and they are best known for their ability to elaborate myelin (Geren 1954). In the PNS, every axon is enveloped along its entire length by a sheath of Schwann cells which, depending on the function of the axon, may or may not produce a myelin sheath.

Embryologically, all Schwann cells are derived from the neural crest, and they descend from a common precursor (Zorick & Lemke 1996, Jessen et al. 1997). The regulation of their final cell fate, i.e. myelinating or non-myelinating, is ultimately determined by the axons with which they find themselves in contact (Bunge et al. 1990).

20

Furthermore, these two phenotypes are apparently reversible and interconvertible (Bray et al. 1981, Bunge et al. 1990, Jessen & Mirsky 1992). Under certain conditions, myelinating Schwann cells will readily abandon their complex structure and distinctive molecular make-up and revert to a phenotype similar to that of immature Schwann cells prior to myelination. Typically, such regression is triggered by axonal degeneration in vivo or by removing Schwann cells from the influence of axons and placing them in serum-containing media in culture. If such cells are placed in contact with appropriate axons they again differentiate and form myelin or, if exposed to axons that are normally unmyelinated in vivo, they adopt the phenotype of mature nonmyelinating Schwann cells (Jessen et al. 1997).

Schwann cells may loosely ensheath, without myelinating, multiple slow-conducting axons in nerves (Zorick & Lemke 1996). The separation between axons is achieved by each one being invaginated at a different site on the Schwann cell surface, so that each one lies in its own groove. Often, the axons are completely enfolded so that the outer lips of the groove come together to form a mesaxon (Johnston & Peters 1973). Thus, in the case of non-myelinated axons, each Schwann cell can contribute to the ensheathment of several fibers, while in myelinated fibers, each Schwann cell produces a single myelin segment around a single axon.

Aside from morphological differences, myelinating and nonmyelinating Schwann cells are readily distinguished from one another by their gene expression profiles. While myelinating Schwann cells express high levels of myelin-specific genes, the nonmyelinating ones instead express a set of nonmyelinating markers that are also expressed by immature Schwann cells. Some markers are common to both types of mature Schwann cells, including the lipid galactocerebroside (GalC) and the S100 protein (Zorick & Lemke 1996). S100 is a family of calcium-binding proteins, and the family member expressed at highest levels by glial cells is S100β (Zimmer & Van Eldik 1988, Mata et al. 1990).

2.1.2.2. Axoglial contact and the initiation of myelinogenesis

In the immature developing nerve, Schwann cell progenitors migrate from the neural crest and proliferate along pre-existing axonal tracts during embryonic development, loosely investing bundles of axons with cytoplasmic extensions. The establishment of axonal contact triggers Schwann cell proliferation (Wood & Bunge 1975, Salzer & Bunge 1980). Eventually, a one-to-one relationship is established between the Schwann cell and a segment of an axon to be myelinated. Simultaneously, the Schwann cell elongates laterally along the axon and secretes a basal lamina at the outer surface of the axon-Schwann cell complex (Bunge et al. 1986, Eldridge et al. 1987, 1989, Kidd & Heath 1988, Owens & Bunge 1989). The ability of Schwann cells to make a basal lamina correlates with their ability to proceed to myelination (Bunge et al. 1986), and the deposition of the basal lamina is one of the factors driving Schwann cell differentiation towards a myelinating phenotype (Carey et al. 1986, Eldridge et al. 1989).

Each Schwann cell has the potential to form a myelin sheath, but will only do so after it has contacted a suitable axon (Weinberg & Spencer 1976, Aguayo et al. 1976, Bray et

21

al. 1981). Myelination is regulated precisely according to the diameter and conduction properties of the axon (Duncan 1934, Friede 1972), and it can occur only if the axonal diameter is greater than 0.7 µm (Webster 1971, Windebank et al. 1985). The presence of axons is required for the expression of myelin genes during development (Wood & Engel 1976, Sternberger et al. 1979, Uyemura et al. 1979, Carson et al. 1983, Frail & Braun 1984) and the maintenance of a myelinating phenotype. The loss of axonal contact leads to the downregulation of myelination-related gene expression (Brockes et al. 1980, Mirsky et al. 1980, Poduslo 1984, LeBlanc et al. 1987, Gupta et al. 1988, 1990, Trapp et al. 1988, Brunden & Brown 1990, LeBlanc & Poduslo 1990).

Before myelination can start, contact must be established between the Schwann cell and an axon (Wood & Bunge 1975, Salzer & Bunge 1980). Axonal signals are mandatory at all stages of Schwann cell precursor development into myelin-forming cells. For example, it has been shown that proliferation, survival, and differentiation of Schwann cell precursors does not occur in the absence of neurons (Jessen & Mirsky 1991, Dong et al. 1995). It is also well established that the signal for nerve ensheathment originates in axons with which Schwann cells interact (Aguayo et al. 1976, Owens & Bunge 1989, Scherer et al. 1994).

Despite extensive studies, the nature of the axonal signals that initiate the myelination process and regulate the rate and extent of myelin formation are still unknown (reviewed by Mirsky & Jessen 1996, Garbay et al. 2000). While the axon-Schwann cell contact is required for the initiation of myelin synthesis (Wood & Bunge 1975, Salzer & Bunge 1980), diffusible substances are also implicated in the axon-derived myelination signal (Bolin & Shooter 1993). A role for locally produced progesterone in myelination has been suggested (Koenig et al. 1995, Melcangi et al. 2000). Since continuous axonal contact is required, it is most likely that the signal is at least partially mediated by, at present unidentified, cell surface receptors on both the axonal and Schwann cell membranes. One molecule suggested to play a role in axonal recognition is MAG, a cell adhesion molecule expressed only by myelinating glial cells (Owens & Bunge 1989, 1991, Owens et al. 1990, discussed in detail in chapter 2.3). Another cell adhesion molecule of the immunoglobulin (Ig) superfamily expressed by neurons, L1, has also been sugggested to have a critical role in the initiation phase of myelination (Seilheimer et al. 1989, Wood et al. 1990), but myelination does occur in the nervous system of L1-deficient mice (Dahme et al. 1997).

Schwann cell myelination has been shown to be regulated by specific patterns of neural impulses (Stevens et al. 1998). There are major differences in axonal firing patterns during the myelinating and premyelinating periods of dorsal root ganglion (DRG) development (Fields 1998). Although Schwann cells are closely associated with DRG axons when they are firing at low frequency (<0.5 Hz), myelination does not begin until the prenatal period, when the firing pattern increases to 1-10 Hz (Fitzgerald & Fulton 1992). Stevens and co-workers (1998) found that stimulation of DRG neurons at a frequency of 0.1 Hz, but not at a high frequency of 1 Hz, resulted in the inhibition of myelination of DRG axons by Schwann cells. Furthermore, stimulation at 0.1 Hz lowered the expression of L1 by the DRG neurons, indicating that the low-frequency activity seen in DRG neurons before the onset of myelination may be an inhibitory factor to myelin formation via its effect on L1 expression at the axonal surface.

When the Schwann cell has received the myelination signal from the axon, its gene

22

expression profile is modified to allow the large-scale synthesis of myelin components (Owens & Bunge 1989, Zorick & Lemke 1996, Garbay et al. 2000). This regulation at the gene expression level involves transcription factors, of which Pax-3 (Kioussi et al. 1995), Oct-6 (Monuki et al. 1989, 1993), and Krox-20 (Swiatek & Gridley 1993, Topilko et al. 1994) have been shown to participate in the development of a myelinating Schwann cell phenotype.

2.1.2.3. Envelopment and spiralling

In sciatic nerves of rats, axonal ensheathment starts at birth, and rapid myelination continues for another two weeks (Webster 1971). Once contact has been made between the Schwann cell and the axon, the Schwann cell plasma membrane starts to wrap itself spirally around the axon and upregulates the synthesis of myelin components (Owens & Bunge 1989).

The mesaxon is defined by the non-compacted layers of Schwann cell membrane around the axon. Thus, before compaction, there is only one mesaxon, and the internal and external mesaxons, separated by compacted myelin lamellae, are only defined after compaction has started. When looking at a single Schwann cell during the formation of the first membrane spiral turn around the axon, major level-to-level differences are observed in the size, contour, and direction of the mesaxon. These differences appear to be determined by the shape of the Schwann cell nucleus and variations in the distribution of organelles. As the number of turns increases, the cytoplasm is gradually confined to a series of longitudinal strips near the external mesaxon (Webster 1971).

An important aspect of the spiraling of the Schwann cell plasma membrane around the axon is that myelin growth occurs by the rotation of the internal mesaxon (Bunge et al. 1989), as opposed to the long-held belief that new myelin layers are added at the outermost limits of the existing sheath by the rotation of the Schwann cell body around the axon. It is obvious that in such a mechanism, direct contacts must be established temporarily between the innermost layer of myelin and the axonal membrane. Such interactions would have to involve cell surface adhesion molecules present in the glial periaxonal membrane. Again, MAG has been suggested to play a dynamic role in interacting with the axon during spiral growth. Such a function for MAG is indeed supported by recent evidence from mice deficient in both myelin protein zero (P0) and MAG (Carenini et al. 1999, see chapter 2.2.6.5).

2.1.2.4. Compaction

When several layers of loosely wrapped Schwann cell plasma membrane have enveloped the axon, a large portion of the structure compacts. While the cytoplasmic leaflets of apposing membranes come close to each other and practically fuse, cytoplasm is excluded from the majority of compact myelin, as are numerous proteins, including MAG, characteristic of the noncompacted regions of myelin (reviewed by Quarles et al.

23

1997). In fact, only a few proteins have unequivocally been characterised as components of compact PNS myelin (see chapter 2.2.1). Of these proteins, the two most abundant, P0 and myelin basic protein (MBP), are believed to be to a large extent responsible for the stabilisation of compact myelin (reviewed by Quarles et al. 1997). While MBP, with its strong positive charge, has been suggested to play a role in linking two apposing cytoplasmic faces of myelin membranes (Moscarello 1997), P0 seems to be crucial for keeping the extracellular faces of two compact myelin membranes at a distance of 2 nm from one another (Shapiro et al. 1996, Suter 1997).

During compaction, the extracellular intermembrane space of 12-14 nm characteristic of noncompact myelin decreases to the 2 nm characteristic of compact myelin. As a result, layers of lipid and protein have been laid down in a series on concentric layers or lamellae, each of which has a thickness of about 18 nm (Johnston & Peters 1973). The spiralling of the inner lip of the Schwann cell membrane around the axon continues, and the greatest increase in myelin membrane area occurs while the sheath is a compact lamellar spiral (Webster 1971). In fact, the area of myelin membrane increases exponentially with time, showing a 6500-fold increase in the rat sciatic nerve between the ages of one day and 2.5 months (Webster 1971). In sciatic nerves of adult rats, the thickest myelin sheaths contain approximately 100 lamellae, and very few fibers have less than 10 lamellae (Friede & Samorajski 1968).

As a result of compaction, in Schwann cells with compact sheaths consisting of more than 6 turns, almost all of the cytoplasm is located along the external mesaxon and outer layer of the myelin spiral. In addition, cytoplasm can be found at Schmidt-Lanterman incisures, at the nodal ends of the sheaths, and in the small periaxonal collar located along the innermost layer of compact myelin. (Webster 1971)

2.1.3. Some differences between central and peripheral nervous system myelin

While there is a strong overall similarity in the mechanism of myelination and the structure of the myelin sheath in the CNS and PNS, important differences also exist. While a Schwann cell only creates a single myelin segment around a single axon (Geren 1954, Webster 1971), an oligodendrocyte can make numerous myelin sheaths around several axons (Peters & Proskaur 1969, Peters et al. 1970, Johnston & Peters 1973). For this reason, the oligodendrocyte nucleus is not found in intimate association with the outside of the sheath, as is the Schwann cell nucleus. Instead, it is some distance away, the myelin sheath being formed at the end of an oligodendrocyte process (reviewed by Pfeiffer et al. 1993). In addition, the outer cytoplasmic collar of myelin is incomplete in the CNS and, at the node of Ranvier, the axon is bare, instead of being covered by processes of the myelin-forming cell, as in the PNS (Peters et al. 1970, Johnston & Peters 1973).

While Schmidt-Lanterman incisures are not present in CNS myelin, compact CNS myelin contains a network of interlamellar junctions called the radial component, which extends from the outermost to the innermost layer of membrane (Kosaras & Kirschner 1990). The radial component has been isolated and shown to be enriched in proteins that

24

are also found in the cytoskeletal veins of cultured oligodendrocyte membrane sheets, i.e. tubulin, actin, 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNPase), and certain isoforms of MBP (Karthisagan et al. 1994). It has been suggested that the radial component of myelin serves as a cytoskeletal scaffolding structure (Dyer 1997).

Important differences also exist between PNS and CNS myelin at the biochemical level. In compact CNS myelin, the quantitatively major protein is the proteolipid protein (PLP) (Eng et al. 1968), while P0 and peripheral myelin protein 22 (PMP22) are found only in PNS compact myelin. In noncompacted regions of the sheath, the CNS myelin-specific proteins include the putative cell adhesion molecules myelin/oligodendrocyte glycoprotein (MOG; Linington et al. 1984, Gardinier et al. 1997) and the oligodendrocyte-myelin glycoprotein (OMGP; Mikol & Stefansson 1988).

2.2. Biochemistry of the peripheral nerve myelin sheath

Myelin is composed mainly of lipids with a quantitatively minor, but functionally important contribution of proteins. On average, myelin consists of 80 % lipid and 20 % protein (reviewed by Garbay et al. 2000). Since glycoproteins are important constituents of plasma membranes in general (Winzler 1970), it is not surprising that many of the proteins in both compact and non-compact myelin are glycosylated. Many of the proteins present in myelin have not been found in other tissues or cell types, and the function of these myelin-specific proteins in myelin has been studied most extensively. However, despite the relatively low number of myelin-specific proteins, and the vast number of different research approaches used, surprisingly little is known about the molecular mechanisms that drive myelin formation and assure its maintenence. The expression of the myelin constituents is under tight regulation during development, and in fact, the expression profiles of the different myelin proteins resemble one another (reviewed by Garbay et al. 2000), directly reflecting the accumulation of myelin membrane in the nerve. The consituents of myelin that have been suggested to play important structural or functional roles in the PNS myelin sheath, and those that are of special interest to this study, are briefly introduced in the following paragraphs.

2.2.1. Proteins of compact myelin

Schematic representations of the proteins present in compact myelin are shown in figure 2. These proteins include both integral and extrinsic membrane proteins.

2.2.1.1. Protein zero

P0 is the most abundant protein of PNS myelin, accounting for over 50 % of the total myelin protein (Everly et al. 1973, Kitamura et al. 1976). It is a small transmembrane

25

glycoprotein of approximately 30 kilodalton (kDa) (Everly et al. 1973, Kitamura et al. 1976), and it is believed to be adhesive via both its extracellular and cytoplasmic domains (Kirschner & Ganser 1980, Lemke & Axel 1985, D'Urso et al. 1990, Filbin & Tennekoon 1992). The extracellular domain of P0 contains a single Ig-like domain and carries the HNK-1 carbohydrate epitope (Bollensen & Schachner 1987, Voshol et al. 1996), while the cytoplasmic domain carries a basic charge and is believed to hold the cytoplasmic leaflets of myelin together by electrostatic interactions with membrane lipids (Lemke & Axel 1985). There is also evidence for interactions between P0 and the cytoskeleton (Wong & Filbin 1994, 1996, Filbin et al. 1997), but as there are no cytoskeletal elements in compact myelin, where P0 is localised, these interactions can only take place before myelin compaction has occurred.

Studies on the three-dimensional (3D) structure of the P0 extracellular domain have indicated that this domain forms a homotetramer (Shapiro et al. 1996, Inouye et al. 1999, Sedzik et al. 1999). Direct interactions between tetramers from apposing myelin membranes have been suggested to play an important role in the formation and maintenance of the compact myelin structure (Shapiro et al. 1996, Suter 1997). Recently, an interaction between P0 and the peripheral myelin protein 22 (PMP-22) was demonstrated (D'Urso et al. 1999), but the significance of this finding remains to be assessed.

The importance of P0 for normal structure of compact myelin is highlighted by studies on P0-deficient mice (Giese et al. 1992). In the mutants, Schwann cells are able to establish a 1:1 relationship with axons, but subsequent steps of myelination are impaired. Some myelin is made, but it is unusually thin and poorly compacted. The results demonstrate that the expression of P0 is essential for the normal spiraling, compaction,

26

PLP

+-

P0

-+

PMP-22

-+

MBP

++

P2

++

CNSPNS

extracellular

plasmamembrane

cytoplasm

Fig. 2. Proteins of compact myelin (modified after Quarles et al. 1997). The drawings show the apparent relationships of each protein to the lipid bilayer of the myelin membrane. Oligosaccharides are represented by triangles. The presence of these proteins in the compact myelin of the CNS and PNS is indicated under the drawing. PLP, proteolipid protein; P0, protein zero; PMP-22, peripheral myelin protein 22; MBP, myelin basic protein; P2, protein 2.

and maintenance of PNS myelin. In humans, several point mutations in the gene encoding P0 have been identified and

classified as Charcot-Marie-Tooth disease (CMT) (Charcot & Marie 1886, Tooth 1886) type 1B (Su et al. 1993, reviewed by Chance & Lupski 1994, Suter & Snipes 1995). In addition, rare P0 mutations have been found in patients diagnosed with Dejerine-Sottas syndrome (Fabrizi 1999, reviewed by Chance & Lupski 1994, Roa et al. 1996).

2.2.1.2. Myelin basic protein

MBP is a highly basic extrinsic membrane protein present at the cytoplasmic surfaces of compact myelin membranes (Korngurth & Anderson 1965, Herndon et al. 1973, Guarnieri et al. 1974, Sternberger et al. 1978, Omlin et al. 1982). MBP represents 5-15 % of PNS myelin protein (Benjamins & Morell 1978). Its strong positive charge suggests that it may contribute to the compaction of myelin, together with P0, via electrostatic interactions with the cell membrane (Moscarello 1997). Several isoforms of MBP exist that are produced by alternative messenger ribonucleic acid (mRNA) splicing and posttranslational modifications (reviewed by Moscarello 1997). In the PNS, four alternatively spliced forms of MBP are expressed in substantial amounts (Barbarese et al. 1977, Greenfield et al. 1982).

MBP binds zinc (Riccio et al. 1995, Tsang et al. 1997), and zinc has been shown to prevent its dissociation from isolated myelin membranes (Earl et al. 1988). A 3D structure has recently (Beniac et al. 1997, Risdale et al. 1997) been determined for MBP by single-particle electron crystallography. MBP was found to be a C-shaped molecule when adsorbed to a lipid monolayer, but the relevance of this 3D structure to MBP function has not been assessed.

In the shiverer mouse mutant, the gene for MBP is mostly deleted (Roach et al. 1985). While the CNS of these animals is almost completely devoid of myelin, the PNS myelin appears to be both qualitatively and quantitatively normal (Privat et al. 1979, Kirschner & Ganser 1980, Rosenbluth 1980, Inoue et al. 1981). Thus, the absence of MBP does not prevent the formation of compact myelin in the PNS. Of special interest with regard to human disease is that MBP is the major autoantigen in multiple sclerosis (MS), an autoimmune demyelinating disease of the human CNS (reviewed by Conlon et al. 1999).

2.2.1.3. Peripheral myelin protein-22

PMP-22 is a quantitatively minor component of PNS compact myelin (Haney et al. 1996), whose function is not presently known, although it is suspected to participate in adhesive processes via its glycosylated extracellular domain (Snipes et al. 1993). It contains four transmembrane domains (D'Urso & Muller 1997), and its extracellular domain carries the HNK-1 carbohydrate epitope (Snipes et al. 1993). It is mainly expressed by myelinating Schwann cells, but is also found in various neural and

27

nonneural tissues (Welcher et al. 1991, Baechner et al. 1995, Parmantier et al. 1995).Based on morphological phenotypes of PMP-22 point mutants, the trembler mice

(Suter et al. 1992a, b), PMP-22 appears to be involved in the initial phases of myelination, maintenance of myelin, and neuron-glia interactions (reviewed by Garbay et al. 1999). The importance of PMP-22 for PNS myelin has been confirmed by studies on PMP-22 -deficient mice (Adlkofer et al. 1995), which concluded that PMP-22 is involved in the initiation of myelination, the determination of myelin thickness, and the stability of the myelin sheath.

PMP-22 is involved in a line of human neuropathies. A subgroup of CMT type 1A is caused by point mutations in the gene encoding PMP-22 (Valentijn et al. 1992). Cell culture studies have indicated that these mutations cause impaired intracellular trafficking of the PMP-22 protein, which is not incorporated into the plasma membrane (Naef & Suter 1999, Tobler et al. 1999). In addition, changes in the gene dosage of PMP-22 are linked to human neuropathies (Patel et al. 1992), an observation that is supported by studies on mice overexpressing PMP-22 (Huxley et al. 1996), and on PMP-22-deficient mice (Adlkofer et al. 1995, 1997).

2.2.1.4. Myelin protein 2

Myelin protein 2 (P2) is a 14.8-kDa cytosolic protein expressed only in the nervous system, located on the cytoplasmic side of compact myelin membranes (Trapp et al. 1984a, Narayanan et al. 1988). It is expressed by Schwann cells that have established a one-to-one relationship with an axon (Trapp et al. 1984a). P2 is a highly basic protein, but it shares no significant similarity at the amino acid sequence level to the MBPs. It is a member of a family of fatty acid-binding proteins, with a high affinity for oleic acid, retinoic acid, and retinol (Uyemura et al. 1984). Postulated roles for the fatty acid-binding proteins include the uptake and transport of fatty acids to intracellular organelles, the targeting of fatty acids to specific metabolic pathways, and the maintenance of intracellular pools of fatty acids. Thus, P2 could be involved in the assembly, remodeling, and maintenance of myelin.

2.2.2. Proteins of noncompact myelin

As discussed above, the myelin sheath also contains several membrane compartments with a noncompact structure, such as the paranodal loops, the Schmidt-Lanterman incisures, and the periaxonal collar. These myelin-related membranes define cytoplasmic pockets in the sheath, where dynamic processes, such as transmembrane signaling and ion flux, are likely to occur. These structures also carry a defined set of proteins, including MAG (section 2.3), connexins, periaxin, S100β, and the neural cell adhesion molecule (N-CAM), that are thought to be of importance for myelin formation and maintenance.

Connexins are proteins that form the so-called gap junctions between adjacent cells or

28

cell compartments, allowing small molecules to diffuse freely through membranes (reviewed by Goodenough et al. 1996). The presence of Cx32 in PNS myelin was not noticed until mutations in the gene encoding Cx32 were found to be the cause for X-linked CMT (Bergoffen et al. 1993, reviewed by Scherer et al. 1997). The localisation of Cx32 in myelin closely resembles that of MAG, in that Cx32 is localised in Schmidt-Lanterman incisures and paranodal loops (Scherer et al. 1995a). The expression of the Cx32 mRNA and protein parallels that of other myelin proteins during development and after nerve injury (Scherer et al. 1995a).

Periaxin was first identified as a relatively abundant 147-kDa protein of myelinating Schwann cells in a screen for novel cytoskeleton-associated proteins with a role in peripheral nerve myelination (Gillespie et al. 1994). Periaxin is detectable at an early stage of PNS development, and it is initially concentrated in the plasma membrane, the abaxonal membrane, and the adaxonal membrane. As the myelin sheath matures, periaxin becomes more concentrated in the abaxonal membrane and plasma membrane (Scherer et al. 1995b). Periaxin has been suggested to be involved in attaching Schwann cells to axons and basal laminae by linking the cytoskeleton with membrane components (Gillespie et al. 1994, Scherer et al. 1995b); however, no direct evidence for the function of periaxin exists.

Originally identified as highly soluble proteins of the nervous system (Moore 1965, Moore et al. 1968) that are present in glial cells (Hydén & McEwen 1966), the S100 proteins comprise a group of small, acidic proteins belonging to the EF-hand family of calcium-binding proteins (reviewed by Donato 1999). They are believed to participate in the regulation of energy metabolism, the regulation of cell shape and polarity, and intracellular calcium-regulated signal transduction (reviewed by Donato 1999). The 10-kDa calcium- and zinc-binding (Baudier & Gerard 1983, Baudier et al. 1983, 1984) S100β protein is present in the cytoplasm and associated with plasma membranes of myelinating Schwann cells (Spreca et al. 1989). It is absent from compact myelin membranes (Spreca et al. 1989, Mata et al. 1990), but is present at the Schmidt-Lanterman incisures and in the paranodal loops (Mata et al. 1990). In cultured Schwann cells, a fraction of S100β is associated with the cytoskeleton, and calcium plays a regulatory role in this association (Rambotti et al. 1990). A role for S100β in the regulation of the Schwann cell microtubular cytoskeleton during myelination has been suggested (Mata et al. 1990), as it is able to interact with tubulin and microtubule-associated proteins (MAP) (Baudier et al. 1987, Baudier & Cole 1988) and to regulate the assembly of tubulin into microtubules (Baudier et al. 1982, Endo & Hidaka 1983, Donato 1983, 1987, Hesketh & Baudier 1986). Another widely studied function of the S100 proteins is the regulation of phosphorylation of several protein kinase substrates (Patel et al. 1983, Baudier et al. 1987, Baudier & Cole 1988, Deloulme et al. 1990, Baudier et al. 1992, Lin et al. 1994, Sheu et al. 1994, 1995). Calmodulin, another small calcium-binding protein that acts by regulating the phosphorylation of target proteins, is also localised in the same regions as S100β (Mata & Fink 1988).

The 120-kDa glycosylphosphatidylinositol (GPI) -anchored form of N-CAM has been shown to be present in CNS myelin, where its extraction behaviour represents that of the so-called lipid rafts. Following detergent extraction and density gradient centrifugation, N-CAM is present in the floating fraction that also contains MAG (Krämer et al. 1997). In the PNS, N-CAM is considered one of the markers for nonmyelinating Schwann cells

29

(Zorick & Lemke 1996), but it seems to be able to partially substitute for MAG in the PNS of MAG-deficient mice (Schachner & Bartsch 2000). The exact role of N-CAM in myelination has not been elucidated.

2.2.3. Protein kinases implicated in myelination

Signal transduction events and gene regulation play a major role in glial cell differentiation towards a myelinating phenotype, and during the formation and maintenance of a normal myelin sheath (Bhat 1995). However, only a few protein kinases have actually been studied in myelinating cells and been shown to be important for myelination.

Several lines of evidence support the notion that the Fyn tyrosine kinase, a member of the src family of protein kinases, may play a role in myelination. Fyn is activated during the initial stages of CNS myelination, and its association with MAG has been reported (Umemori et al. 1994). Fyn is present in the DIG fraction from myelinating oligodendrocytes (Krämer et al. 1999). Fyn has been shown to be involved in the expression of the MBP gene (Umemori et al. 1999) and, in Fyn-deficient mice, CNS myelination is impaired (Umemori et al. 1994, Schachner & Bartsch 2000). Furthermore, Fyn is involved in the differentiation of oligodendrocytes (Osterhout et al. 1999). These studies indicate that Fyn participates in the initial events of CNS myelination as a signaling molecule downstream of MAG, and that it may also regulate myelin gene expression.

Protein kinase C (PKC) is a family of phospholipid-dependent, ubiquitous serine/threonine kinases known to phosphorylate a large number of proteins (Dekker & Parker 1994). In rat CNS myelin, there is high PKC activity, and the PKC levels from postnatal day 14 to 42 parallel the deposition of myelin proteins (Yoshimura et al. 1992). Several myelin-specific proteins, including MAG (Kirchhoff et al. 1993, Yim et al. 1995), MBP (Vartanian et al. 1986), and CNPase (Vartanian et al. 1992), are also substrates for PKC. PKC is involved in regulating myelin protein gene expression in cultured oligodendrocytes (Asotra & Macklin 1993), and it has been suggested to regulate process outgrowth from myelinating oligodendrocytes (Yong & Oh 1997).

Protein kinase A, or cyclic 3',5'-adenosine monophosphate (cAMP) -dependent protein kinase, is another serine/threonine kinase present in a variety of cell types, which is activated by the cAMP-induced dissociation of the regulatory subunit from the catalytic subunit (reviewed by Daniel et al. 1998). cAMP is involved in the differentiation of Schwann cell progenitors into mature myelin-producing cells (reviewed by Walikonis & Poduslo 1997), and it is likely that its action involves the activation of PKA. Indeed, a cAMP-dependent kinase activity phosphorylating several myelin proteins has been detected in purified rabbit peripheral nerve myelin (Zabrenetzky et al. 1984).

30

2.2.4. The myelinating glial cell cytoskeleton and myelination

It is clear that great morphological adaptation must occur in the process of myelination. A main player in the morphology of a cell is its cytoskeleton. Numerous studies have been performed in order to characterise the roles of individual cytoskeletal components for myelinating glial cell morphology, but the actual amount of information obtained from these studies is relatively small. A central role for the glial cell cytoskeleton has nevertheless been suggested in the process of myelination (Wilson & Brophy 1989, Dyer 1997). Two of the most abundant cytoplasmic proteins, actin and tubulin, have been the subjects of most studies related to the role of the cytoskeleton in myelin formation and maintenance.

Actin is an ubiquitous protein, the polymerisation of which leads to the formation of microfilaments (reviewed by Steinmetz et al. 1997). Actin microfilaments are crucial for dynamic events in the cell, such as movement and cell division. Actin has been shown to be important for the morphological changes as well as for the induction of myelin gene expression occurring in Schwann cells during the myelination process (Fernandez-Valle et al. 1997). In myelinating glia, actin is localised in the noncompact regions of myelin (Trapp et al. 1989a).

Tubulin is an abundant protein, the polymerisation of which, powered by guanosine 5'-triphosphate (GTP) hydrolysis, leads to the formation of microtubules (Fig. 3). Microtubules are formed by the lateral aggregation of protofilaments, each of which is composed of a chain of stable αβ tubulin dimers (Stephens 1982). The protofilaments are arranged in parallel in order to form the cylindrical microtubule wall. Both α and β tubulin bind a guanine nucleotide; while GTP is nonexchangeably bound to α tubulin, GTP and guanosine 5'-diphosphate (GDP) can exchange on β tubulin. Microtubules are dynamic structures important for the generation and maintenance of cell polarity, intracellular transport, and organelle movement. The recently determined 3D structure for tubulin (Nogales et al. 1998) has provided a detailed insight into microtubule properties and function (Downing & Nogales 1998). Furthermore, a number of accessory proteins capable of stabilising or destabilising microtubule polymers have been identified (reviewed by Cassimeris 1999). The best characterised family of microtubule-associated proteins is the tau/MAP family of proteins (Drewes et al. 1998). Of these proteins, tau, MAP2, and MAP4 have three to four distinct short tubulin-binding sequences (Maccioni et al. 1989, Drewes et al. 1998) that have been used as models for tubulin-binding proteins.

The role of microtubules in myelination has mainly been studied from the point of view of intracellular transport of myelin constituents. Microtubules are abundant in the cytoplasm of myelinating glia (Kidd et al. 1996). Normal microtubule turnover is required for the maintenance of membrane sheets by cultured oligodendrocytes (Benjamins & Nedelkoska 1994), and it is also essential for the correct transport of P0 and MAG to their sites of residence in the myelin sheath (Trapp et al. 1995). Of special interest in the context of myelination are the effects of microtubule-binding drugs on myelinating glial cells. Taxol, a microtubule-stabilising drug used in cancer treatment, causes peripheral neuropathy in some patients, characterised by massive accumulation of microtubules in Schwann cell cytoplasm (Vuorinen et al. 1989, Vuorinen & Röyttä 1990). On the other hand, colchicine, a microtubule-destabilising drug, has been reported

31

to cause a variety of abnormalities in myelin, including demyelination and the inhibition of myelination, when injected into rabbit optic nerves (Matthieu et al. 1981).

2.2.5. Lipid composition of the myelin membrane

Even though the lipid bilayer serves as a permeability barrier into which the various membrane proteins are inserted, particular lipids may modulate the biological activity of membrane-associated enzymes and other membrane proteins, and participate in cell

32

outside inside+ end

- end

β

α

β

β

β

α

α

α

taxol

colchicine

GTP

GDP

GTP/GDP

MAP

Fig. 3. A schematic diagram of tubulin and microtubule structure, indicating the locations of important functional sites (modified after Downing & Nogales 1998). Shown are the binding sites for GTP/GDP, taxol, colchicine, and the tau/MAP family of tubulin-binding proteins. Portions of two vertical protofilaments are shown; a microtubule is typically a cylider formed from 13 protofilaments. The plus end is the faster growing end of the microtubule, and the binding sites for the MAPs reside on the outer face of the tubule.

adhesion and signal transduction events (Gurr & James 1971, Ledeen 1989, Saqr et al. 1993, Quarles et al. 1997). All the major lipid classes and their respective subclasses encountered in other membranes are also represented in the myelin membrane (Hamberger & Svennerholm 1971) and, strictly speaking, there are no myelin-specific lipids. However, the lipid composition of the myelin membrane is unique, and the fatty acyl components of myelin lipids also present a distinctive distribution when compared with those observed in non-myelin membranes (Hamberger & Svennerholm 1971, reviewed by Garbay et al. 2000).

Myelin contains 70-80 % lipid (reviewed by Garbay et al. 2000), including some components that are characteristic of myelin and myelin-forming cells, such as GalC, its sulfated derivative sulfatide, and ethanolamine plasmalogen (reviewed by Morell et al. 1994). Of particular interest with regard to myelination is that GalC has been shown to be part of a signaling pathway that affects calcium influx and cytoskeletal organisation in cultured oligodendrocytes (Dyer & Benjamins 1990). Furthermore, mice lacking the ability to synthesise GalC or sulfatide form dysfunctional and unstable myelin (Coetzee et al. 1996, 1998, Stoffel & Bosio 1997).

Gangliosides are a class of complex sialic acid-containing glycosphingolipids, which are present in high concentration in neurons (Derry & Wolfe 1967), but are also present in glia (Hamberger & Svennerholm 1971). The ganglioside composition of PNS myelin is different from CNS myelin, the major ganglioside being a glucosamine-containing monoganglioside, LM1 (Morell et al. 1994). Interestingly, mice lacking complex gangliosides develop both CNS and PNS pathologies, the latter being characterised by demyelination and axonal degeneration (Sheikh et al. 1999).

Cholesterol accounts for approximately 30 % of the total lipids in the PNS (reviewed by Garbay et al. 2000). In mouse and rabbit sciatic nerves, cholesterol accumulates continuously throughout the period of myelinogenesis and during the subsequent period of myelin maturation (Yates & Wherrett 1974, Juguelin et al. 1986). This accumulation pattern is consistent with the role proposed for cholesterol in the stabilisation and compaction of the multilamellar myelin membrane (Nussbaum et al. 1969, Detering & Wells 1976).

The myelin sheath synthesised by oligodendrocytes and Schwann cells has a similar lipid composition to the detergent-insoluble glycosphingolipid-enriched complexes (DIG) (Parton & Simons 1995), in that it has a high glycosphingolipid content and is rich in cholesterol (Krämer et al. 1997). It is thought that DIGs represent functional microdomains within the plasma membrane (Sargiacomo et al. 1993), which are sorted to the apical surface of polarised epithelial cells (Simons & Wandinger-Ness 1990). Along with these so-called lipid rafts, certain proteins, including several cytoplasmic signal transduction molecules, are targeted to specific locations within the cell (Simons & Ikonen 1997).

2.2.6. Inorganic components of myelin

The myelin sheath has attracted attention in terms of electrostatic interactions of metals with charged membrane lipids, or with isolated myelin membranes. In the course of these

33

studies, the concentrations of several inorganic species have been determined in isolated myelin and in the perinuclear cytoplasm of glial cells (table 1).

The most studied metal in the context of myelination is zinc, a transition metal essential for normal cellular function (Vallee & Falchuk 1993). Zinc is present in myelin at a higher concentration than any other trace element (Berlet et al. 1994), and it seems to have unique effects on myelin structure. Zinc appears to favor myelin compaction (Inouye & Kirschner 1984) and to inhibit the release of MBP from myelin (Earl et al. 1988). Interactions of zinc with myelin proteins, including MBP, PLP, and P0, have been suggested (Inouye and Kirschner 1984), but to date, experimental evidence only exists for MBP as a zinc-binding protein (Riccio et al. 1995, Tsang et al. 1997).

34

Table 1. Levels of some inorganic components in isolated myelin membranes and glial cell cytoplasm.________________________________________________________________________Method Compartment Component (unit) Concentration

PNS CNS________________________________________________________________________

Electron probe glial cytoplasm water (%) 76 80X-ray microanalysis K (mmol/kg dry weight) 580 530(Stys et al. 1997) Na -"- 170 190

Cl -"- 120 100Ca -"- 3.3 4.0

myelin water (%) 35 53K (mmol/kg dry weight) 150 250Na -"- 210 170Cl -"- 110 90Ca -"- 3.3 3.0

Flame atomic myelin Mn (µmol/g protein) - 0.24absorption Co -"- - 0.50spectrophotometry) Cu -"- - 0.41(Berlet et al. 1994 Zn -"- - 1.41

Hg -"- - 0.66Cd -"- - 0Pb -"- - 0.9Mg -"- - 69.45Ca -"- - 30.14

________________________________________________________________________

2.3. The myelin-associated glycoprotein

The presence of MAG in isolated CNS myelin fractions was originally reported in the early 1970s (Quarles et al. 1972, 1973), and the protein has been the subject of numerous studies for more than 25 years. However, controversies on the function and importance of MAG still exist, and ambiguous results concerning MAG can frequently be found in the literature.

2.3.1. The mag gene

In rodents, the mag gene spans approximately 15-16 kilobase pairs (Fujita et al. 1989, Nakano et al. 1991), contains 13 exons, and expresses multiple mRNAs (Lai et al. 1987a, Tropak et al. 1988). The mag gene possesses a TATA-less promoter region (Laszkiewicz et al. 1997) with several short sequence homologies with two other myelin proteins, MBP and PLP (Nakano et al. 1991). It is possible that such homologous regulatory elements could play a role in the expression of myelin-specific genes. The developmental activation of the gene coincides with profound DNA demethylation that may provide auxiliary regulatory mechanisms (Konat 1996). The mouse mag gene has been mapped to chromosome 7, and the human gene resides in the long arm of chromosome 19, in the region 19q13.1-3 (Barton et al. 1987, D'Eustachio et al. 1988, Spagnol et al. 1989, Garcia et al. 1995).

Three types of mRNA, differing by the presence or absence of exons 2 and 12, are transcribed from the mag gene (Fujita et al. 1989): one type in which both exons 2 and 12 are absent, one in which exon 2 is spliced out, and one in which only exon 12 is missing. No mRNAs containing both exons have been found. In the CNS, most mRNA contains exon 2, while in the PNS, the form without exon 2 predominates (Tropak et al. 1988). The significance of the presence or absence of exon 2 in MAG mRNAs is not known. Since exon 2 is in the non-coding region of the mRNA, it does not affect the structure of the encoded protein. Exon 12 contains 45 base pairs (bp), including an in-frame stop codon. The 10 amino acids encoded by this exon comprise the isoform-specific carboxy terminus of S-MAG. When exon 12 is spliced out, the resulting mRNA, which contains a second in-frame stop codon in exon 13, codes for the 54-residue isoform-specific carboxy terminus of L-MAG.

Exon 4 encodes the signal peptide sequence and the first N-terminal residues of the mature MAG protein. Exons 5-9 each encode one of the five Ig domains of the extracellular region of MAG. The transmembrane domain is encoded by exon 10, while the isoform-specific cytoplasmic domains are encoded by exons 11 and 12 (S-MAG) and 11 and 13 (L-MAG), respectively (Fujita et al. 1989, Lai et al. 1987a, Nakano et al. 1991).

35

2.3.2. The expression and subcellular localisation of MAG

MAG is expressed only by myelinating glial cells in the nervous system. It is noteworthy that MAG is one of the few myelin-specific proteins expressed by myelinating glia in both branches of the nervous system (Quarles et al. 1972, 1973, Figlewicz et al. 1981), suggesting a crucial role in myelin formation and maintenance. As described above, two different isoforms of MAG are produced by alternative mRNA splicing.

In the rodent CNS, the expression of the two MAG isoforms is differentially regulated during postnatal development (Frail & Braun 1984). The L-MAG mRNA is expressed at highest levels during the early developmental stages of the rat, reaching its highest concentration during the third week when myelination is most intense. Thereafter, L-MAG mRNA levels decrease, to reach the adult level at the age of two months. S-MAG mRNA, on the other hand, is very little expressed during the early stages of myelination, but becomes more and more abundant as myelination progresses. S-MAG mRNA is the predominant form in the adult rat brain (Frail & Braun 1984, Lai et al. 1987a, Tropak et al. 1988). In adult mouse brain, S- and L-MAG polypeptides are both present in significant amounts, with S-MAG predominating (Pedraza et al. 1991), while L-MAG predominates in young animals (Braun et al. 1990, Inuzuka et al. 1991, Pedraza et al. 1991).

The expression of MAG in the rat PNS preceeds that of P0 and MBP (Stahl et al. 1990), the two major structural proteins of PNS myelin. S-MAG mRNA predominates throughout development, reaching its highest level in the rat during the second postnatal week. L-MAG mRNA is present in very small amounts at all ages (Tropak et al. 1988). Correlatively, S-MAG protein represents the large majority of MAG in the nerves (Pedraza et al. 1991). However, as in the CNS, L-MAG expression occurs during the early stages of myelination (Pedraza et al. 1991), peaking in the rat at 8 days after birth, while S-MAG is maximally expressed at 15 days (Inuzuka et al. 1991).

In the human nervous system, the expression of the MAG isoforms apparently differs from rodents in some aspects. In post-mortem brain, both the L-MAG mRNA and protein are present at a much higher level than those of S-MAG. On the other hand, in the PNS, S-MAG is clearly the predominant isoform. (Miescher et al. 1997)

In line with the mRNA expression, the MAG protein has only been detected in myelin-forming cells. In the CNS, MAG is enriched exclusively in the periaxonal membrane of myelin internodes (Sternberger et al. 1979), while in the PNS, it is found in the periaxonal membrane, paranodal loops, Schmidt-Lanterman incisures, and the inner and outer mesaxons (Sternberger et al. 1979, Trapp & Quarles 1982, Martini & Schachner 1986, Trapp et al. 1989a). In other words, MAG is present in the noncompacted membranes of the myelin sheath. Studies of human post-mortem samples have raised the possibility that, in the PNS, the small amount of L-MAG present may be confined to the periaxonal membrane, while S-MAG seems to be present in all noncompacted myelin membranes listed above (Miescher et al. 1997).

In transfected kidney epithelial cells, differential localisation has been observed for the two MAG isoforms. While in both transiently and stably transfected cell lines, L-MAG was sorted primarily to the basolateral membrane, the sorting of S-MAG varied under different conditions. The conclusion was drawn that L-MAG contains an invariable basolateral sorting signal, while the sorting of S-MAG is dependent upon extrinsic

36

factors, such as coexpression by adjacent cells. (Minuk & Braun 1996)The extraction behaviour of MAG from nervous tissue has shed light on the

macromolecular complexes that could be associated with MAG. In particular, extraction studies of CNS myelin (Pereyra et al. 1988) have indicated that MAG cofractionates with a group of proteins that have also been shown to be enriched in the radial component (Karthisagan et al. 1994). It has thus been suggested that these proteins - including tubulin, CNPase, MBP, and MAG - could be parts of a heteromolecular complex with interactions spanning several layers of the myelin membrane (Pereyra et al. 1988). Analysis of DIGs from cultured oligodendrocytes and purified CNS myelin (Krämer et al. 1997) has shown that a portion of MAG is present in the DIG fraction from myelin, but not in the one obtained from nonmyelinating oligodendrocytes. This suggests that at least a fraction of MAG is sorted along with lipid rafts. Furthermore, a portion of MAG is present in the cytoskeletal fraction from cultured oligodendrocytes, and this association seems to be regulated by phosphorylation (Bambrick & Braun 1991).

2.3.3. The MAG protein

MAG has a molecular weight of 100 kDa (Quarles et al. 1972), of which approximately 30 % is carbohydrate (Frail & Braun 1984). The polypeptide backbone of MAG is comprised of either 582 or 626 amino acid residues (Arquint et al. 1987, Lai et al. 1987a, Salzer et al. 1987). A schematic representation of the two isoforms of MAG is shown in figure 4.

2.3.3.1. The extracellular domain

On the basis of the primary structure, the extracellular domain of MAG is predicted to be composed of five separate Ig domains (Lai et al. 1987b). Thus, MAG is considered a member of the Ig superfamily of cell adhesion molecules (reviewed by Brümmendorf & Rathjen 1994). At the amino acid sequence level, the extracellular domain shares a significant degree of homology with other Ig family members, such as N-CAM and sialoadhesin. While the first Ig domain is of the variable (V) type, Ig domains 2 to 5 are homologous to domains of the constant type (Brümmendorf & Rathjen 1994).

Of special interest is the homology between the first two Ig domains of MAG and the corresponding domains in members of the sialoadhesin family (Kelm et al. 1996), including sialoadhesin, the avian Schwann cell myelin protein (SMP), CD22, and CD33. The first extracellular domain of all the family members is a V-type Ig domain, containing the sialic acid binding site, with unusual structural features (Brümmendorf & Rathjen 1994). The two distinctive properties of this domain are the presence of an intrasheet disulfide bridge, and the linkage of the first and second Ig domains via another disulfide bridge (Pedraza et al. 1990). The structure of this domain from sialoadhesin has recently been determined by X-ray crystallography in complex with the ligand sialyllactose (May et al. 1998), and comparison of the family members has highlighted

37

residues involved in direct sialic acid binding, that are almost completely conserved throughout the family. A crucial residue for the binding of sialic acid by MAG is Arg118 within the first Ig-like domain (Tang et al. 1997). The sialoadhesin structure should also be useful for studies on the function of the other sialic acid binding proteins, and on the differences between them. A model for the structure of the MAG first Ig domain, based on this structure, is shown in figure 5.

2.3.3.2. The cytoplasmic domain

As the result of alternative mRNA splicing of exon 12, two isoforms of MAG are produced, differing only by the carboxy-terminal portion of their cytoplasmic domain (Lai et al. 1987a, Tropak et al. 1988). The length of the cytoplasmic tail of S-MAG is 46 residues, and that of L-MAG 90 amino acids. The MAG isoforms share a common cytoplasmic domain of 36 amino acids, which is followed by the isoform-specific carboxy-terminal domains of 10 or 54 amino acids for S- and L-MAG, respectively

38

extracellular

plasmamembrane

cytoplasm

S-MAG L-MAG

Fig. 4. The two isoforms of MAG. S- and L-MAG differ from each other only by the carboxy-terminal portions of their respective cytoplasmic tails.

(Salzer et al. 1987). At the primary sequence level, neither of the MAG cytoplasmic domains (MAGcts) shares a significant degree of homology with other proteins. However, there are a few short sequence motifs, such as phosphorylation consensus sequences, that have been used to predict the functional properties of the MAG cytoplasmic domains (Arquint et al. 1987, Salzer et al. 1987). Apart from studies on MAG phosphorylation, described below, no studies aimed at characterising the structure or function of the MAGcts have been reported.

2.3.4. Posttranslational modifications

2.3.4.1. Phosphorylation

The only intracellular ligands for the MAG proteins detected so far are enzymes involved in signal transduction. Phosphorylation of L-MAG has been reported to occur on serine,

39

Fig. 5. Left: A ribbon model for the structure of the first amino-terminal Ig domain of MAG. The β-strands are indicated by arrows. Shown are the amino (N) and carboxy (C) termini of the domain. The model was built using the automated SWISS-MODEL server (Peitsch 1996, Guex & Peitsch 1997) and SwissPdbViewer software (Guex & Peitsch 1997) on the basis of the high sequence homology between the corresponding domain in MAG and sialoadhesin, the structure of which has been determined (May et al. 1998). Right: A surface representation of the same model, indicating the sialic acid binding site. The sialyllactose molecule (black) was imported from the liganded sialoadhesin structure, and the terminal sialic acid (Sia) is indicated.

threonine, and tyrosine residues (Edwards et al. 1988, Afar et al. 1990, Kirchhoff et al. 1993), while almost all phosphorylation of S-MAG occurs on serine (Agrawal et al. 1990, Kirchhoff et al. 1993).

Both MAG isoforms are phosphorylated on serine residues by PKC (Kirchhoff et al. 1993). Fyn (Umemori et al. 1994) and other src family tyrosine kinases (Edwards et al. 1988) have also been reported to phosphorylate MAG in vitro. A PKA consensus phosphorylation site is present in the L-MAG-specific domain, but evidence from cell culture studies has suggested that MAG is not phosphorylated by PKA (Afar et al. 1990, Kirchhoff et al. 1993, Yim et al. 1995). The phosphorylation of the MAG cytoplasmic domains has been suggested to regulate the interactions of MAG with its intracellular ligands, such as the cytoskeleton. The functional importance of the MAGct phosphorylation events and the identities of other MAG cytoplasmic ligands were still unknown at the outset of this study.

2.3.4.2. Glycosylation

The extracellular domain of MAG possesses several putative N-glycosylation sites, and these sites have been shown to be targets for glycosylation (Burger et al. 1993, Tropak & Roder 1997). Most of the oligosaccharides are of the complex type due to the presence of sialic acid and sulfate moieties (Quarles et al. 1992, Bartoszewicz et al. 1995).

Among the carbohydrate motifs, MAG possesses the L2/HNK-1 epitope (McGarry et al. 1983, Inuzuka et al. 1984a, Nobile-Orazio et al. 1984) which is present in P0, PMP-22, and several other cell adhesion molecules, and whose expression varies during peripheral nerve development and regeneration (Martini & Schachner 1986, Martini et al. 1988). The L2/HNK-1 -reactive oligosaccharides of MAG are primarily on Ig domains 4 or 5 (Fahrig et al. 1993, Pedraza et al. 1995). The L2/HNK-1 epitope has attracted considerable interest with regard to both cell-cell interactions and demyelinating diseases. A role for the L2/HNK-1 carbohydrate structure in peripheral nerve myelination is supported by the identification of binding sites for this determinant in PNS membranes (Needham & Schnaar 1993).

2.3.4.3. Acylation

A cysteine residue within the MAG transmembrane domain is a target for palmitoylation, occurring posttranslationally via a thioester linkage (Pedraza et al. 1990). A similarly positioned cysteine residue is also found in the related proteins SMP (Dulac et al. 1992) and CD22 (Stamenkovic & Seed 1990). An intramembrane palmitoylated cysteine residue has also been reported in the myelin protein P0 (Agrawal et al. 1983). The covalent linkage of an acyl group could play a role in the targeting of MAG to specific membrane subdomains.

40

2.3.4.4. Proteolysis

In the CNS, MAG has been shown to be degraded by an endogenous Ca2+-activated neutral protease present in myelin, forming a 90-kDa soluble derivative containing the whole MAG extracellular domain (dMAG) (Inuzuka et al. 1984b, Sato et al. 1984, Möller 1996). Originally, the enzyme responsible for the proteolysis was thought to be similar to calpain II, but recent evidence suggests that the activity actually resembles that of cathepsin L (Stebbins et al. 1997). The rate at which MAG is degraded by this activity is species-dependent, being considerably faster in human brain than in the brain of any other species tested (Möller 1996), including other primates, and the degradation is increased in multiple sclerosis lesions when compared to normal brain myelin (Möller et al. 1987, Quarles 1989, Möller 1996). The proteolytic site has recently been mapped to the extracellular domain/transmembrane domain junction (Stebbins et al. 1997). The putative smaller degradation product of MAG, containing the transmembrane and cytoplasmic domains, has not been studied. While the proteolysis of MAG may have a role in the progression of several pathological neurological conditions, dMAG has also been detected in healthy human cerebrospinal fluid (Yanagisawa et al. 1985). Thus, the proteolytic modification of MAG may also be part of normal tissue remodeling.

2.3.5. The proposed functions of MAG

Although a number of functions have been assigned to MAG, relatively few functional properties are actually backed up by experimental evidence. The hypotheses on the potential functions of MAG have been based mainly on the subcellular localisation of MAG, its biochemical properties, and the phenotypic abnormalities observed in the MAG-deficient mice. The presence of a large adhesion-capable extracellular domain and two alternative cytoplasmic tails has raised the possibility that MAG might act as a multifunctional protein. The most commonly suggested funtions of MAG are described below.

2.3.5.1. Initiation of myelination

Based on its biochemical properties and its enrichment in periaxonal membranes during axonal ensheathment, MAG was proposed to be important for the initiation and progression of myelination (Owens & Bunge 1989). This hypothesis was further supported by the findings that the overexpression of L-MAG caused accelerated myelination by Schwann cells (Owens et al. 1990), while hypomyelination was observed by Schwann cells that underexpressed MAG (Owens & Bunge 1991). However, the apparently normal initiation of myelin formation in MAG-deficient mice (Li et al. 1994, Montag et al. 1994) has raised questions concerning the importance of MAG during the early stages of myelination.

41

2.3.5.2. Membrane spacing and intercellular adhesion

A common property of all MAG-containing membranes is that they face another membrane with an extracellular distance of 12-14 nm (Trapp & Quarles 1982). The presence of MAG and a normal periaxonal space (12-14 nm) are rarely dissociated, as shown by studies on pathological conditions (Trapp et al. 1984b). The size of the MAG extracellular domain would be suitable for maintaining such a space (Trapp 1990). Furthermore, in the trembler mutant mouse, the frequently observed loose myelin lamellae within the myelin sheath contain MAG (Vallat et al. 1999). The presence of the five Ig homology domains, and the reported adhesion properties of the MAG extracellular domain further stress the possibility that MAG is directly involved in cell adhesion events.

MAG is a member of the sialoadhesin family, a subgroup of the Ig superfamily, that is characterised by the ability to bind sialic acid (Kelm et al. 1996). Sialic acid is a common component of cell surface glycoproteins and glycolipids. Thus, MAG may promote cell adhesion through binding to sialic acid epitopes on other cells. In line with this hypothesis, MAG has been shown to specifically bind gangliosides (Yang et al. 1996, Collins et al. 1997), sialic acid-containing lipids present in both axonal and glial cell membranes.

Primary sequence analysis has highlighted the presence of an arginine-glycine-aspartic acid (RGD) consensus motif (Pierschbacher & Ruoslahti 1984) for integrin binding within the first Ig domain of MAG. However, it has been suggested that this motif is buried within the Ig fold, and would thus not be able to bind integrin (Pedraza et al. 1990). Experimental evidence supporting this hypothesis has also been obtained, as RGD-containing peptides were unable to inhibit the binding of MAG to neurons (Sadoul et al. 1990), and anti-RGD antibodies do not detect the motif in MAG (Pedraza et al. 1990). Furthermore, the RGD motif is buried in the molecular model shown in figure 5. Other protein ligands for the MAG extracellular domain have been detected. They include unidentified axonal counterparts (DeBellard et al. 1996), and the extracellular matrix proteins tenascin-R (Yang et al. 1999) and collagen (Fahrig et al. 1987). The significance of the observed interactions has, however, not been established, although an interaction between MAG and tenascin-R at the nodes of Ranvier has been proposed (Schachner & Bartsch 2000).

2.3.5.3. Membrane motility

One of the most drastic changes in glial cell morphology during myelination is the synthesis of large amounts of membrane, and the coordinated wrapping of this membrane around the axon. It is clear that a cytoskeleton-based machinery is required for such movements. As all MAG-containing membranes, especially the periaxonal membrane during axonal ensheathment and the mesaxon during spiral growth, have the ability to move, it has been proposed that MAG would be involved in membrane movement (Attia et al. 1989, Trapp 1990). The phosphorylation status of MAG has been suggested to play a role in cytoskeletal interactions (Attia et al. 1989), and the association of MAG with

42

cytoskeletal components has been reported to be enhanced upon phosphorylation (Bambrick & Braun 1991). However, as no such ligands for the MAG cytoplasmic domains have previously been identified, the only evidence supporting this hypothesis is the colocalisation of MAG with cytoskeletal components, such as F-actin and spectrin, in PNS myelin (Trapp et al. 1989a). On the other hand, since the initial stages of myelination in the MAG-deficient mice (Li et al. 1994, Montag et al. 1994) seem to proceed normally, apart from subtle abnormalities of the periaxonal collar, the direct participation of MAG in membrane movement remains to be demonstrated.

2.3.5.4. Endocytosis

The only clear difference detected in the subcellular localisation of the two isoforms of MAG in myelinating glia is the enrichment of L-MAG in endosomes in the CNS. The first association of L-MAG with endosomes was demonstrated by studies which determined the localisation of MAG at times when the L-MAG (7 days), or the S-MAG (adult) were quantitatively the abundant form (Trapp et al. 1989b). During early stages of myelination, MAG was enriched in endosomes, indicating that L-MAG may be associated with an endocytic pathway that originates in the periaxonal membrane and terminates in the oligodendrocyte perinuclear cytoplasm. These early observations have been confirmed and extended by studies on the quaking mutant mice (Bo et al. 1995), showing that L-MAG, but not S-MAG, is selectively removed from the periaxonal membrane in the CNS by receptor-mediated endocytosis. In the quaking mutant, the endocytosis of L-MAG is increased. A tyrosine-based internalisation motif (Ktistakis et al. 1990) is present in the L-MAG-specific cytoplasmic domain (Bo et al. 1995), but its role in the endocytic process has not been established. The physiological importance of L-MAG endocytosis is presently unclear.

2.3.5.5. Signal transduction

The localisation of MAG in the noncompacted structures of the myelin sheath (Sternberger et al. 1979, Trapp & Quarles 1982, Martini & Schachner 1986, Trapp et al. 1989a), where cytoplasm is present, in addition to the presence of several phosphorylation sites within the MAG cytoplasmic domains, has led to the hypothesis that MAG, especially L-MAG, might play a role in signal transduction, both inside the glial cell and between the glial cell and the neuron.

Strong evidence exists for a role of L-MAG in signal transduction events related to myelinogenesis. Both of the MAG isoforms can bind to the Fyn kinase (Jaramillo et al. 1994). The binding of MAG antibodies to the extracellular domain of L-MAG results in the activation of Fyn inside the cell (Umemori et al. 1994). When phosphorylated on Tyr620, L-MAG binds to phospholipase Cγ (PLCγ) (Jaramillo et al. 1994). While src family kinases, including Fyn, are components of the lipid rafts (Krämer et al. 1999), and as MAG has also been shown to be present in this fraction from CNS myelin (Krämer et

43

al. 1997), it is possible that the MAG-Fyn interaction occurs in the DIG microdomains of noncompact myelin.

2.3.5.6. Effects on axonal properties

An important barrier to axonal regeneration after trauma is an axon growth inhibitory activity that is present in myelin of the CNS and PNS (David et al. 1995). MAG has been identified as one of the myelin components that can act as an inhibitor of neurite outgrowth (reviewed by McKerracher & David 1997). Of the neurons tested, only DRG neurons have been found to exhibit a clear age-dependent reaction to MAG, with the negative response developing after postnatal day 3 (DeBellard et al. 1996). All other neurons, regardless of age, had significantly shorter neurites when plated on MAG. Cell culture studies have implicated MAG in the promotion of neurite outgrowth from neonatal DRG cells (Johnson et al. 1989). The putative inhibitory effect of MAG on neurite outgrowth is likely to be most important in the CNS, where the removal of myelin debris from the site of injury is a much slower process than in the PNS.

Myelination has trophic effects on axons, causing larger axonal caliber (Windebank et al. 1985), increased neurofilament spacing (de Waegh et al. 1992, Sanchez et al. 1996), and greater neurofilament numbers (Sanchez et al. 1996). Although little is known about the molecular basis of this phenomenon, MAG has been suggested to play a role in the process (Yin & Trapp 1997). Significant alterations have been detected in axons of myelinated fibers in MAG-deficient mice, suggesting that MAG may function as a ligand that influences the organisation of the axonal cytoskeleton (Yin et al. 1998).

2.2.6. Pathological conditions related to MAG

To date, no human or animal diseases have been characterised that would be caused by mutations in the mag gene. However, MAG may play roles in the progression of several human and animal pathological conditions (Quarles 1989).

2.2.6.1. Animal mutants

A variety of animal mutants with neurological disorders have been studied with regard to MAG. There are many mammalian, especially mouse, mutants with deficiencies in myelination. Furthermore, several spontaneous animal mutants have been used as models of similar human diseases for years (Snipes & Suter 1995).

Even though mutations in the mag gene have not been linked to phenotypes with deficiencies in myelination, changes in the expression and post-translational modification of MAG are common. The decrease in the expression level of MAG observed in some mutants is often linked to an overall decrease in the level of myelin protein expression

44

(Frail & Braun 1985, Yanagisawa & Quarles 1986). In the quaking mouse, the splicing of MAG mRNA and the glycosylation of the MAG protein are abnormal (Frail & Braun 1985, Bartoszewicz et al. 1995). In the nerves of trembler mice, MAG has a higher-than-normal molecular weight (Inuzuka et al. 1985), caused by abnormal glycosylation (Bartoszewicz et al. 1996).

The phenotype of the quivering (qv) mouse mutant includes peripheral neuropathy and deafness, and the qv locus in mouse chromosome 7 lies in very close proximity of the mouse mag gene (Barton et al . 1987, D'Eustachio et al. 1988, Grosson et al. 1996). The gene responsible for the qv phenotype has not been identified. A form of inherited deafness, DNFA4, has been proposed to be the human equivalent of qv (Petit 1996). Furthermore, the mutation causing DFNA4 has been mapped to human chromosome 19q13.1, corresponding to the region of the human mag gene (Chen et al. 1995). It has not been ruled out that a small mutation in the mag gene could be responsible for the qv phenotype, although the MAG mRNA and protein are both expressed in the mutants (Barton et al. 1987). Thus, mag can still be considered a candidate gene for the qv mouse and human DFNA4 phenotypes.

The hypomyelinating mouse mutant claw paw, caused by an autosomal recessive mutation of an unknown gene, is characterised by delayed myelination in the PNS (Henry et al. 1991). The delay has been attributed to a breakdown in the signaling mechanism between the axon and the Schwann cell. The mag gene is located near the claw paw mutation on mouse chromosome 7 (Barton et al. 1987, Henry et al. 1991). However, several lines of evidence have indicated that MAG would not be the affected gene in claw paw mice. The MAG protein is expressed in the mutants (Koszowski et al. 1998), and the sequence of MAG mRNA is normal (Niemann et al. 1998). The splicing pattern of the MAG mRNA is normal in the CNS (Niemann et al. 1998), but surprisingly, it has not been studied in the PNS, which is the affected tissue.

2.2.6.2. Human peripheral neuropathies

Approximately 10 % of patients with peripheral neuropathy of unknown cause have an associated monoclonal gammopathy (Bosch & Smith 1993). MAG is the most common protein autoantigen in these gammopathies, and the autoantibodies are usually directed towards the extracellular carbohydrate epitopes of MAG (Frail et al. 1984). Injection of these antibodies into cat sciatic nerve causes demyelination (Hays et al. 1987). It is of interest that monoclonal anti-MAG antibodies in patients with demyelinating neuropathy in association with IgM gammopathy have specificities similar to the HNK-1 antibody (Quarles et al. 1992).

In neuropathies associated with IgM monoclonal gammopathies of undetermined significance, the epitopes recognised with the highest affinity by the autoantibodies are present on MAG and could interfere with cell adhesion and cellular signaling processes. In addition, binding to antigenically similar glycoproteins, such as P0, PMP-22 and some acidic glycolipids, may be a contributory factor. It is generally accepted that the anti-MAG autoantibodies are inducing a progressive demyelinating polyneuropathy by modifying axon-Schwann cell interactions. (Miescher & Steck 1996)

45

2.2.6.3. Multiple sclerosis

Multiple sclerosis (MS), one of the very first human diseases to be described scientifically (Carswell 1838, Charcot 1868), is a chronic inflammatory demyelinating disease of the CNS, which is believed to result from an autoimmune attack against the myelin sheath (Steinman 1996, Conlon et al. 1999). Myelin proteins that have been investigated as possible target antigens of the autoimmune response include MBP, PLP, MAG, myelin oligodendrocytic basic protein (MOBP), and MOG (Johnson et al. 1986, Linington et al. 1988, Määttä et al. 1998, Kaye et al. 2000). In animal models, immunisation against each of these proteins is capable of inducing experimental allergic encephalomyelitis, an autoimmune laboratory model of demyelinating disease. Because myelin appears to be a target for the autoimmune attack in MS, genes coding for myelin proteins are potential candidates for involvement in the pathogenesis of MS. Little is known of the association between MS and genes encoding myelin proteins. A recent study suggests that the MAG, MBP, OMGP, and PLP genes do not have a significant effect on the susceptibility to MS, while a potential role of the MOG gene could not be excluded (Seboun et al. 1999).

Patients with multiple sclerosis are characterised by an autoimmune respone directed towards myelin proteins, including MAG (Link et al. 1989, Johnson et al. 1986, Baig et al. 1991, Soderstrom et al. 1994), and MAG is rapidly lost in areas of active disease (Gendelman et al. 1985, Brady & Quarles 1988). At the edges of developing plaques, MAG is reduced more than other myelin proteins (Quarles et al. 1992). While it seems unlikely that mutations concerning MAG would be involved in MS (Seboun et al. 1999), MAG is clearly present in the structures that are degenerated in MS lesions. Indeed, the first pathology observed in early MS lesions is in the periaxonal regions of myelin sheaths (Rodriguez & Scheithauer 1994). While MAG seems to be one of the targets of the autoimmune attack in MS, the significance of this phenomenon has not been confirmed.

2.2.6.4. Gliomas

Schwannomas and oligodendrogliomas arise when myelinating glial cells are transfomed into cancer cells. In oligodendrogliomas, deletions in the region of the mag gene, the long arm of chromosome 19, are common (Weber et al. 1996). Loss of alleles from chromosome 19q, observed in over 80 % of cases (Reifenberger et al. 1994), is the most frequent genetic alteration in oligodendrogliomas. Indeed, oligodendroglial tumors tend to lose the entire long arm of chromosome 19q (Rosenberg et al. 1996). Such frequent allelic losses strongly suggest the presence of one or more putative tumor suppressor genes on chromosome 19q. Despite intensive efforts, however, no 19q glioma tumor suppressor genes have been identified (Yong et al. 1995, Ueki et al. 1997, Peters et al. 1999).

The MAG protein is not expressed by tumor cells in oligodendrogliomas and Schwann cell tumors, whereas myelin sheaths and their remnants within the tumors contain MAG (Schwechheimer et al. 1992). Considering the putative roles of MAG in

46

cell adhesion, signal transduction, and the maintenance of glial cell polarity, it is a feasible hypothesis that mutations in the mag gene could also contribute to the progression of a certain subset of glial cell cancers, originating from myelinating glia. The presence of mutations in the mag gene in gliomas has not been studied.

2.2.6.5. The MAG knockout mice

The MAG-deficient (MAG -/-) mice were generated by inactivation of the mag gene by homologous recombination in embryonic stem cells (Li et al. 1994, Montag et al. 1994). CNS myelin formation is delayed in the mutant animals, and the cytoplasmic collars of mature CNS myelin are frequently missing or reduced. Redundant myelination is common in the CNS of adult MAG mutants (Bartsch et al. 1995), and 8-month-old mice display evidence of dying-back oligodendrogliopathy (Lassmann et al. 1997).

Surprisingly, the PNS phenotype of the MAG-deficent mice was milder than expected, especially in young animals. The initial stages of PNS myelination seemed to proceed normally, and myelin sheaths appreared normal. In adult mutants, however, the structure of myelin was reported to be abnormal (Fruttiger et al. 1995). The hypothesis that MAG was dispensable for PNS myelin formation was supported by the observation that MAG-deficient Schwann cells isolated from the mutants formed normal myelin in cell culture conditions (Carenini et al. 1998).

In order to characterise the functions of the MAG isoforms, mice deficient in only the L-MAG isoform were generated. CNS myelin of the L-MAG-deficient mice displays most of the abnormalities reported for the MAG knockouts. In contrast to the MAG null mutants, however, PNS axons and myelin of older L-MAG-deficient animals do not degenerate, indicating that S-MAG is sufficient to maintain PNS myelin integrity. (Fujita et al. 1998)

In the MAG-deficient mice, increased expression of N-CAM was detected at areas normally containing MAG, and it was suggested that N-CAM may functionally substitute for MAG during myelinogenesis, but not myelin maintenance. This hypothesis was further supported when mice deficient for both MAG and N-CAM were generated (Carenini et al. 1997). In these animals, abnormalities in PNS myelin are detected earlier than in MAG -/- mice. Given that the main isoform of N-CAM detected in myelin is the 120-kDa GPI-anchored form (Krämer et al. 1997), lacking a cytoplasmic tail, it seems likely that N-CAM is able to substitute for MAG in extracellular adhesion, but would not be able to fulfill the cytoplasmic functions of MAG.

The PNS of P0 null mice displays a severe dysmyelinating phenotype, while several other cell recognition molecules, including MAG, are upregulated (Giese et al. 1992). In mice deficient for both MAG and P0 (Carenini et al. 1999), the Schwann cell spiraling was impaired, when compared to the P0 null mice. Similar, but somewhat milder, results were obtained for mice deficient for P0 and N-CAM. These findings demonstrated that MAG and N-CAM can play significant roles during PNS myelin formation. In addition, because these roles were only seen in the absence of P0, it seems that these myelin-related molecules can play distinct, but also partially overlapping roles.

In the CNS of Fyn-deficient mice, a significant hypomyelination is observed. In

47

addition, hypomyelination in the CNS of these mice is more severe than in the CNS of MAG mutants. Interestingly, in mice deficient in both MAG and Fyn, the hypomyelination is more severe, and a critical role for both Fyn and L-MAG in the initiation of myelination has been suggested based in these observations. (Schachner & Bartsch 2000)

As a conclusion from genetic knockout studies, it can be said that MAG seems to be indispensable for the establishment of normal myelin structure. These studies have been suffering from the apparent functional substitution for MAG by other proteins in the mutants. Such problems are common in gene inactivation studies, as backup processes have evolved to ensure the progression of important developmental processes, such as myelinogenesis, even in the cases of mutations.

48

3. Aims of the present study

The myelin sheath is a structurally complex unit, and its correct formation and lifetime maintenance are likely to largely depend on defined protein-protein interactions. Despite the large number of studies on the myelin-specific proteins, very few such interactions have been described.

On the other hand, the existence of two isoforms of MAG, differing only by their cytoplasmic domains, has been an unsolved puzzle ever since their presence was first detected. Therefore, the initial goal of this study was to gain information on the functional and structural properties of the MAGcts, and any differences that might exist therein between the two MAG isoforms.

During the course of the study, several new sub-goals were set:

- to identify ligands binding directly to the MAG cytoplasmic domains, and to characterise these interactions in vitro and in vivo (III, IV, V, VI)

- to study the phosphorylation of recombinant MAG cytoplasmic domains in vitro (I, VI)

- to develop tools and methods for further studies on MAG and other myelin proteins (I, II, III, IV, V, VII)

- to study the amino acid dimorphism of the rat S-MAG cytoplasmic domain, and its potential functional consequences (I, V)

- to begin structural analyses on the MAG cytoplasmic domains

4. Materials and methods

Detailed descriptions of most methods can be found in the original articles.

4.1. Nucleic acid methods

4.1.1. Isolation and analysis of RNA (I, VII)

For the developmental study of S-MAG subtype mRNA expression (I), total ribonucleic acid (RNA) was prepared from the sciatic nerves of individual rats, aged 5, 15, 21 and 32 days and 8 months, as described by Chomczynski and Sacchi (1987). The S-MAGct complementary deoxyribonucleic acid (cDNA) was amplified from the total RNA samples by reverse transcription polymerase chain reaction (RT-PCR) and digested with the Sma I endonuclease that distinguishes between the cDNAs encoding the rat S-MAG subtypes. After agarose gel electrophoresis, densitometry was used to determine the relative amounts of the two S-MAG transcripts in the samples. As a control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was analysed in a similar fashion from the same samples and used for correction. Additionally, a subset of the samples was analysed in a similar manner, but using 32P-labelled oligonucleotide primers and autoradiographic detection.

For quality control purposes, a sensitive method was developed for the estimation of total RNA quantity in dilute samples (VII). The method involves the electrophoretic separation of total RNA samples in polyacrylamide gels under nondenaturing conditions, and the visualisation of RNA by silver staining. The RNA silver staining method was optimised from a previously published deoxyribonucleic acid (DNA) silver staining method (Li & Ownby 1993). The correlation between the total RNA amount and mRNA concentration was studied by comparing the staining of different ribosomal RNA species and the GAPDH mRNA level determined by RT-PCR from the same samples.

4.1.2. Cloning of cDNA (I-IV)

For cloning the MAG cytoplasmic domain cDNAs, first-strand cDNA (a gift from B. Garbay, Bordeaux, France) prepared from sciatic nerve total RNA from 5-to-15-day-old rats was used. The cDNAs coding for either the full-length or truncated S- and L-MAG cytoplasmic domains were amplified by RT-PCR using gene-specific primers. The cDNAs were subcloned into the pGEX-4T-1 vector (Pharmacia, Uppsala, Sweden) for subsequent recombinant glutathione S-transferase (GST) fusion protein expression, and the correctness of the inserts was verified by dideoxy sequencing (Sanger et al. 1977).

4.2. Protein analysis methods

4.2.1. Proteins, antibodies, and peptides (I-V)

The GST-MAGct fusion proteins (table 2) and GST-S100β (Nishikawa et al. 1997) were produced in Escherichia coli (E. coli) BL21, and purified by affinity chromatography (Smith & Johnson 1988) on glutathione sepharose 4B (Pharmacia). When necessary, the GST portion of the fusion protein was removed by thrombin digestion. The fusion proteins were labelled with 125I (Amersham, Buckinghamshire, UK) by the chloramine-T method (Greenwood et al. 1963) to a specific activity of approximately 30 µCi/µg.

51

Table 2. The GST-MAGct fusion proteins used in this study.________________________________________________________________________Name Amino acid range Molecular weight

(Salzer et al. 1987) (kDa)________________________________________________________________________GST-S-MAG-R579 534-582 of S-MAG 32(also referred to as (the R579 isoform ofGST-S-MAGct) rat S-MAG)

GST-S-MAG-P579 534-582 of S-MAG 32(the P579 isoform ofrat S-MAG)

GST-L-MAGct 534-626 of L-MAG 37

GST-L-MAG-∆585 534-585 of L-MAG 33

GST-C-MAG 534-572 of MAG, plus 32the C-terminal sequenceGNDPPRTATP

________________________________________________________________________

Biotinylation of the GST-MAGct fusion proteins was performed using biotinamidocaproate N-hydroxysuccinimide ester according to established protocols (Harlow & Lane 1988).

The GST-S-MAGct and GST-L-MAGct proteins were used to immunise white Russian rabbits. The purified IgG fractions of the resulting antibodies were used in immunocytochemistry.

Monoclonal antibodies towards S100β and β-tubulin, polyclonal antibodies against tubulin, and fluorescently labelled secondary antibodies were from Sigma. Other secondary antibodies, the ABComplex, and the polyclonal S100 antibody were from Dako (Copenhagen, Denmark).

Synthetic peptides corresponding to regions in the rat L-MAGct were purchased from Research Genetics. These included SEKR-15 (SEKRLGSERRLLGLR; L-MAG amino acids 573-587), SEKR-30 (SEKRLGSERRLLGLRGEPPELNLSYSHSNL; L-MAG amino acids 573-602), and LGSE-26 (LGSERRLLGLRGEPPELNLSYSHSNL; L-MAG amino acids 577-602). It should be noted that the two asparagines in the latter two peptides correspond to aspartic acid residues in the rat protein.

4.2.2. Protein interactions and function

4.2.2.1. Zinc binding assays (IV)

Several different methods were used to study the binding of zinc by the S-MAGct. The aggregation of the GST-S-MAGct fusion protein was followed by differential centrifugation as a function of zinc concentration. The polymerisation of the protein was also followed by covalent crosslinking with disuccinimidyl suberate (DSS) at different zinc concentrations. Immobilised metal ion affinity chromatography (Porath et al. 1975) with immobilised Zn2+ was used to confirm the presence of surface residues able to coordinate zinc.

Quantitative zinc-binding studies were performed by two independent methods. First, a colorimetric assay based on the free zinc-induced changes in the absorption spectrum of oxidised hematoxylin was developed to estimate the number of zinc atoms bound per one GST-S-MAGct fusion protein molecule. The dissociation constant and stoichiometry for the binding of Zn by the S-MAGct were additionally determined using radioactive 65Zn, by separating the unbound 65Zn from the bound fraction by centrifugal ultrafiltration. The radioactivities were determined by scintillation counting.

The specificity of the S-MAGct towards zinc was determined using 65Zn. The S-MAG cytoplasmic domain (10 µM) was incubated with a constant (20 µM) concentration of 65Zn, in the presence of different concentrations (20 µM - 2 mM) of ZnCl2, CaCl2, MgCl2, and CoCl2, in Hepes-buffered saline (HBS; 10 mM Hepes (N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid)), 150 mM NaCl, pH 7.5). After incubation for 30 min at room temperature, the samples were subjected to centrifugal ultrafiltration through a 3 kDa molecular weight cutoff filter. The radioactivity of the flowthrough was determined by scintillation counting. The amount of 65Zn bound to the

52

protein was calculated by subtracting the radioactivity of the flowthrough from the radioactivity of the original sample.

Phenyl sepharose chromatography of the S-MAGct was performed in the presence and absence of 20 µM or 2 mM ZnCl2, in order to detect changes in the surface properties of the S-MAGct induced by zinc.

E. coli BL21 cells were transformed with the pGEX-4T-1 and pGEX-S-MAG plasmids. Isolated colonies were picked, and the bacteria were grown at +37°C for 2 h. Each culture was then divided in half; isopropyl-β-D-thiogalactoside was added to one of the duplicate cultures to 0.5 mM, and the induction was carried out at +37°C for 1 h. Each culture was again divided into several equal portions, zinc chloride was added to final concentrations ranging from 0 to 10 mM, and the cultures were continued for 1 h. Each culture was then diluted and plated. After an overnight culture, colonies were counted. Statistical analyses of the data were performed using Student’s t-test.

4.2.2.2. Ligand blotting and immunoblotting (III, V)

The use of human postmortem nervous tissue samples in this study was approved by the Finnish National Board of Medicolegal Affairs (Decision no. 102/32/200/99). Human sciatic nerve homogenate or C6 glioma cell lysate proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli 1970) and transferred onto nitrocellulose using a semi-dry blotter (Kem-En-Tec, Copegnhagen, Denmark). The filters were blocked, and the 125I-labelled GST-S-MAGct or GST-L-MAGct fusion protein probe was added to a concentration of 10 ng/ml. As a control, either an excess of nonlabelled GST was added, or a parallel experiment was performed using 125I-GST. After incubation, the filters were washed and exposed to X-ray films.

For immunoblotting analyses, proteins were separated by SDS-PAGE and transferred to either nitrocellulose or polyvinylene difluoride membranes by semi-dry electroblotting. After transfer, the membranes were blocked, incubated with the primary antibodies, washed, and incubated with biotinylated secondary antibodies. Detection was performed using the ABComplex and enhanced chemiluminescent detection.

4.2.2.3. Affinity chromatography and covalent crosslinking (III, V)

For affinity chromatography, GST fusion proteins were immobilised on glutathione sepharose. Then, either purified proteins or peptides, or an aliquot of a cell or a tissue lysate was added, and the mixture was incubated in a buffer suitable for the assay. The glutathione sepharose was collected by centrifugation, washed with the corresponding buffer, and the proteins present in the pellet were studied by SDS-PAGE, with or without glutathione elution.

For the interaction studies by covalent crosslinking, the bovine S100β protein and the SEKR-15 peptide were dissolved in HBS. The crosslinking mixture contained 10 µM S100β and 200 µM SEKR-15 in HBS, supplemented with either 0.5 mM ethylene

53

diamine tetraacetic acid (EDTA), or CaCl2. After a 10-min incubation at room temperature, the homobifunctional crosslinking reagent DSS was added to 0.4 mM. Crosslinking was stopped after 10 min with Laemmli sample buffer, and the samples were analysed by SDS-PAGE and Coomassie staining.

4.2.2.4. Polymerisation of microtubules in vitro (V)

Tubulin was purified from rat brain by performing three consecutive cycles of temperature-sensitive polymerisation-depolymerisation (Wiche & Cole 1976). Bovine brain tubulin, isolated in a similar manner, was purchased from Sigma.

For the quantitative analysis of tubulin polymerisation, 50 µg of rat brain tubulin alone, or with 1 nmol of GST or GST-S-MAG, was incubated at +37°C for 5 minutes in a total volume of 100 µl. Polymerisation was started by the addition of GTP to 1 mM, and the polymerisation rate was followed by measuring the turbidity at 350 nm at timed intervals.

For the analysis of S-MAG coassembly with microtubules, 50 µg of rat brain tubulin and 20 µg of GST, 25 µg of GST-S-MAG ct, or 10 µg of the S-MAG cytoplasmic domain were incubated at +37°C for 30 min, in a total volume of 500 µl. Then, GTP was added to 0.8 mM, and microtubule polymerisation was carried out at +37°C for 30 min. Taxol was added to 2.5 µg/ml, and incubation was continued at +20°C for 30 min. The microtubule pellets were collected by ultracentrifgation at 100000 g for 1 h at +20°C, dissolved in Laemmli sample buffer, and analysed by SDS-PAGE and Coomassie staining. The gels were also analysed densitometrically, correcting the measured intensity for the molecular weight of each protein, with NIH Image 1.62 (NIH, http://rsb.info.nih.gov/nih-image/) to approximate the molar ratio of tubulin to other proteins and the effect of the added proteins on the amount of polymerised tubulin. The reactions were also carried out using different concentrations of S-MAGct, in order to determine the kinetic parameters of the interaction.

4.2.2.5. L-MAG dimerisation assays

The GST-L-MAGct protein was used to study the autoassociation of the L-MAGct in several different assays.

Gel filtration chromatography was performed at room temperature in tris(hydroxymethyl)aminomethane (Tris) -buffered saline (TBS; 10 mM Tris, 150 mM NaCl, pH 7.8) containing 25 mM CaCl2, using a Pharmacia HPLC apparatus and a Superdex G-200 column (Pharmacia). The amount of sample used for each run was 100 µg. The protein content of the eluate fractions was monitored by a UV detector. Standard curves were constructed using elution volumes of standard proteins.

Ligand blotting assays were performed with the 125I-GST-L-MAGct probe. Different GST-MAGct fusion proteins were immobilised on nitrocellulose, and subjected to a denaturation/renaturation cycle, using guanidine hydrochloride (Vinson et al. 1988). The

54

filters were blocked in 5 % non-fat dry milk in HBS for 30 min at +4°C, and nonlabelled GST was added to a concentration of 1 µg/ml. After 30 min at +4°C, the 125I-labelled GST-L-MAGct was added to 10 ng/ml, and incubation was continued at +4°C for 1 h. After washing thoroughly with HBS, the filter was air-dried and exposed to X-ray film at -80°C.

Ultracentrifugation was used to study the aggregation state of the GST-L-MAGct. Ten µg of GST or GST-L-MAGct were incubated in TBS containing 10 % sucrose, and subjected to centrifugation at 150000 g for 90 min. The top of the supernatant fraction and the pellet were analysed by SDS-PAGE and Coomassie staining.

DSS crosslinking was also employed on the GST-MAGct fusion proteins. Briefly, 1 µg of protein was incubated in HBS, and subjected to chemical crosslinking by the addition of DSS to 0.4 mM. After 10 min at room temperature, the reactions were stopped by the addition of Laemmli sample buffer, and the samples were analysed by SDS-PAGE and silver staining.

4.2.2.6. Protein phosphorylation in vitro (I, VI)

GST, GST-S-MAGct, or GST-L-MAGct were phosphorylated by PKA in vitro, in the presence and absence of a 10-fold excess (5 µM) of bovine S100β. The radioactivity of the GST fusion proteins was determined by scintillation counting after purification by glutathione affinity chromatography.

For phosphoamino acid analysis of PKA-phosphorylated proteins, the proteins were bound to pieces of polyvinylene difluoride membrane, and the bound proteins were hydrolyzed with 250 µl of 6 M HCl at +110°C for 1 h. The hydrolysates were analysed by one- (pH 1.9), or two-dimensional (pH 1.9 and pH 3.5) thin-layer electrophoresis on cellulose, followed by autoradiography. Nonradioactive phosphoamino acid standards were co-electrophoresed with the samples and visualised by ninhydrin staining.

In the PKC phosphorylation reactions, the concentrations of GST and the GST-MAGct fusion proteins were 1 µM and, when present, the concentration of bovine brain S100β was 1.6 µM. The reactions were carried out for 40 min at +37°C and terminated by adding Laemmli sample buffer. The products were separated by SDS-PAGE, and the gels were coomassie-stained and exposed to X-ray films. Densitometry of the autoradiograms was performed with NIH Image 1.62 software.

GST, GST-S-MAGct, and GST-L-MAGct were phosphorylated by casein kinase 2 (CK2) (Promega, Madison, WI, USA), and the reaction was allowed to proceed at +37°C for 10 or 60 min. The reaction was stopped by adding Laemmli sample buffer, and the products were analysed by SDS-PAGE, autoradiography, and densitometry.

55

4.2.3. Structural analyses

4.2.3.1. Primary sequence analyses (III, V)

Protein sequence pattern searches and concensus secondary structure predictions were performed on the Network Protein Sequence Analysis server at http://www.ibcp.fr. Sequence alignments were done either manually, or using ClustalX (Thompson et al. 1997). The tau/MAP microtubule-binding repeats and flanking domains were retrieved from the ProDom database (Corpet et al. 1998).

A Macintosh HyperCard stack that allows the detection of homologous regions in a protein to a library of peptide sequences was developed. The algorithm was based on the PAM250 protein alignment matrix, and peaks are observed at areas that share most homology with the sequences in the peptide library. The program was used to identify potential S100β-binding sites in the MAG cytoplasmic domains by using a previously characterised set of S100β-binding peptides from a phage display screen (Ivanenkov et al. 1995) as the library.

4.2.3.2. Circular dichroism spectroscopy (V)

Circular dichroism (CD) spectroscopy was used to study the folding state and secondary structure content of the isolated S-MAGct. Assays were performed at a protein concentration of 0.5 mg/ml in 50 mM potassium phosphate (pH 7), in the presence or absence of 20 µM zinc chloride. Wavelengths 195-260 nm were scanned at room temperature using the Jasco J-715 spectropolarimeter (Jasco, Tokyo, Japan). Similarly, the peptides SEKR-15, SEKR-30, and LGSE-26 from the L-MAG-specific region of the cytoplasmic domain were analysed, at a concentration of 2 mg/ml.

The secondary structure contents of the S-MAGct and the L-MAGct-specific peptides were estimated from the measured data with the least squares method, by using the standard CD spectra for polylysine in the three conformational states: helix, sheet, and coil (Greenfield & Fasman 1969). A computer program for the analysis was written for Macintosh computers that fits the CD spectrum of a protein or peptide to these three states. The program gave correct results when tested on combinations of theoretical spectra (unpublished data), and the results of the fitting for the analysed samples were as expected from secondary structure predictions. The data for the S-MAGct were fitted in the wavelength region 195-250 nm, and the data for the peptides were fitted between 210-250 nm, due to interference at and below 200 nm caused by an unknown compound in the peptide preparations, possibly sodium.

56

4.2.3.3. X-ray crystallography

The GST-S-MAGct protein was produced in E. coli BL21 and purified by glutathione affinity chromatography. To increase the yield, the protein was eluted from glutathione sepharose by using 50 mM triethanolamine, pH 10.8. The buffer was exchanged and the protein was concentrated in a 30 kDa molecular weight cutoff centrifugal ultrafiltration tube. The initial protein concentration used in the crystallisation experiments was 5-50 mg/ml.

The crystallisation conditions for the GST-S-MAGct and the S-MAGct were screened using incomplete factorial screens (Zeelen et al. 1994) and a Gilson Abimed 222 XL liquid handler. Crystals were grown in hanging drops containing 2 µl of protein and 2 µl of the well solution. After the detection of crystals in the factorial screens, the crystallisation conditions were optimised with respect to pH and precipitant concentration.

Data collection from GST-S-MAGct crystals was performed either at the Department of Biochemistry, University of Oulu, or at the DESY synchrotron facility in Hamburg, on the X31 beamline using a MAR345 detector (Mar Research). For the structural analysis, the synchrotron data were processed using programs DENZO and SCALEPACK of the HKL package (Otwinowski & Minor 1997). Molecular replacement, using three structures for GST from the PDB database (1GTA, McTigue et al. 1995; 1BG5, Zhang et al. 1998; 1GNE, Lim et al. 1994), was done using the CCP4 suite (Collaborative Computational Project Number 4 1994) program AMoRe (Navaza 1994). Further refinement by simulated annealing, as well as energy minimisation, was performed using the CNS program (Rice & Brünger 1994, Brünger et al. 1998), which was also used for calculating the electron density maps. Model building was performed on the basis of the calculated maps using the program O (Jones et al. 1991).

4.3. Cell culture and immunocytochemistry (V)

The immortalised Schwann cell line S16, originally produced by repetitive passaging of rat Schwann cells (Goda et al. 1991), was used to determine the subcellular localisation of MAG, and any changes in the latter produced by microtubule-affecting drugs. As a control, the S16Y cell line (Toda et al. 1994), reported not to produce detectable levels of MAG, was used. The cells were routinely maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10 % fetal calf serum (FCS), at +37°C in a humidified atmosphere containing 5 % CO2. For the immunostaining experiments, the cells were cultured on glass coverslips in N2 differentiation medium (1:1 DMEM and Ham’s F-12 Medium, with 1:100 dilution of N2 supplement (Gibco)) for 1-3 days. When necessary, the cells were treated with taxol or colchicine prior to fixation.

The cells were fixed, and blocked in 10 % BSA. The primary antibodies were added at a final dilution of 1:100-1:200 in 5 % BSA in phosphate-buffered saline (PBS) and, after incubation, the cells were washed with PBS, and the secondary antibodies were added in 5 % BSA in PBS. Control experiments using secondary antibodies alone were performed at the same time under identical conditions. After washing, the coverslips were mounted

57

in Immu-Mount (Shandon, Pittsburgh, PA, USA), and viewed using a Zeiss Axiovert 405 M fluorescence microscope. The cells were photographed using a low-light-level Extended Isis video camera (Photonic Sciences, Mountford, UK). The images were digitised using a DT3851 frame grabber (Data Translation, Marlboro, MA, USA) and the Global Lab Image program (Data Translation). The grayscale images from double-stained samples were colorised and superimposed using Adobe Photoshop.

For coprecipitation experiments, the rapidly-growing C6 glioma cell line (Benda et al. 1968) was used. The C6 cells were maintained in 10 % FCS in DMEM, and prior to the experiment, they were grown for 2 days in DMEM supplemented with 1 mM dibutyryl cAMP. The cells were lysed in 20 mM Hepes (pH 7.5), supplemented with 1 % Triton X-100, and precleared by centrifugation.

58

5. Results

5.1. Analyses of primary structure

The primary structures of the MAG cytoplasmic domains were analysed in detail in order to gain information on potential structural and functional features, and differences therein between the MAG isoforms. Secondary structure predictions were performed using several different algorithms, and theoretical physicochemical properties for the different parts of the MAG cytoplasmic domains were analysed (table 2). In addition to MAG, the sequence of the cytoplasmic domain of the avian SMP protein was analysed similarly, as

RRKKNVTESPSFSAGDNPHVLYSPEFRISGAPDKYESREVSTRDCH-582

RRKKNVTESPSFSAGDNPHV--LYSPEFRISGAPDKYE

RRKKGAG-SPEVT-PVQPMAGPGGDPDLDLRPQQVRW-|||| : || :: :| : |:::: ::

SEKRLGSERRLLGLR---GEPPELDLSYSHSDLGKRPTKDSYTLTEELAEYAEIRVK-626

S-MAG

L-MAG

SMP

--LRGAMERWALGVKEGSGAPQEVTPT-SHPPM--KPTRGPL---EDPPEYAEIRVK-620 | || :||:: | | |: : || : :||: |: ||||||||

Fig. 6. Comparison between the MAG cytoplasmic domains and the corresponding region in SMP. Typical results from secondary structure predictions with different algorithms are shown. Arrows indicate predicted β-strands, and boxes denote putative α-helices. The L-MAG and SMP sequences have been aligned taking into account secondary structure predictions and putative functional sites. Potential phosphorylation sites in MAG are also indicated: underlined, PKC; double underlined, PKA; bold, Tyr620 known to be phosphorylated by Fyn kinase.

controversy exists on whether or not SMP could be the avian homologue of MAG. Protein sequence databases were also used to find sequences homologous to that of the MAGct, but no clear homologies were detected (data not shown).

Secondary structure predictions indicated clear differences between the MAG isoforms (Fig. 6). The portion of the MAGct common to both isoforms is predicted to be mostly without secondary structure, with a putative β-hairpin in the middle region. No organised secondary structure was predicted to be present in the isoform-specific carboxy-terminal portion of S-MAG. On the other hand, all the methods used predicted α-helices both at the beginning and the end of the L-MAG-specific sequence, and a β-strand between these helices. Despite the relatively low sequence similarity between SMP and L-MAG, the results of the prediction for SMP closely resembled those for the L-MAGct. When these predicted structures and the putative functional sites observed in this study (see below) were taken into account, an improved sequence alignment between L-MAG and SMP could be constructed, showing much higher homology between these two proteins than originally reported (Dulac et al. 1992).

60

Table 2. Predicted isoelectric points of different portions of the MAG cytoplasmic domain. The corresponding regions of SMP were included in the analysis, and the homology in this case is based on an improved alignment taking into account the functional properties of MAG encountered in this study, and the predicted secondary structure elements. The level of sequence identity between MAG and SMP based on the original alignment of (Dulac et al. 1992) is given in parentheses. No isoform of SMP corresponding to S-MAG has been reported. The first four amino acids (537-540) of the cytoplasmic domain were not included in the analysis, since this stretch of basic residues is most likely merely anchoring the protein to the membrane, according to the "positive inside" rule (von Heijne 1989).________________________________________________________________________domain species pI % identity with rat MAG________________________________________________________________________common (amino acids 541-572) rat 4.38 100

mouse 4.38 94human 4.01 84quail (SMP) 3.99 12.5 (6)

S-MAG-specific (amino acids 573-582) rat (R579) 7.08 100

rat (P579) 5.41 90mouse 7.08 100human 5.52 60

L-MAG-specific (amino acids 573-626) rat 5.88 100

mouse 5.88 98human 5.88 98quail (SMP) 7.16 39 (22)

________________________________________________________________________

5.2. Production of recombinant proteins and antibodies (I-IV)

As a first step in the analysis of the functions of the MAGcts, these domains were produced as recombinant proteins in E. coli. After optimising the production conditions, both GST-S-MAGct and GST-L-MAGct could be obtained pure at levels permitting protein interaction studies in vitro . The GST-MAGct fusion proteins were used to raise polyclonal antibodies in rabbits.

The GST fusion proteins were labelled with 125I or biotin, and they were used for in vitro functional studies, including binding studies, in vitro phosphorylation, and cell overlay experiments. The antibodies against the entire L-MAGct and only the L-MAG-specific domain worked in a variety of applications, including blotting, immunocytochemistry, and immunoprecipitation.

5.3. The amino acid dimorphism of the rat S-MAG (I)

Sequence analysis of the GST-S-MAGct expression constructs revealed the possibility of a nucleotide dimorphism in the S-MAG-specific exon 12 coding sequence. A simple RT-PCR-restriction fragment length polymorphism (RFLP) method was developed to study the presence and expression of the S-MAG dimorphism in rat sciatic nerves. From the total of 29 samples analysed, 11 individuals had only the 100-bp product, 14 had both the 100- and the 127-bp products, and 4 had only the 127-bp product. The fact that the observed RFLP resulted from the expected nucleotide dimorphism in exon 12 was confirmed by subcloning and sequencing the RT-PCR products from rat sciatic nerve RNA samples.

The total expression of S-MAG mRNA in the rat sciatic nerves relative to the amount of GAPDH expression was determined at different ages by RT-PCR-RFLP and densitometry. The relative and total expression levels of Sma I-sensitive (S-MAG-R579) and Sma I-resistant mRNA (S-MAG-P579) were also determined using a 32P-end-labelled primer. The expression of total S-MAG mRNA relative to the expression of GAPDH, irrespective of the S-MAG subtypes expressed by the individuals, was highest at the ages of 5-15 days, and declined thereafter. In animals expressing both mRNA subtypes, the total expression of the S-MAG-P579 mRNA was significantly higher (three- to four-fold) than that of S-MAG-R579 during the first two weeks.

The cytoplasmic domains of both S-MAG subtypes were expressed as GST fusion proteins. As the induction time was prolonged, the GST-S-MAG-P579 fusion protein underwent increasing degradation, while the majority of the GST-S-MAG-R579 fusion protein remained intact, regardless of the induction time. The degradation was the result of proteolytic cleavages in the carboxy-terminal portion of the S-MAG protein.

Using the same samples, an electrophoretic method was developed for the estimation of total RNA quantity in dilute samples (VII). The linear range for the assay, slightly depending on the RNA species used in the measurements, is between 10 and 1000 ng.

61

5.4. Characterisation of S-MAG as a zinc-binding protein (IV)

During protein interaction studies under various conditions, the GST-S-MAGct protein behaved differently from the other proteins used. This phenomenon could be attributed to the presence of zinc cations, and thus, the ability of the S-MAGct to bind zinc was investigated.

The aggregation of the fusion proteins was studied by incubating the proteins with increasing concentrations of zinc chloride. While the GST-S-MAGct protein was lost from solution in a zinc concentration-dependent manner, the GST-L-MAGct and the GST-C-MAGct proteins were unaffected by the zinc concentrations employed. These data indicate that the loss of GST-S-MAGct from solution in the presence of zinc is mediated by the 10 carboxy-terminal amino acids of the protein. Covalent crosslinking with the homobifunctional reagent DSS indicated that increasing concentrations of zinc chloride induced the polymerisation of the GST-S-MAGct, but not the GST-L-MAGct, fusion protein. The results indicate that the polymerisation is dependent on the presence of the isoform-specific sequence of the S-MAGct.

Comparison of the chromatographic behaviour of the GST and GST-S-MAGct proteins on a Zn-loaded chelating sepharose column showed that the GST-S-MAGct, but not GST alone, was retained in the column and eluted with EDTA, indicating that the S-MAGct is able to reversibly bind to the Zn-containing matrix.

A spectrophotometric assay based on the zinc-induced change in the absorption spectrum of oxidised hematoxylin was developed to quantitate the level of free zinc in the binding assays. The presence of the GST-S-MAGct, but not GST alone, reduced the zinc-induced absorbance increase of hematoxylin at 600 nm in a concentration-dependent manner. The results indicate that approximately one zinc cation is bound per one S-MAGct.

In order to determine the dissociation constant for the zinc-S-MAG interaction, the S-MAGct was cleaved from the fusion protein and used for zinc-binding assays. The results obtained with the hematoxylin assay indicate a maximum of 0.6 zinc ions bound per one molecule of S-MAG, and yield a Kd value of 7.1 µM. The binding of zinc by the S-MAGct was also studied using the radioactive 65Zn isotope and ultrafiltration. The results were similar to those obtained using the hematoxylin method, the number of zinc ions bound per S-MAG protein being 0.6 and the Kd being 5.8 µM.

The binding specificity of the S-MAGct for zinc was studied by incubating the S-MAGct with a constant amount of 65Zn, in the presence of a 1-100 -fold excess of either Ca2+, Mg2+, Zn2+, or Co2+. As expected, nonradioactive zinc effectively displaced the 65Zn from the protein. Cobalt ions were also able to displace zinc from the S-MAGct, but less efficiently. Very weak displacement was observed with Mg2+, while Ca2+ had no detectable effect.

The hydrophobic properties of the purified S-MAGct in the presence and absence of zinc were studied by chromatography on phenyl sepharose. The S-MAGct quantitatively bound to phenyl sepharose in the presence of zinc cations, and it could be eluted from the matrix with EDTA. In the absence of zinc, no binding of the S-MAGct to the column was observed.

CD spectroscopic analysis (V) of the S-MAGct in the presence and absence of zinc revealed that the binding of zinc had no detectable effect on the secondary structures of

62

the domain. From the results, it could be calculated that the S-MAGct is mostly an unordered polypeptide in solution. A small amount of secondary structure, calculated to be 20-30 % of helical structure is also present. Secondary structure predictions place a single short β-hairpin in the S-MAGct, with the rest being predicted as random coil (Fig. 6).

5.5. S-MAG is a microtubule-associated protein (V)

5.5.1. S-MAG binds directly to tubulin

Initially, a 50-kDa protein was found to bind to the S-MAGct in ligand blotting and affinity coprecipitation assays. Affinity chromatographic and ligand blotting analyses using pure tubulin confirmed the tubulin-S-MAGct interaction. The observed interaction was strictly dependent on the presence of the 10-amino acid S-MAG-specific carboxy terminus.

The effect of the presence of the S-MAGct on temperature-dependent microtubule assembly was studied by turbidimetry using the GST-MAGct fusion proteins and tubulin purified from rat brain. Although the total amount of tubulin polymerised at equilibrium was not affected in this assay, the presence of the S-MAGct decreased the initial rate of microtubule assembly.

When microtubules were assembled in vitro and stabilised with taxol, the GST-S-MAGct fusion protein and the S-MAGct were both found in the microtubule pellet, while the GST protein was not. The total amount of polymerised, taxol-stabilised tubulin was similar in all samples. SDS-PAGE followed by densitometric analysis of the Coomassie-stained gels revealed that, in the microtubule fraction, the molar ratio of S-MAGct to tubulin was 1.1:1.

A preliminary analysis of the kinetics of the S-MAGct-tubulin interaction was carried out by in vitro assembly and taxol stabilisation of microtubules in the presence of different amounts of S-MAGct. The levels of S-MAGct in the microtubule pellet were determined, and Kd and Bmax were calculated (Fig. 7).

5.5.2. Cell culture studies

When grown in the differentiation-inducing N2 medium, a fraction of the S16 cells starts to differentiate, and they grow very large in area, adopting a flat morphology. At the same time, these cells start to express S-MAG and other myelin proteins (Sasagasako et al. 1996). The morphology of the differentiating S16 cells resembles the flat appearance of Schwann cells isolated from adult sciatic nerve (Kidd et al. 1996), while the small bipolar S16 cells have a morphology reminiscent of neonatal Schwann cells (Kidd et al. 1996). After 3 days in the N2 medium, most of the S-MAG -expressing cells grew on top of each other, forming a large network-like structure. On the other hand, cells not present

63

in these structures retained their bipolar morphology and did not express MAG. S-MAG was detected evenly distributed all over the differentiating S16 cell

membrane. In addition, strong punctate staining was frequently observed in the cell perinuclear region, at cell-cell contact sites between two MAG-positive cells, at ruffled membranes, and in small processes.

Double staining for tubulin and S-MAG revealed several interesting features. The punctate staining for S-MAG often found in the perinuclear region was apparently ordered along microtubular structures. Most processes emitted by the S16 cells stained strongly for microtubules, and a partial colocalisation of S-MAG and tubulin was usually evident at the growing tip. Indeed, the subcellular regions rich in S-MAG almost always contain microtubules.

Taxol, a microtubule-stabilising agent, and colchicine, which destabilises microtubules, were employed to further characterise the S-MAG-tubulin interaction in cultured S16 cells. The more or less severe changes in microtubule staining and cell morphology were frequently associated with changes in the localisation of S-MAG. The uniform staining of the plasma membrane was conserved in the presence of taxol and colchicine, but both of these drugs also caused an increase in the amount of punctate staining for S-MAG around the nucleus. Furthermore, taxol induced the growth of small S-MAG-containing processes.

Biotinylated GST-MAGct fusion proteins and fluorescein-conjugated streptavidin were employed to determine the subcellular distribution of S16 cell components binding to the MAGcts in a cell overlay experiment. The L-MAGct-binding activity was

64

0

0.1

0.2

0.3

0 1 2 3 4 5

0

0.05

0.1

0 0.5 1

A B

[S-MAGct] (µM) B (µM)

Fig. 7. Preliminary binding analysis between S-MAGct and tubulin. A. The binding curve for S-MAGct to taxol-stabilised microtubules. B. A Scatchard analysis of the data shown in A. The intercept between the extrapolated line and the x-axis gives a Bmax of 0.86 µM, and from the slope of the line, a Kd of 7 µM can be determined. The concentration of tubulin in the assay was 0.8 µM, and thus, the Bmax can also be expressed as 1.1 molecules of S-MAGct bound per tubulin monomer. B, S-MAGct bound to the microtubule pellet; B/F, ratio of bound and free S-MAGct.

localised in the cell perinuclear region. S-MAGct-binding activity, on the other hand, was clearly concentrated in thin processes. The staining of cells with a similar morphology for β-tubulin indicated that these processes are rich in microtubules, and thus, the S-MAGct-binding components apparently colocalise with tubulin.

5.6. Crystallisation of the S-MAG cytoplasmic domain

Both the GST-S-MAGct fusion protein and the purified S-MAGct were used in crystallisation experiments. In initial incomplete factorial screens (Zeelen et al. 1994), specific conditions leading to crystal growth were observed. Optimisation screens were then performed by varying the conditions, such as pH and the precipitant concentration. The conditions found to give the biggest crystals of GST-S-MAGct were 2 M ammonium sulfate, pH 5.5, at +22°C (Fig. 8).

One of the GST-S-MAGct crystals was analysed by X-ray crystallography, using data collected at the EMBL-DESY synchrotron in Hamburg. The resolution limit used for data processing was 2.6 Å. The unit cell dimensions of the crystal were 92.525 x 92.525 x 57.665 Å, and there was one molecule per asymmetric unit. The space group of the crystal was P43212. The first 213 amino acids from the previously determined structures for GST and two GST fusion proteins were used as templates in molecular replacement and electron density map calculation. After model building and refinement, the R factor was 29.7 % and R(free) 35.9 %. Very little electron density could be observed at the area where the S-MAGct was supposed to lie, while the GST portion of the fusion protein was well defined in the map (Fig. 9). This suggests that either only a degradation product of the fusion protein had been crystallised, or the S-MAGct is flexible even within the crystal, or the strategy employed in the molecular replacement was not sufficient for

65

Fig. 8. Two different kinds of crystals were grown under the same conditions from the GST-S-MAGct. The diamond-shaped crystals (arrow) were used for crystallography experiments, while the crystals with different morphology (arrowheads) did not diffract X-rays. Due to phase separation in the crystallisation drop, the picture could not be better focused. The longest dimension of the crystals in this view is 0.3-0.4 mm.

66

A

B

Fig. 9. A. A diffraction image for a GST-S-MAGct crystal. Diffraction can be detected down to 2.8 Å (second outermost ring). B. Failure to detect significant electron density for the S-MAGct portion of the GST-S-MAGct fusion protein. While the first 213 residues of GST were well-defined, electron density was very weak for the linker region and the S-MAGct. Asp213 is the last residue clearly seen in the electron density map. The map was contoured at 1 σ.

determining the structure of a fusion protein. When the drops containing these crystals were analysed by SDS-PAGE several months after the drops were made and the crystals had formed, only a very small amount of degradation was detectable (data not shown), indicating that it is unlikely that the crystals, formed in 1-5 days, would consist of only GST.

Crystals were also obtained from a preparation of pure S-MAGct, and the size of these crystals could be increased by the addition of zinc chloride. However, even the biggest S-MAGct crystals obtained were too small for X-ray diffraction analyses.

5.7. The interaction between S100β and L-MAG (III)

Ligand blotting experiments, performed using 125I-GST-L-MAGct as the probe, revealed a strong interaction with a protein of about 10 kDa in a C6 cell lysate and a cleared human sciatic nerve homogenate. No signals were detected under identical conditions when using the 125I-GST-S-MAGct as the probe. The detection of the 10-kDa band with the 125I-GST-L-MAGct probe required the presence of Ca2+. The above results suggested that the 10-kDa ligand could be the S100β protein. Thus, one filter was reprobed with an anti-S100 antibody, and an identical signal was observed as with the GST-L-MAGct probe.

The GST and GST-L-MAGct proteins were immobilised on glutathione sepharose and their ability to bind bovine S100β was tested. While the S100β was not copurified with the GST protein, it was present in the affinity-purified GST-L-MAGct sample. Human brain S100 proteins (a mixture of α and β isoforms) and bovine brain S100β were immobilised on nitrocellulose, and their ability to bind the radiolabelled GST-L-MAGct was studied by ligand blotting. The iodinated GST-L-MAGct bound to both S100 protein preparations, while the GST alone did not.

The binding of the S100β protein to the region at the beginning of the L-MAG-specific sequence was confirmed by covalent crosslinking of the S100β protein and the SEKR-15 peptide. A crosslinking product of the expected size, 12 kDa, could be detected in the presence of Ca2+. The binding affinity and stoichiometry were studied by following the binding of the SEKR-15 peptide by immobilised GST-S100β, using GST as a control. The amount of bound peptide increased with the total peptide concentration, reaching a plateau when approximately one mole of peptide was bound per one mole of S100β. Scatchard analysis of the experimental data yielded values for Bmax (1.2 mol SEKR-15 bound / mol S100β) and Kd (7 µM).

By L-MAG cytoplasmic domain sequence homology searches, several protein domains were found that were known to interact with S100 proteins. The region of strongest homology between these proteins and MAG was at the beginning of the L-MAG-specific sequence, and the highest degree of homology was found at the S100β-binding site of neuromodulin. Secondary structure predictions suggested that this segment of L-MAG and the homologous regions in the previously identified S100 ligands are likely to have a similar secondary structure, a basic amphipathic α-helix. The S100β-binding peptides characterised earlier (Ivanenkov et al. 1995) were aligned with

67

the MAG cytoplasmic domains. The highest scores, when compared to a random set of peptides analysed similarly, were observed at the beginning of the L-MAG-specific sequence.

The tendency of the SEKR-15 peptide region to fold into a helix was analysed by CD spectroscopy. Fitting of the results on theoretically calculated secondary structure contents is shown in table 3. The results indicate that the SEKR-15 peptide can adopt a helical conformation in solution. A similar content of helical character was observed in a longer peptide, SEKR-30, while the helicity was lost in the LGSE-26 peptide, which is identical to the SEKR-30 peptide but lacks the first four N-terminal residues (SEKR).

5.8. Phosphorylation of recombinant MAG in vitro (I, VI)

Significant differences were observed between the PKA phosphorylation levels of the different GST-MAGct fusion proteins. While the GST-S-MAGct was phosphorylated to low levels, the GST-L-MAGct was more strongly phosphorylated, with 5- to 10-fold more radioactivity incorporated during the assay time. GST alone was not detectably phosphorylated. Thus, there is at least one PKA phosphorylation site in the L-MAG-specific cytoplasmic domain, between amino acids 573 and 626. The major PKA-phosphorylated residue was phosphoserine in both GST-S-MAGct and GST-L-MAGct. Phosphothreonine, accounting for about one-third of the phosphorylation, was also detected in GST-L-MAGct.

In the presence of S100β, the phosphorylation of both the full-length and the truncated L-MAG cytoplasmic domains by PKA was decreased. The presence of S100β reduced the levels of phosphoserine in the L-MAGct, while phosphothreonine levels remained unaffected.

Both GST-S-MAGct and GST-L-MAGct were strongly phosphorylated to approximately equal levels by PKC in vitro. The phosphorylation time-courses of the two S-MAG cytoplasmic domains were similar, and the ratio of phosphorylated GST fusion protein to PKC was constant and similar for both S-MAG subtypes.

68

Table 3. Analysis of the presence of helical structure in the region of the SEKR-15 peptide by circular dichroism spectroscopy. The measured CD spectrum was fitted by the least squares method to combinations of the spectra of polylysine in the three major conformations: helix, sheet, and coil.________________________________________________________________________Peptide Best fitted solution

L-MAG (relative amounts in %) amino acids Helix Sheet Coil

________________________________________________________________________SEKR-15 573-587 6.8 18.2 75SEKR-30 573-602 9.5 38.1 52.4LGSE-26 577-602 0 29.1 70.9________________________________________________________________________

Phosphorylation of the GST-MAGct fusion proteins by CK2 was not detectable. On the other hand, while strong time-dependent autophosphorylation of the regulatory CK2β subunit was detected in all samples, the rate of autophosphorylation was highest in the presence of the MAGcts.

5.9. Dimerisation of the L-MAG cytoplasmic domain

The first evidence for the dimerisation of the L-MAGct came from gel filtration studies with the GST-L-MAGct fusion protein (table 4). Furthermore, covalent crosslinking using DSS indicated a quantitative crosslinking of the GST-L-MAGct fusion protein to a tetramer, while no crosslinking was observed for GST or GST-S-MAGct (table 4). Ligand blotting with a radioactively labelled GST-L-MAGct probe indicated binding to immobilised GST-L-MAGct, but not to GST or GST-S-MAGct (Fig. 10A). This interaction was only detected when the nitrocellulose filter went through a denaturation/renaturation cycle using guanidine hydrochloride (not shown). These results confirm the requirement of the L-MAG-specific sequence for dimerisation. Sedimentation analysis by ultracentrifugation (Fig. 10B) showed that, while a fraction of the GST-L-MAGct protein pelleted through a layer of 10 % sucrose, GST alone did not. This result confirms the polymerising property of the fusion protein and provides additional evidence for the dimerisation of the L-MAGct.

69

Table 4. Gel filtration and covalent crosslinking analyses of GST-MAGct proteins.________________________________________________________________________

Gel filtration________________________________________________________________________method sample monomer peak conclusion

(kDa) (kDa)________________________________________________________________________Gel filtration GST 29 56 dimer

GST-S-MAGct 32 64 dimerGST-L-MAGct 37 140 tetramer

________________________________________________________________________DSS crosslinking GST 29 29 monomer& SDS-PAGE GST-S-MAGct 32 32 monomer

GST-L-MAGct 37 80-90 dimer________________________________________________________________________

70

A

B

GST

GST-S-MAGct

GST-L-MAGct

1 2 3 4

46

30

21

0.5 µg 2.5 µg

Fig. 10. Evidence for the dimerisation of the L-MAGct. A. The indicated amounts of GST-MAGct fusion proteins were immobilised on nitrocellulose, and binding of radiolabelled GST-L-MAGct was analysed. B. GST and GST-L-MAGct were centrifuged through 10 % sucrose for 1.5 h at 150000 g. The pellet and top fractions were analysed by SDS-PAGE and Coomassie staining. Lanes 1-2, GST; lanes 3-4, GST-L-MAGct. Lanes 1 and 3, supernatant fraction; lanes 2 and 4, pellet fraction.

6. Discussion

Surprisingly few studies have focused on the potential functional differences of the two MAG isoforms, differing from each other only by their cytoplasmic carboxy terminus. S- and L-MAG are expressed differentially during the period of myelination (Tropak et al. 1988), and gene knockout studies (Li et al. 1994, Montag et al. 1994, Fujita et al. 1998) have indicated different roles for the proteins. Differential phosphorylation for the MAG isoforms has been observed both in vivo and in vitro (Edwards et al. 1989, Afar et al. 1990, Bambrick & Braun 1991, Kirchhoff et al. 1993, Jaramillo et al. 1994, Yim et al. 1995).

The present study, performed to gain information on the properties and function of the MAG cytoplasmic domains, revealed that the L-MAGct interacts with S100β, which regulates its phosphorylation by PKA in vitro, and that the S-MAGct is a zinc-binding protein that could also be characterised as a microtubule-associated protein. The results imply that the S- and L-MAG cytoplasmic domains have important functional differences.

6.1. The cytoplasmic domain of S-MAG

S-MAG has traditionally been treated as the less interesting of the MAG isoforms, mainly based on the inability to assign any specific functions to its cytoplasmic domain. Although S-MAG has been shown to be phosphorylated by PKC (Kirchhoff et al. 1993) and to interact with Fyn (Jaramillo et al. 1994), the significance of these findings has remained unclear. Recently, however, gene knockout studies have indicated that S-MAG has an important, but unidentified, role in peripheral nerve myelin, as the PNS pathology observed in mice deficient in both S- and L-MAG (Li et al. 1994, Montag et al. 1994) is lacking in the L-MAG-deficient animals (Fujita et al. 1998). A vital step in the identification of the function of any protein is the identification of molecules with which it can interact. Prior to this study, no specific ligands for the cytoplasmic domain of S-MAG had been reported. This could be explained either by the unlikely possibility that there were no specific cytoplasmic ligands for this protein, or by the failure to create the correct conditions for the detection of the interactions.

6.1.1. The dimorphism of amino acid 579

In rat brain, a nucleotide dimorphism affecting amino acid 579 of S-MAG was observed by Tropak and coworkers (1988), but was not investigated further. The results obtained in this study demonstrate that the S-MAG [Arg/Pro]579 dimorphism can also be detected in rat peripheral nerves, and that it is of allelic origin. In sciatic nerves of young heterozygous rats, the expression of the S-MAG-P579 allele is significantly higher than that of the S-MAG-R579 allele, and the production and analysis of recombinant proteins corresponding to the two subtypes of S-MAGct indicated that the protein with proline at position 579 was degraded in its S-MAG -specific portion already in the early stages of induction. On the other hand, the phosphorylation of the S-MAG cytoplasmic domain by PKC in vitro was not affected by the dimorphism, indicating that either Ser577 is not a major phosphorylation target of PKC, or its phosphorylation does not require a basic amino acid residue at position +2. The strong upregulation of the S-MAG-P579 mRNA expression during myelin formation in the sciatic nerve could be a mechanism to compensate for accelerated proteolytic degradation, or for impaired function of important regions in the corresponding protein product. That such a mechanism has evolved and can be so clearly seen at a defined age, suggests that the fully functional cytoplasmic domain of S-MAG may play an important role during the early stages of peripheral nerve myelination.

6.1.2. The binding of zinc

Myelin contains high levels of divalent metal cations, including zinc (Berlet et al. 1994), most of which is bound to proteins. The binding of divalent cations, including calcium and zinc, often induces the formation of a hydrophobic face or pocket in the protein that is involved in ligand interaction. This is a common mechanism found, for example, in proteins of the EF hand family, including S100 proteins (Baudier et al. 1982, Baudier & Gerard 1983). Several authors have proposed a role for zinc cations in the compaction of the mature myelin sheath (Inouye & Kirschner 1984, Earl et al. 1988, Riccio et al. 1995, Tsang et al. 1997). Furthermore, MBP has been shown to be a zinc-binding protein (Berlet et al. 1994, Cavatorta et al. 1994, Tsang et al. 1997), and zinc has been shown to promote its aggregation (Cavatorta et al. 1994, Riccio et al. 1995).

In the present study, it is shown that the cytoplasmic domain of S-MAG binds zinc cations. The dissociation constant for the zinc-S-MAGct complex was determined to be 6-7 µM by two independent methods. Dissociation constants in this range for protein-zinc complexes are common. For example, similar dissociation constants have been observed for the binding of zinc by protein kinase C inhibitor-1 (Kd = 4.3 µM, Mozier et al. 1991), and by a 19-kDa protein that regulates 6-phosphofructo-1-kinase (Kd = 6 µM , Brand et al. 1988). The binding of zinc by calmodulin is clearly weaker (Kd = 80-300 µM) than by S-MAG, while the S100 β protein binds zinc somewhat more strongly (Kd = 0.1-1 µM) than S-MAG (Baudier et al. 1983). The affinity of the S-MAGct for zinc ions suggests that the protein is also able to bind zinc under physiological conditions.

Zinc-binding sites in proteins are also able to bind other heavy metal ions, such as

72

Co2+. In this study, the binding of zinc to the S-MAGct was prevented by excess amounts of Co2+. The most likely explanation is that Co2+ binds to the same site, the affinity of Co2+ for the binding site being at least an order of magnitude weaker than that of Zn2+. Ca2+ and Mg2+ were not able to significantly decrease the binding of zinc by the S-MAGct. These data indicate that the binding properties of the Zn2+-binding site of S-MAGct are similar to zinc-binding sites in other proteins, and further suggest a physiological relevance for the binding of zinc by the S-MAGct.

Originally, zinc-binding domains in proteins were found to be involved in DNA binding and enzymatic catalysis. More recently, zinc-binding proteins have been described, where the binding of zinc facilitates conformational changes that are required for protein-protein interactions (Mackay & Crossley 1998). This has lead to the hypothesis that folding of the polypeptide chain around a zinc atom, a strong Lewis acid, is a good means to form a small, compact protein domain under reducing conditions, such as those encountered in the cytoplasm (Mackay & Crossley 1998). Hydrophobic interaction chromatography indicated that the binding of zinc by the S-MAGct induced a conformational change that increases the surface hydrophobicity of this domain. On the other hand, CD spectroscopy indicated that no detectable changes occur in the secondary structure profile of the S-MAGct upon Zn-binding, indicating that the conformational change induced by zinc in the S-MAGct may be restricted to amino acid side chains, and to regions of the S-MAGct that are devoid of an organised secondary structure. Whether the zinc-induced conformational change in the S-MAGct affects the affinity of this domain towards its ligands, remains to be determined.

6.1.3. Association with tubulin

Based on the colocalisation of MAG with cytoskeletal proteins, such as actin and spectrin (Trapp et al. 1989a), it has been postulated that MAG interacts with the glial cell cytoskeleton during myelin formation and maintenance. In this study, the first isoform-specific cytoplasmic protein ligand for S-MAG is identified as tubulin, the core component of cytoplasmic microtubules. The characterisation of S-MAG as a microtubule-associated protein places it in a dynamic structural complex linking the axon and the myelinating glial cell cytoskeleton.

Microtubules - polar themselves by nature - are known to be important in both the generation and maintenance of cell polarity (reviewed by Drewes et al. 1998). The polarisation of a cell involves the polar rearrangement of microtubules, the transport of molecules and organelles along microtubules, and the stabilisation of microtubules by anchor points provided by microtubule-associated proteins. Myelinating glial cells are polarised in a fashion reminiscent of epithelial cells, with their apical surface facing the neuron, and the basolateral surface facing the surrounding basal lamina. In myelinating glial cells, microtubules in the perinuclear cytoplasm organise synthetic organelles and initiate microtubule-based transport pathways (Kidd et al. 1994, Trapp et al. 1995). Furthermore, continuous microtubule turnover is required for the maintenance of membrane sheets by cultured oligodendrocytes (Benjamins & Nedelkoska 1994).

A 50-kDa human sciatic nerve protein bound to the recombinant S-MAGct in a ligand

73

blotting assay. The protein was further identified as tubulin by coprecipitation of C6 cell proteins with the S-MAGct, and by affinity chromatography using pure bovine tubulin. In addition to soluble tubulin, the S-MAGct bound to in vitro assembled microtubules, with an S-MAGct:tubulin molar ratio of 1:1, and it was able to regulate the rate of microtubule polymerisation. The behaviour of the S-MAGct in these assays was reminiscent of that of the classical microtubule-associated proteins of the tau/MAP family, and suggests that S-MAG may play a role in the regulation of the myelinating glial cell microtubule network.

Affinity chromatography demonstrated that the tubulin-S-MAGct interaction was strictly dependent on the presence of the 10-amino acid S-MAG-specific carboxy terminus. This region is homologous to the tubulin-binding repeats of the tau/MAP family. These repeats, of which three or four are usually found in a microtubule-associated protein (Drewes et al. 1998), are able to bind directly to microtubules (Ennulat et al. 1989, Lee et al. 1989). While S-MAG only carries one copy of this repeat, its presence is crucial for the tubulin-binding property. In line with the results of the CD spectroscopy for the S-MAGct, showing that it does not contain appreciable amounts of organised secondary structure in solution, the tau/MAP family proteins, and especially their tubulin-binding domains, have similarly been characterised as relatively unfolded polypeptide chains (Mandelkow et al. 1996). A growing number of proteins has been characterised as intrinsically unstructured, and the lack of defined structure in these proteins has been suggested to play a role in protein turnover and in protein-protein interactions (Wright & Dyson 1999). Attempts to determine the 3D structure of S-MAGct by X-ray crystallography during this study proved unsuccesful, which may have been caused by the apparently unfolded nature of this domain. While it is possible that problems in the structure determination could have been solved by searching for other crystal forms, this was not within the scope of the present work.

In addition to the even distribution of S-MAG in the differentiating S16 cell plasma membrane, a punctate pattern of staining was frequently observed that seemed to be oriented along microtubular structures. This staining may reflect the presence of MAG-containing transport vesicles in the cytoplasm, or MAG-rich microdomains in the plasma membrane. The punctate staining was increased in the perinuclear region of the cell when microtubule turnover was disrupted by either taxol or colchicine. In myelinating Schwann cells, MAG is transported to the myelin sheath in vesicles, in a process involving microtubules, and treatment with colchicine causes accumulation of MAG-containing vesicles in the Schwann cell perinuclear cytoplasm (Trapp et al. 1995). In the experiments reported in this study, the differentiating S16 cells were at a stage of rapid enlargement and morphological change, and it is likely that the vesicle-like staining observed for S-MAG in differentiating cells indeed originated from S-MAG -transporting structures that were arrested in the cytoplasm when microtubule turnover was disrupted.

The growth of cellular processes involves the polymerisation and rearrangement of microtubules, and the selective stabilisation and guidance of microtubules are required for process growth in the right direction (Suter & Forscher 1998). In this study, the colocalisation of S-MAG with tubulin at subcellular sites most likely to be mobile structures, i.e. the ruffled membrane borders and Schwann cell processes, suggests that S-MAG could act in directing and stabilising the dynamic microtubules within these structures. The relocalisation of S-MAG from small Schwann cell processes when

74

microtubule turnover was disrupted provides evidence for a functional interaction between S-MAG and microtubules in growing Schwann cell processes.

The effects on the localisation of S-MAG induced by taxol, also used in the treatment of certain human tumors, could account for the side-effect of the drug observed in some patients: a peripheral sensory neuropathy characterised by microtubule aggregation in the Schwann cell cytoplasm (Vuorinen et al. 1989, Vuorinen and Röyttä 1990). The relocalisation or the inefficient transport of S-MAG in these cells could lead to the disruption of the axon-Schwann cell interaction, contributing to the development of the taxol-induced neuropathy. A similar mechanism can be envisaged for the taiep rat, a neurological mutant characterised by the accumulation of microtubules in the oligodendrocyte cytoplasm (Duncan et al. 1992, Couve et al. 1997). In the taiep rat, the overall levels of myelin proteins in purified CNS myelin are decreased, but the level of MAG is significantly more reduced than that of other proteins (Möller et al. 1997).

To my knowledge, only two integral membrane proteins have previously been reported to interact directly with tubulin via their cytoplasmic domains. An integral membrane protein of the rough endoplasmic reticulum, p63, binds microtubules in vivo and in vitro, and a peptide corresponding to the cytoplasmic tail of p63 promotes microtubule polymerisation in vitro (Klopfenstein et al. 1998). The cytoplasmic domains of N-methyl D-aspartate receptor subunits have also been shown to interact directly with tubulin, decreasing the rate of microtubule formation in vitro (van Rossum et al. 1999). It is likely that interactions between microtubules and certain membrane proteins are important during morphological adaptations involving rearrangements of the microtubule cytoskeleton.

In conclusion, the S-MAGct is both structurally and functionally related to the microtubule-associated proteins of the tau/MAP family, of which S-MAG can now be considered to be a member. The identification of S-MAG as a MAP suggests that, in addition to its function as a cell adhesion molecule, it may play roles in the generation and maintenance of glial cell polarity, intracellular transport, and the regulation of the microtubular cytoskeleton.

6.2. The cytoplasmic domain of L-MAG

More experimental data is available on the function of the L-MAG cytoplasmic domain than on that of the S-MAGct, and it seems that L-MAG may be involved in signal transduction events in myelinating glial cells. L-MAG is a substrate for several src family tyrosine kinases, including Fyn, and the phosphorylation site at Tyr620 resembles closely the autophosphorylation site of the epidermal growth factor receptor. The binding of antibodies to the extracellular domain of L-MAG in transfected cells triggers Fyn activity (Umemori et al. 1994), and the tyrosine-phosphorylated form of L-MAG can bind to PLCγ (Jaramillo et al. 1994). L-MAG is also phosphorylated by PKC (Kirchhoff et al. 1993). Furthermore, the presence of putative endocytotic signals in the L-MAGct has been suggested to play a role in the sorting of MAG to endocytotic vesicles (Bo et al. 1995). Gene knockout studies have indicated that L-MAG is an essential component of CNS myelin (Fujita et al. 1998).

75

Dimerisation is a frequently observed phenomenon in cellular signal transduction molecules, a classical example being the epidermal growth factor receptor (Schlessinger 1988). The dimerisation of the recombinant L-MAGct in vitro was observed in gel filtration chromatography, ultracentrifugation, ligand blotting, and covalent crosslinking experiments. The same experiments performed using the GST and GST-S-MAGct proteins indicated that the L-MAG-specific region was required for dimerisation. It is not clear at present how dimerisation affects the function of L-MAG.

Based on the inability of forskolin and cAMP analogues to induce MAG phosphorylation in cultured cells (Afar et al. 1990, Kirchhoff et al. 1993, Yim et al. 1995), it has been concluded that MAG is not phosphorylated by PKA. Here, the recombinant L-MAGct was readily phosphorylated by PKA in vitro on serine and threonine, and one phosphorylation site is a threonine residue in the L-MAG-specific carboxy terminus, between amino acids 573 and 626. One classical consensus sequence for PKA phosphorylation (Kennelly & Krebs 1991) is present in this region, at Thr607. Furthermore, secondary structure predictions and analyses by CD spectroscopy indicate that Thr607 is located in an exposed turn structure (P. Kursula, unpublished observations), making it accessible for phosphorylation by PKA. As the other threonine residues in the L-MAG-specific domain have no similarity to the PKA consensus motif, it can be assumed that Thr607 of L-MAG is a PKA phosphorylation site. These results, while contradictory to some earlier evidence from cultured cells, indicate that it is possible that PKA phosphorylates L-MAG. In this respect, it is important to note that in one of the earlier studies (Afar et al. 1990), increased phosphorylation of L-MAG, but not S-MAG, in transfected cells was reported to occur in some experiments in the presence of dibutyryl cAMP. As the stoichiometric level of phosphorylation in vitro, using pure protein, was relatively low in the present study, although clearly detectable and isoform-specific, it is likely that the methods previously used to study MAG phosphorylation by PKA in cell cultures were not sensitive enough.

A putative cytoplasmic ligand for L-MAG was identified as the S100β protein. The S100 proteins comprise a group of small, acidic proteins belonging to the EF-hand family of calcium-binding proteins (reviewed by Donato 1999). They are believed to participate in the regulation of energy metabolism, the regulation of cell shape and polarity, and intracellular calcium-regulated signal transduction. The S100 family member expressed at highest levels by C6 glioma cells is S100β (Zimmer & Van Eldik 1988), and S100β is also specifically expressed in myelinating glial cells (Mata et al. 1990). Here, the interaction observed in ligand blotting experiments between L-MAGct and S100β was confirmed by affinity chromatography, dot blot, and crosslinking assays using pure S100β protein. The interaction was calcium-dependent, and the S100β-binding site was localised to the beginning of the L-MAG-specific sequence of the cytoplasmic domain, in a region predicted to form an amphipathic basic α-helix. A potential functional significance for this interaction is provided by the regulation of L-MAG phosphorylation by PKA. Furthermore, preliminary results indicate that calmodulin, a protein both structurally and functionally related to S100β, interacts with the same region in L-MAG, but in a calcium-independent manner (P. Kursula, unpublished observations). As calmodulin is also localised in the cytoplasmic pockets of noncompact PNS myelin (Mata & Fink 1988), the possibility that it is another cytoplasmic ligand for L-MAG should be further investigated.

76

The phosphorylation of many proteins, such as p87 (Kligman & Patel 1986), neurogranin (Sheu et al. 1994), and neuromodulin (Sheu et al. 1994, Lin et al. 1994) by PKC is affected by the presence of S100, implying that a major mechanism of S100 action involves intracellular protein kinase substrates. Interestingly, the S100β-binding site at the beginning of the L-MAG-specific carboxy terminus also harbors two serine residues that fit perfectly into a PKC phosphorylation concensus sequence. However, S100β had no significant effect on the strong phosphorylation of L-MAG by PKC. On the other hand, serine phosphorylation of the L-MAGct by PKA was observed, and this phosphorylation was specifically downregulated in the presence of S100β and calcium, suggesting that there is an S100β-regulated PKA phosphorylation site near or at the S100β-binding domain of L-MAG. Based on the PKA consensus motif, the most likely phosphorylation target in this region is Ser579.

The Ca2+-bound form of the S100β presents a hydrophobic face, which is absent in the apo-S100β form, and that is involved in the calcium-dependent interaction between S100β and the CapZ TRTK-12 peptide (Kilby et al. 1997). The sequence homology between the CapZ TRTK-12 peptide and the putative amphipathic basic α-helical region at the beginning of the L-MAG-specific region of the cytoplasmic domain suggests that the interaction described in this study implicates the same hydrophobic region of S100β and the amphipathic α-helix of the L-MAGct. The tendency of this region in the L-MAGct to fold into an α-helix in solution was confirmed by CD spectroscopy of a series of three overlapping peptides.

Interactions between S100 and many cytoskeleton-associated proteins, such as tubulin (Donato 1987), which is also present in the cytoplasm of the non-compacted regions of the myelin sheath (Kidd et al. 1996) where it may interact with the S-MAGct (V), tau (Baudier et al. 1987), the glial fibrillary acidic protein (Bianchi et al. 1993), the actin-capping protein CapZ (Ivanenkov et al. 1995), and annexin II (Bianchi et al. 1992), indicate that the S100 proteins are able to regulate the stability of all three major components of the cytoskeleton: actin microfilaments, microtubules, and intermediate filaments, and thus may have important roles in regulating cell shape and polarity. In fact, the prevention of S100β expression in cultured C6 glioma cells results in changes in cell shape, cytoskeletal organisation, and cell polarity (Selinfreund et al. 1990). MAG also colocalises with cytoskeletal proteins in myelinating glial cells, including F-actin and spectrin (Trapp et al. 1989a). It has been suggested, but not previously demonstrated, that the MAG cytoplasmic domain interacts with cytoskeletal components, and that these interactions, important in the wrapping of the glial cell membrane around the axon, could be regulated by the phosphorylation status of the MAG cytoplasmic tail (Attia et al. 1989). Moreover, the extraction behaviour of native MAG from nervous tissue suggests an association with membrane skeletal structures (II), and a similar extraction behaviour has been observed for a fraction of bovine brain S100 proteins (Donato et al. 1989). While direct interactions between S-MAGct and tubulin (V) are likely to be important for myelin structure, the calcium-dependent L-MAG-S100β interaction could also indirectly play a role in the regulation of the glial cell cytoskeleton and, hence, in the active process of changing the cell polarity and shape during myelination, and in the long-term maintenance of the myelin sheath.

The results of the present study further strengthen the long-held view of L-MAG as a signaling molecule, and indicate a putative role for this molecule in calcium-regulated

77

signaling events within the myelin sheath. The detailed molecular aspects of L-MAGct function - including the relationships between L-MAG dimerisation, interactions with S100β, and phosphorylation - as well as the upstream and downstream factors related to L-MAG function, are subjects for future studies.

6.3. Concluding remarks

At the beginning of this study, very little information was available on the functional and structural properties of the MAG cytoplasmic domains, and the differences between the two MAG isoforms. The results of this study indeed suggest that S- and L-MAG are very likely to have different functions in the myelinating glial cell cytoplasm. Previously, it had been suggested that L-MAG may act as a signaling molecule, while the role of S-MAG could be purely structural. The several functional phosphorylation sites in the L-MAG cytoplasmic domain, its interaction with S100β, and its dimerisation are strong evidence for a signaling role for L-MAG. On the other hand, the interaction of S-MAG with microtubules shows that it really may play an important structural role in the myelin sheath. This role as a molecule directly linking the axonal membrane and the glial cell cytoskeleton is likely to be very dynamic. The information and tools provided by this study will be useful for further studies on the structure, function, and interactions of MAG and other myelin proteins. A schematic representation summarising our current knowledge on the interactions of the two MAG isoforms with other molecules is given in figure 11.

78

79

extracellular

plasmamembrane

cytoplasm

S-MAG L-MAG

sialic acid-carrying glycolipidsand glycoproteins

extracellular matrix proteins

PKC

PKA

S100β

Zn

tubulin

dimerisation

dimerisation ?Fyn

PLCγ

Fig. 11. The current view of the interactions of MAG. While no discrimination has been made between the MAG isoforms regarding their extracellular ligands, the results of this study and previous research have been summarised for each cytoplasmic domain. The molecules found to interact with MAG in vitro in the present study are underlined.

7. References

Adlkofer K, Frei R, Neuberg DH, Zielasek J, Toyka KV & Suter U (1997) Heterozygous peripheral myelin protein 22-deficient mice are affected by a progressive demyelinating tomaculous neuropathy. J Neurosci 17: 4662-4671.

Adlkofer K, Martini R, Aguzzi A, Zielasek J, Toyka KV & Suter U (1995) Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deficient mice. Nature Genet 11: 274-280.

Afar DE, Salzer JL, Roder J, Braun PE & Bell JC (1990) Differential phosphorylation of myelin-associated glycoprotein isoforms in cell culture. J Neurochem 55: 1418-1426.

Agrawal HC, Noronha AB, Agrawal D & Quarles RH (1990) The myelin-associated glycoprotein is phosphorylated in the peripheral nervous system. Biochem Biophys Res Commun 169: 953-958.

Agrawal HC, Schmidt RE & Agrawal D (1983) In vivo incorporation of 3H-palmitic acid into P0 protein, the major intrinsic protein of rat sciatic nerve myelin. J Biol Chem 258: 6556-6560.

Aguayo AJ, Epps J, Charron L & Bray GM (1976) Multipotentiality of Schwann cells in cross-anastomosed and grafted myelinated and unmyelinated nerves: quantitative microscopy and radioautography. Brain Res 104: 1-20.

Andres KH (1965) Über die Feinstruktur besonderer Einrichtungen in markhaltigen Nervenfasern des Kleinhirns der Ratte. Z Zellforsch 65: 701-712.

Arquint M, Roder J, Chia LS, Down J, Wilkinson D, Bayley H, Braun P & Dunn R (1987) Molecular cloning and primary structure of myelin-associated glycoprotein. Proc Natl Acad Sci USA 84: 600-604.

Arroyo EJ & Scherer SS (2000) On the molecular architecture of myelinated fibers. Histochem Cell Biol 113: 1-18.

Asotra K & Macklin WB (1993) Protein kinase C activity modulates myelin gene expression in enriched oligodendrocytes. J Neurosci Res 34: 571-588.

Attia J, Tropak M, Johnson PW, Newerly-Abranow W, Pawson T, Roder JC & Dunn RJ (1989) Modulated adhesion: a proposal for the role of myelin-associated glycoprotein in myelin wrapping. Clin Chem 35: 717-720.

Baechner D, Liehr T, Hameister H, Altenberger H, Grehl H, Suter U & Rautenstrauss B (1995) Widespread expression of the peripheral myelin protein-22 gene (PMP22) in neural and non-neural tissues during murine development. J Neurosci Res 42: 733-741.

Baig S, Olsson T, Yu-Ping J, Hojeberg B, Cruz M & Link H (1991) Multiple sclerosis: cells secreting antibodies against myelin- associated glycoprotein are present in cerebrospinal fluid. Scand J Immunol 33: 73-79.

Balice-Gordon RJ, Bone LJ & Scherer SS (1998) Functional gap junctions in the schwann cell myelin sheath. J Cell Biol 142: 1095-1104 .

Bambrick LL & Braun PE (1991) Phosphorylation of myelin-associated glycoprotein in cultured oligodendrocytes. Dev Neurosci 13: 412-416.

Barbarese E, Braun PE & Carson JH (1977) Identification of prelarge and presmall basic proteins in mouse myelin and their structural relationship to large and small basic proteins. Proc Natl Acad Sci USA 74: 3360-3364.

Bargmann W & Lindner E (1964) Über den Feinbau des Nebennierenmarkes des Igels (Erinaceus europaeus L). Z Zellforsch 64: 868-912.

Barton DE, Arquint M, Roder J, Dunn R & Francke U (1987) The myelin-associated glycoprotein gene: mapping to human chromosome 19 and mouse chromosome 7 and expression in quivering mice. Genomics 1: 107-112.

Bartoszewicz ZP, Lauter CJ & Quarles RH (1996) The myelin-associated glycoprotein of the peripheral nervous system in trembler mutants contains increased alpha 2-3 sialic acid and galactose. J Neurosci Res 43: 587-593.

Bartoszewicz ZP, Noronha AB, Fujita N, Sato S, Bo L, Trapp BD & Quarles RH (1995) Abnormal expression and glycosylation of the large and small isoforms of myelin-associated glycoprotein in dysmyelinating quaking mutants. J Neurosci Res 41: 27-38.

Bartsch U, Bandtlow CE, Schnell L, Bartsch S, Spillmann AA, Rubin BP, Hillenbrand R, Montag D, Schwab ME & Schachner M (1995) Lack of evidence that myelin-associated glycoprotein is a major inhibitor of axonal regeneration in the CNS. Neuron 15: 1375-1381.

Baudier J, Briving C, Deinum J, Haglid K, Sorskog L & Wallin M (1982) Effect of S-100 proteins and calmodulin on Ca2+-induced disassembly of brain microtubule proteins in vitro. FEBS Lett 147: 165-168.

Baudier J & Cole RD (1988) Interactions between the microtubule-associated tau proteins and S100b regulate tau phosphorylation by the Ca2+/calmodulin-dependent protein kinase II. J Biol Chem 263: 5876-5883.

Baudier J, Delphin C, Grunwald D, Khochbin S & Lawrence JJ (1992) Characterization of the tumor suppressor protein p53 as a protein kinase C substrate and a S100b-binding protein. Proc Natl Acad Sci USA 89: 11627-11631.

Baudier J & Gerard D (1983) Ions binding to S100 proteins: structural changes induced by calcium and zinc on S100a and S100b proteins. Biochemistry 22: 3360-3369.

Baudier J, Glasser N, Haglid K & Gerard D (1984) Purification, characterization and ion binding properties of human brain S100b protein. Biochim Biophys Acta 790: 164-173.

Baudier J, Haglid K, Haiech J & Gerard D (1983) Zinc ion binding to human brain calcium binding proteins, calmodulin and S100b protein. Biochem Biophys Res Commun 114: 1138-1146.

Baudier J, Mochly-Rosen D, Newton A, Lee SH, Koshland DE, Jr. & Cole RD (1987) Comparison of S100b protein with calmodulin: interactions with melittin and microtubule-associated tau proteins and inhibition of phosphorylation of tau proteins by protein kinase C. Biochemistry 26: 2886-2893.

Benda P, Lightbody J, Sato G, Levine L & Sweet W (1968) Differentiated rat glial cell strain in tissue culture. Science 161: 370-371.

Beniac DR, Luckevich MD, Czarnota GJ, Tompkins TA, Ridsdale RA, Ottensmeyer FP, Moscarello MA & Harauz G (1997) Three-dimensional structure of myelin basic protein. I. Reconstruction via angular reconstitution of randomly oriented single particles. J Biol Chem 272: 4261-4268.

Benjamins JA & Morell P (1978) Proteins of myelin and their metabolism. Neurochem Res 3: 137-174.

Benjamins JA & Nedelkoska L (1994) Maintenance of membrane sheets by cultured oligodendrocytes requires continuous microtubule turnover and Golgi transport. Neurochem Res 19: 631-639.

Bergoffen J, Scherer SS, Wang S, Scott MO, Bone LJ, Paul DL, Chen K, Lensch MW, Chance PF & Fischbeck KH (1993) Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262: 2039-2042.

Berlet HH, Bischoff H & Weinhardt F (1994) Divalent metals of myelin and their differential binding by myelin basic protein of bovine central nervous system. Neurosci Lett 179: 75-78.

Bhat NR (1995) Signal transduction mechanisms in glial cells. Dev Neurosci 17: 267-284.

81

Bianchi R, Giambanco I & Donato R (1993) S-100 protein, but not calmodulin, binds to the glial fibrillary acidic protein and inhibits its polymerization in a Ca(2+)-dependent manner. J Biol Chem 268: 12669-12674.

Bianchi R, Pula G, Ceccarelli P, Giambanco I & Donato R (1992) S-100 protein binds to annexin II and p11, the heavy and light chains of calpactin I. Biochim Biophys Acta 1160: 67-75.

Bo L, Quarles RH, Fujita N, Bartoszewicz Z, Sato S & Trapp BD (1995) Endocytic depletion of L-MAG from CNS myelin in quaking mice. J Cell Biol 131: 1811-1820.

Bolin LM & Shooter EM (1993) Neurons regulate Schwann cell genes by diffusible molecules. J Cell Biol 123: 237-243.

Bollensen E & Schachner M (1987) The peripheral myelin glycoprotein P0 expresses the L2/HNK-1 and L3 carbohydrate structures shared by neural adhesion molecules. Neurosci Lett 82: 77-82.

Bosch EP & Smith BE (1993) Peripheral neuropathies associated with monoclonal proteins. Med Clin North Am 77: 125-139.

Brady RO & Quarles RH (1988) Developmental and pathophysiological aspects of the myelin-associated glycoprotein. Cell Mol Neurobiol 8: 139-148.

Brand IA, Heinickel A, Kratzin H & Soling HD (1988) Properties of a 19-kDa Zn2+-binding protein and sequence of the Zn2+-binding domains. Eur J Biochem 177: 561-568.

Braun PE, Horvath E & Edwards AM (1990) Two isoforms of myelin-associated glycoprotein accumulate in quaking mice: only the large polypeptide is phosphorylated. Dev Neurosci 12: 286-292.

Bray GM, Rasminsky M & Aguayo AJ (1981) Interactions between axons and the sheath cells. Annu Rev Neurosci 4: 127-162.

Brockes JP, Raff MC, Nishiguchi DJ & Winter J (1980) Studies on cultured rat Schwann cells. III. Assays for peripheral myelin proteins. J Neurocytol 9:67-77.

Brümmendorf T & Rathjen FG (1994) Cell adhesion molecules 1: Immunoglobulin superfamily. Protein Profile 1: 963-979.

Brunden KR & Brown DT (1990) P0 mRNA expression in cultures of Schwann cells and neurons that lack basal lamina and myelin. J Neurosci Res 27: 159-168.

Brünger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T & Warren GL (1998) Crystallography & NMR System: A new software suite for macromolecular structure determination. Acta Cryst D54: 905-921.

Bunge MB, Clark MB, Dean AC, Eldridge CF & Bunge RP (1990) Schwann cell function depends upon axonal signals and basal lamina components. Ann NY Acad Sci 580: 281-287.

Bunge RP, Bunge MB & Bates M (1989) Movements of the Schwann cell nucleus implicate progression of the inner (axon-related) Schwann cell process during myelination. J Cell Biol 109: 273-284.

Bunge RP, Bunge MB & Eldridge CF (1986) Linkage between axonal ensheathment and basal lamina production by Schwann cells. Annu Rev Neurosci 9: 305-328.

Burger D, Pidoux L & Steck AJ (1993) Identification of the glycosylated sequons of human myelin-associated glycoprotein. Biochem Biophys Res Commun 197: 457-464.

Carenini S, Montag D, Cremer H, Schachner M & Martini R (1997) Absence of the myelin-associated glycoprotein (MAG) and the neural cell adhesion molecule (N-CAM) interferes with the maintenance, but not with the formation of peripheral myelin. Cell Tissue Res 287: 3-9.

Carenini S, Montag D, Schachner M & Martini R (1998) MAG-deficient Schwann cells myelinate dorsal root ganglion neurons in culture. Glia 22: 213-220.

Carenini S, Montag D, Schachner M & Martini R (1999) Subtle roles of neural cell adhesion molecule and myelin-associated glycoprotein during Schwann cell spiralling in P0-deficient mice. Glia 27: 203-212.

Carey DJ, Todd MS & Rafferty CM (1986) Schwann cell myelination: induction by exogenous basement membrane-like extracellular matrix. J Cell Biol 102: 2254-2263.

Carson JH, Nielson ML & Barbarese E (1983) Developmental regulation of myelin basic protein expression in mouse brain. Dev Biol 96: 485-492.

Carswell R (1838) Pathological Anatomy: Illustration of the Elementary Forms of Disease. Longman Publishing, London.

82

Cassimeris L (1999) Accessory protein regulation of microtubule dynamics throughout the cell cycle. Curr Opin Cell Biol 11: 134-141.

Cavatorta P, Giovanelli S, Bobba A, Riccio P, Szabo AG & Quagliariello E (1994) Myelin basic protein interaction with zinc and phosphate: fluorescence studies on the water-soluble form of the protein. Biophys J 66: 1174-1179.

Chance PF & Lupski JR (1994) Inherited neuropathies: Charcot-Marie-Tooth disease and related disorders. Baillieres Clinical Neurology 3: 373-385.

Charcot J-M (1868) Histologie de la sclérose en plaques. Gaz Hop 41: 554-558.Charcot J-M & Marie P (1886) Sur une forme particulière d'atrophie musculaire progressive

souvent familiale débutant par les pieds et les jambes et atteignant plus tard les mains. Rev Méd 6: 97-138.

Chen AH, Ni L, Fukushima K, Marietta J, O'Neill M, Coucke P, Willems P & Smith RJH (1995) Linkage of a gene for dominant non-syndromic deafness to chromosome 19. Hum Mol Genet 4: 1073-1076.

Chomczynski P & Sacchi N (1987) Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 162: 156-159.

Coetzee T, Fujita N, Dupree J, Shi R, Blight A, Suzuki K, Suzuki K & Popko B (1996) Myelination in the absence of galactocerebroside and sulfatide: Normal structure with abnormal function and regional stability. Cell 86: 209-219.

Coetzee T, Suzuki K & Popko B (1998) New perspectives on the function of myelin galactolipids. Trends Neurosci 21: 126-130.

Collaborative Computational Project Number 4 (1994) The CCP4 suite: Programs for protein crystallography. Acta Cryst D50: 760-763.

Collins BE, Yang LJ, Mukhopadhyay G, Filbin MT, Kiso M, Hasegawa A & Schnaar RL (1997) Sialic acid specificity of myelin-associated glycoprotein binding. J Biol Chem 272: 1248-1255.

Conlon P, Oksenberg JR, Zhang J & Steinman L (1999) The immunobiology of multiple sclerosis: An autoimmune disease of the central nervous system. Neurobiol Disease 6: 149-166.

Corpet F, Gouzy J & Kahn D (1998) The ProDom database of protein domain families. Nucleic Acids Res 26: 323-326.

Couve E, Cabello F, Krsulovic J & Roncagliolo M (1997) Dysmyelination and microtubular defect in oligodendrocytes of the myelin mutant taiep rat. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 371.

D'Eustachio P, Colman DR & Salzer JL (1988) Chromosomal location of the mouse gene that encodes the myelin- associated glycoproteins. J Neurochem 50: 589-593.

D'Urso D, Brophy PJ, Staugaitis SM, Gillespie CS, Frey AB, Stempak JG & Colman DR (1990) Protein zero of peripheral nerve myelin: biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 4: 449-460.

D'Urso D, Ehrhardt P & Müller HW (1999) Peripheral myelin protein 22 and protein zero: a novel association in peripheral nervous system myelin. J Neurosci 19: 3396-3403.

D'Urso D & Muller HW (1997) Ins and outs of peripheral myelin protein-22: mapping transmembrane topology and intracellular sorting. J Neurosci Res 49: 551-562.

Dahme M, Bartsch U, Martini R, Anliker B, Schachner M & Mantei N (1997) Disruption of the mouse L1 gene leads to malformations of the nervous system. Nature Genet 17: 346-349.

Daniel PB, Walker WH & Habener JF (1998) Cyclic AMP signaling and gene regulation. Annu Rev Nutr 18: 353-383.

David S, Braun PE, Jackson DL, Kottis V & McKerracher L (1995) Laminin overrides the inhibitory effects of peripheral nervous system and central nervous system myelin-derived inhibitors of neurite growth. J Neurosci Res 42: 594-602.

Davis AD, Weatherby TM, Hartline DK & Lenz PH (1999) Myelin-like sheaths in copepod axons. Nature 398: 571.

de Waegh SM, Lee VM & Brady ST (1992) local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68: 451-463.

83

DeBellard ME, Tang S, Mukhopadhyay G, Shen YJ & Filbin MT (1996) Myelin-associated glycoprotein inhibits axonal regeneration from a variety of neurons via interaction with a sialoglycoprotein. Mol Cell Neurosci 7: 89-101.

Dekker LV & Parker PJ (1994) Protein kinase C - a question of specificity. Trends Biochem Sci 19: 73-77.

Deloulme JC, Sensenbrenner M & Baudier J (1990) Interactions of S100 proteins with protein kinase substrates. Biological implication. Adv Exp Med Biol 269: 153-157.

Derry DM & Wolfe LS (1967) Gangliosides in isolated neurons and glial cells. Science 158: 1450-1452.

Detering NK & Wells MA (1976) The non-synchronous synthesis of myelin components during early stages of myelination in the rat optic nerve. J Neurochem 26: 253-257.

Donato R (1983) Effect of S-100 protein on assembly of brain microtubule proteins in vitro. FEBS Lett 162: 310-313.

Donato R (1987) Quantitative analysis of the interaction between S-100 proteins and brain tubulin. Cell Calcium 8: 283-297.

Donato R (1999) Functional roles of S100 proteins, calcium-binding proteins of the EF-hand type. Biochim Biophys Acta 1450: 191-231.

Donato R, Giambanco I, Aisa MC & Ceccarelli P (1989) Identification of S-100 proteins and S-100-binding proteins in a detergent-resistant EDTA/KCl-extractable fraction from bovine brain membranes. FEBS Lett 247: 31-35.

Dong Z, Brennan A, Liu N, Yarden Y, Lefkowitz G, Mirsky R & jessen KR (1995) Neu differentiation factor (glial growth factor, heregulin) acts as a neuro-glia signal and regulates the survival, proliferation and maturation of the rat Schwann cenn precursor. Neuron 15: 585-596.

Downing KH & Nogales E (1998) Tubulin structure: Insights into microtubule properties and functions. Curr Opin Struct Biol 8: 785-791.

Drewes G, Ebneth A & Mandelkow EM (1998) MAPs, MARKs and microtubule dynamics. Trends Biochem Sci 23: 307-311.

Dulac C, Tropak MB, Cameron-Curry P, Rossier J, Marshak DR, Roder J & Le Douarin NM (1992) Molecular characterization of the Schwann cell myelin protein, SMP: structural similarities within the immunoglobulin superfamily. Neuron 8: 323-334.

Duncan DJ (1934) A relation between axon diameter and myelination determined by measurement of myelinated spinal root fibers. J Comp Neurol 60: 437-472.

Duncan ID, Lunn KF, Holmgren B, Urba-Holmgren R & Brignolo-Holmes L (1992) The taiep rat: a myelin mutant with an associated oligodendrocyte microtubular defect. J Neurocytol 21: 870-884.

Dyer CA (1997) Myelin proteins as mediators of signal transduction. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 69-74.

Dyer CA & Benjamins JA (1990) Galactocerebroside and sulfatide independently mediate Ca2+ responses in oligodendrocytes. J Neurosci Res 30: 699-711.

Earl C, Chantry A, Mohammad N & Glynn P (1988) Zinc ions stabilise the association of basic protein with brain myelin membranes. J Neurochem 51: 718-724.

Edwards AM, Arquint M, Braun PE, Roder JC, Dunn RJ, Pawson T & Bell JC (1988) Myelin-associated glycoprotein, a cell adhesion molecule of oligodendrocytes, is phosphorylated in brain. Mol Cell Biol 8: 2655-2658.

Edwards AM, Braun PE & Bell JC (1989) Phosphorylation of myelin-associated glycoprotein in vivo and in vitro occurs only in the cytoplasmic domain of the large isoform. J Neurochem 52: 317-320.

Eldridge CF, Bunge MB & Bunge RP (1989) Differentiation of axon-related Schwann cells in vitro. II. Control of myelin formation by basal lamina. J Neurosci 9: 625-638.

Eldridge CF, Bunge MB, Bunge RP & Wood PM (1987) Differentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. J Cell Biol 105: 1023-1034.

84

Endo T & Hidaka H (1983) Effect of S-100 protein on microtubule assembly-disassembly. FEBS Lett 161: 235-238.

Eng LF, Chao FC, Gerstl B, Pratt D & Tavaststjerna MG (1968) The maturation of human white matter myelin. Fractionation of the myelin membrane proteins. Biochemistry 7: 4455-4465.

Ennulat DJ, Liem RKH, Hashim KA & Shelanski ML (1989) Two separate 18-amino acid domains of tau promote the polymerization of tubulin. J Biol Chem 264: 5327-5330.

Everly JL, Brady RO & Quarles RH (1973) Evidence that the major protein in rat sciatic nerve myelin is a glycoprotein. J Neurochem 21: 329-334.

Fabrizi GM, Cavallaro T, Morbin M, Simonati A, Taioli F & Rizzuto N (1999) Novel mutation of the P0 extracellular domain causes a Dejerine-Sottas syndrome. J Neurol Neurosurg Psychiatry 66: 386-389.

Fahrig T, Landa C, Pesheva P, Kuhn K & Schachner M (1987) Characterization of binding properties of the myelin-associated glycoprotein to extracellular matrix constituents. Embo J 6: 2875-2883.

Fahrig T, Probstmeier R, Spiess E, Meyer-Franke A, Kirchhoff F, Drescher B & Schachner M (1993) Functional topography of the myelin-associated glycoprotein. I. Mapping of domains by electron microscopy. Eur J Neurosci 5: 1118-1126.

Fernandez-Valle C, Gorman D, Gomez AM & Bunge MB (1997) Actin plays a role in both changes in cell shape and gene-expression associated with Schwann cell myelination. J Neurosci 17: 241-250.

Fields RD (1998) Effects of ion channel activity on development of DRG neurons. J Neurobiol 37: 158-170.

Figlewicz DA, Quarles RH, Johnson D, Barbarash GR & Sternberger NH (1981) Biochemical demonstration of the myelin-associated glycoprotein in the peripheral nervous system. J Neurochem 37: 749-758.

Filbin MT & Tennekoon GI (1992) Myelin P0-protein, more than just a structural protein? Bioessays 14: 541-547.

Filbin MT, Wong MH, Zhang K & Li WH (1997) The role of myelin Po protein. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 103-114.

Fitzgerald M & Fulton B (1992) The physiological properties of developing sensory neurons. In: Scott SA (ed) Sensory neurons. Oxford University press, New York, pp. 287-306.

Frail DE & Braun PE (1984) Two developmentally regulated messenger RNAs differing in their coding region may exist for the myelin-associated glycoprotein. J Biol Chem 259: 14857-14862.

Frail DE & Braun PE (1985) Abnormal expression of the myelin-associated glycoprotein in the central nervous system of dysmyelinating mutant mice. J Neurochem 45: 1071-1075.

Frail DE, Edwards AM & Braun PE (1984) Molecular characteristics of the epitope in myelin-associated glycoprotein that is recognized by a monoclonal IgM in human neuropathy patients. Mol Immunol 21: 721-725.

Friede RL (1972) Control of myelin formation by axon caliber (with a model of the control mechanism). J Comp Neurol 144: 233-252.

Friede RL & Samorajski T (1968) Myelin formation in the sciatic nerve of the rat. A quantitative electron microscopic, histochemical and radioautographic study. J Neuropathol Exp Neurol 27: 546-570.

Fruttiger M, Montag D, Schachner M & Martini R (1995) Crucial role for the myelin-associated glycoprotein in the maintenance of axon-myelin integrity. Eur J Neurosci 7: 511-515.

Fujita N, Kemper A, Dupree J, Nakayasu H, Bartsch U, Schachner M, Maeda N, Suzuki K & Popko B (1998) The cytoplasmic domain of the large myelin-associated glycoprotein isoform is needed for proper CNS but not peripheral nervous system myelination. J Neurosci 18: 1970-1978.

Fujita N, Sato S, Kurihara T, Kuwano R, Sakimura K, Inuzuka T, Takahashi Y & Miyatake T (1989) cDNA cloning of mouse myelin-associated glycoprotein: a novel alternative splicing pattern. Biochem Biophys Res Commun 165: 1162-1169.

85

Garbay B, Heape AM, Sargueil F & Cassagne C (2000) Regulation of myelin synthesis in the peripheral nervous system. Prog Neurobiol 61: 267-304.

Garbay B, Salles J, Knoll A, Boiron-Sargueil F, Heape AM, Bonnet J & Cassagne C (1999) Trembler as a mouse model of CMT1A? Ann N Y Acad Sci 883: 262-272.

Garcia E, Elliott J, Gorvad A, Brandriff B, Gordon L, Soliman KM, Ashworth LK, Lennon G, Burgin M, Lamerdin J & et al. (1995) A continuous high-resolution physical map spanning 17 megabases of the q12, q13.1, and q13.2 cytogenetic bands of human chromosome 19. Genomics 27: 52-66.

Gardinier MV, Ballenthin PA, Kroepfl JF & Viise LR (1997) Myelin/oligodendrocyte glycoprotein: A molecular analysis. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 37-49.

Gendelman HE, Pezeshkpour GH, Pressman NJ, Wolinsky JS, Quarles RH, Dobersen MJ, Trapp BD, Kitt CA, Aksamit A & Johnson RT (1985) A quantitation of myelin-associated glycoprotein and myelin basic protein loss in different demyelinating diseases. Ann Neurol 18: 324-328.

Geren BB (1954) The formation from the Schwann cell surface of the myelin in the peripheral nerves of chick embryos. Exp Cell Res 7: 558-562.

Ghabriel MN & Allt G (1981) Incisures of Schmidt-Lanterman. Prog Neurobiol 17: 25-58.Giese KP, Martini R, lemke G, Soriano P & Schachner M (1992) Mouse P0 gene disruption leads

to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71: 565-576.

Gillespie CS, Sherman DL, Blair GE & Brophy PJ (1994) Periaxin, a novel protein of myelinating Schwann cells with a possible role in axonal ensheathment. Neuron 12: 497-508.

Goda S, Hammer J, Kobiler D & Quarles RH (1991) Expression of the myelin-associated glycoprotein in cultures of immortalized Schwann cells. J Neurochem 56: 1354-1361.

Goodenough DA, Goliger JA & Paul DL (1996) Connexins, connexons, and intercellular communication. Annu Rev Biochem 65: 475-502.

Greenfield N & Fasman GD (1969) Computed circular dichroism spectra for the evaluation of protein conformation. Biochemistry 8: 4108-4116.

Greenfield S, Weise MJ, Gantt G, Hogan EL & Brostoff SW (1982) Basic proteins of rodent peripheral nerve myelin: immunochemical identification of the 21.5K, 18.5K, 17K, 14K, and P2 proteins. J Neurochem 39: 1278-1282.

Greenwood FC, Hunter WM & Glover JS (1963) The preparation of 131I-labelled human growth hormone of high specific radioactivity. Biochem J 89: 114-123.

Grosson CL, Cannon SC, Corey DP & Gusella JF (1996) Sequence of the voltage-gated sodium channel beta1-subunit in wild-type and in quivering mice. Mol Brain Res 42: 222-226.

Guarnieri M, Himmelstein J & McKhann G (1974) Isolated myelin quantitatively adsorbs antibody to basic protein. Brain Res 72: 172-176.

Guex N & Peitsch MC (1997) SWISS-MODEL and the Swiss-Pdb Viewer: An environment for comparative protein modelling. Electrophoresis 18: 2714-2723.

Gupta SK, Poduslo JF, Dunn R, Roder J & Mezei C (1990) Myelin-associated glycoprotein gene expression in the presence and absence of Schwann cell-axonal contact. Dev Neurosci 12: 22-33.

Gupta SK, Poduslo JF & Mezei C (1988) Temporal changes in P0 and MBP gene expression after crush-injury of the adult peripheral nerve. Mol Brain Res 4: 133-141.

Gurr MI & James AT (1971) Lipid Biochemistry: An introduction. Chapman and Hall, London.Hamberger A & Svennerholm L (1971) Composition of gangliosides and phospholipids of

neuronal and glial cell enriched fractions. J Neurochem 18: 1821-1829.Haney C, Snipes GJ, Shooter EM, Suter U, Garcia C, Griffin JW & Trapp BD (1996)

Ultrastructural distribution of PMP22 in Charcot-Marie-Tooth disease type 1A. J Neuropathol Exp Neurol 55: 290-299.

Harlow E & Lane D (1988) Antibodies - A laboratory manual. Cold Spring Harbor Laboratory, New York.

86

Hays AP, Latov N, Takatsu M & Sherman WH (1987) Experimental demyelination of nerve induced by serum of patients with neuropathy and an anti-MAG IgM M-protein. Neurology 37: 242-256.

Henry EW, Eicher EM & Sidman RL (1991) The mouse mutation claw paw : forelimb deformity and delayed myelination throughout the peripheral nervous system. J Hered 82: 287-294.

Herndon RW, Rauch HC & Einstein ER (1973) Immunoelectron microscopic localization of the encephalitogenic basic protein in myelin. Immunol Commun 2: 163-172.

Hesketh J & Baudier J (1986) Evidence that S100 proteins regulate microtubule assembly and stability in rat brain extracts. Int J Biochem 18: 691-695.

Huxley AF & Stämpfli R (1949) Evidence for saltatory conduction in peripheral myelinated nerve fibers. J Physiol 108: 315-339.

Huxley C, Passage E, Manson A, Putzu G, Figarella-Branger D, Pelissier JF & Fontes M (1996) Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum Mol Genet 5: 563-569.

Hydén H & McEwen B (1966) A glial protein specific for the nervous system. Proc Natl Acad Sci USA 55: 354-358.

Inoue Y, Nakamura R, Mikoshiba K & Tsukada Y (1981) Fine structure of the central myelin sheath in the myelin deficient mutant Shiverer mouse, with special reference to the pattern of myelin formation by oligodendroglia. Brain Res 219: 85-94.

Inouye H & Kirschner DA (1984) Effects of ZnCl2 on membrane interactions in myelin of normal and shiverer mice. Biochim Biophys Acta 776: 197-208.

Inouye H, Tsuruta H, Sedzik J, Uyemura K & Kirschner DA (1999) Tetrameric assembly of full-sequence protein zero myelin glycoprotein by synchrotron x-ray scattering. Biophys J76: 423-437.

Inuzuka T, Fujita N, Sato S, Baba H, Nakano R, Ishiguro H & Miyatake T (1991) Expression of the large myelin-associated glycoprotein isoform during the development in the mouse peripheral nervous system. Brain Res 562: 173-175.

Inuzuka T, Quarles RH, Heath J & Trapp BD (1985) Myelin-associated glycoprotein and other proteins in Trembler mice. J Neurochem 44: 793-797.

Inuzuka T, Quarles RH, Noronha AB, Dobersen MJ & Brady RO (1984a) A human lymphocyte antigen is shared with a group of glycoproteins in peripheral nerve. Neurosci Lett 51: 105-111.

Inuzuka T, Sato S, McIntyre LJ & Quarles RH (1984b) Effects of trypsin and plasmin treatment of myelin on the myelin- associated glycoprotein and basic protein. J Neurochem 43: 582-585.

Ivanenkov VV, Jamieson GAJr, Gruenstein E & Dimlich RV (1995) Characterization of S-100b binding epitopes. Identification of a novel target, the actin capping protein, CapZ. J Biol Chem 270: 14651-14658.

Jaramillo ML, Afar DE, Almazan G & Bell JC (1994) Identification of tyrosine 620 as the major phosphorylation site of myelin-associated glycoprotein and its implication in interacting with signaling molecules. J Biol Chem 269: 27240-27245.

Jessen KR & Mirsky R (1991) Schwann cell precursors and their development. Glia 4: 9-18.Jessen KR & Mirsky R (1992) Schwann cells: early lineage, regulation of proliferation and control

of myelin formation. Curr Opin Neurobiol 2: 575-581.Jessen KR, Mirsky R, Dong Z & Brennan A (1997) Developmental regulation of Schwann cell

precursors and Schwann cell generation. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 165-172.

Johnson D, Hafler DA, Fallis RJ, Lees MB, Brady RO, Quarles RH & Weiner HL (1986) Cell-mediated immunity to myelin-associated glycoprotein, proteolipid protein, and myelin basic protein in multiple sclerosis. J Neuroimmunol 13: 99-108.

Johnson PW, Abramow-Newerly W, Seilheimer B, Sadoul R, Tropak MB, Arquint M, Dunn RJ, Schachner M & Roder JC (1989) Recombinant myelin-associated glycoprotein confers neural adhesion and neurite outgrowth function. Neuron 3: 377-385.

Johnston RJ & Peters A (1973) The conducting tissues. In: Passmore R & Robson JS (eds) A companion to medical studies. Volume 1: Anatomy, biochemistry, physiology and related subjects. Blackwell Scientific Publications, Oxford. p. 14.1-14.29.

87

Jones TA, Zou J-Y, Cowan SW & Kjeldgaard M (1991) Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Cryst A47: 110-119.

Juguelin H, Heape A, Boiron F & Cassagne C (1986) A quantitative developmental study of neutral lipids during myelinogenesis in the peripheral nervous system of normal and trembler mice. Brain Res 390: 249-252.

Kandel ER (1991) Nerve cells and behavior: glial cells. In: Kandel ER, Schwartz JH & Jessell TM (eds) Principles of neural science. Elsevier, New York, p. 22-24.

Kaplan MR, Meyer-Franke A, Lambert S, Bennett V, Duncan ID, Levinson SR & Barres BA (1997) Induction of sodium channel clustering by oligodendrocytes. Nature 386: 724-728.

Karthisagan J, Kosaras B, Nguyen J & Kirschner DA (1994) Protein and lipid composition of radial component-enriched CNS myelin. J Neurochem 62: 1203-1213.

Kaye JF, Kerlero de Rosbo N, Mendel I, Flechter S, Hoffman M, Yust I & ben-Nun A (2000) The central nervous system-specific myelin oligodendrocytic basic protein (MOBP) is encephalitogenic and a potential target antigen in multiple sclerosis (MS). J Neuroimmunol 102: 189-198.

Kelm S, Schauer R & Crocker PR (1996) The Sialoadhesins--a family of sialic acid-dependent cellular recognition molecules within the immunoglobulin superfamily. Glycoconj J 13: 913-926.

Kennelly PJ & Krebs EG (1991) Consensus sequences as substrate specificity determinants for protein kinases and protein phosphatases. J Biol Chem 266: 15555-15558.

Kidd G, Andrews SB & Trapp BD (1994) Organization of microtubules in myelinating Schwann cells. J Neurocytol 23: 801-810.

Kidd G, Andrews SB & Trapp BD (1996) Axons regulate the distribution of Schwann cell microtubules. J Neurosci 16: 946-954.

Kidd GJ & Heath JW (1988) Double myelination of axons in the sympathetic nervous system of the mouse. I. Ultrastructural features and distribution. J Neurocytol 17: 245-261.

Kilby PM, van Eldik LJ & Roberts GC (1997) Identification of the binding site on S100B protein for the actin capping protein CapZ. Protein Sci 6: 2494-2503.

Kioussi C, Gross MK & Gruss P (1995) Pax3: a paired domain gene as a regulator in PNS myelination. Neuron 15: 553-562.

Kirchhoff F, Hofer HW & Schachner M (1993) Myelin-associated glycoprotein is phosphorylated by protein kinase C. J Neurosci Res 36: 368-381.

Kirschner DA & Blaurock AE (1992) Organization, phylogenetic variations, and dynamic transitions of myelin. CRC Press, Boca Raton.

Kirschner DA & Ganser AL (1980) Compact myelin exists in the absence of basic protein in the shiverer mutant mouse. Nature 283: 207-210.

Kitamura K, Suzuki M & Uyemura K (1976) Purification and partial characterization of two glycoproteins in bovine peripheral nerve myelin membrane. Biochim Biophys Acta 455: 806-809.

Kligman D & Patel J (1986) A protein modulator stimulates C kinase-dependent phosphorylation of a 90K substrate in synaptic membranes. J Neurochem 47: 298-303.

Klopfenstein DR, Kappeler F & Hauri HP (1998) A novel direct interaction of endoplasmic reticulum with microtubules. EMBO J 17: 6168-6177.

Koenig HL, Schumacher M, Ferzaz B, Thi AN, Ressouches A, Guennoun R, Jung-Testas I, Robel P, Akwa Y & Baulieu EE (1995) Progesterone synthesis and myelin formation by Schwann cells. Science 268: 1500-1503.

Konat GW (1996) Chromatin structure and transcriptional activity of MAG gene. Acta Neurobiol Exp 56: 281-285.

Korngurth S & Anderson L (1965) Localization of a basic protein of various species with the aid of fluorescence and electron microscopy. J Cell Biol 26: 157-166.

Kosaras B & Kirschner DA (1990) Fine-structure and supramolecular organization of the radial component of CNS myelin. In: Duncan ID, Skoff RP & Colman D (eds) Myelination and dysmyelination. The New York Academy of Sciences, New York, p. 430-434.

88

Koszowski AG, Owens GC & Levinson SR (1998) The effect of the mouse mutation claw paw on myelination and nodal frequency in sciatic nerves. J Neurosci 18: 5859-5868.

Krämer EM, Klein C, Koch T, Boytinck M & Trotter J (1999) Compartmentation of Fyn kinase with glycosylphosphatidylinositol- anchored molecules in oligodendrocytes facilitates kinase activation during myelination. J Biol Chem 274: 29042-29049.

Krämer E-M, Koch T, Niehaus A & Trotter J (1997) Oligodendrocytes direct glycosyl phosphatidylinositol-anchored proteins to the myelin sheath in glycosphingolipid-rich complexes. J Biol Chem 272: 8937-8945.

Ktistakis NT, Thomas D & Roth MG (1990) Characteristics of the tyrosine recognition signal for internalization of transmembrane surface glycoproteins. J Cell Biol 111: 1393-1407.

Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685.

Lai C, Brow MA, Nave KA, Noronha AB, Quarles RH, Bloom FE, Milner RJ & Sutcliffe JG (1987a) Two forms of 1B236/myelin-associated glycoprotein, a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc Natl Acad Sci USA 84: 4337-4341.

Lai C, Watson JB, Bloom FE, Sutcliffe JG & Milner RJ (1987b) Neural protein 1B236/myelin-associated glycoprotein (MAG) defines a subgroup of the immunoglobulin superfamily. Immunol Rev 100: 129-151.

Lassmann H, Bartsch U, Montag D & Schachner M (1997) Dying-back oligodendrogliopathy: a late sequel of myelin-associated glycoprotein deficiency. Glia 19: 104-110.

Laszkiewicz I, Grubinska B, Wiggins RC & Konat GW (1997) Structural characterization of myelin-associated glycoprotein gene core promoter. J Neurosci Res 50: 928-936.

LeBlanc AC & Poduslo JF (1990) Regulation of apolipoprotein E gene expression after injury of the rat sciatic nerve. J Neurosci Res 25: 162-171.

LeBlanc AC, Poduslo JF & Mezei C (1987) Gene expression in the presence or absence of myelin assembly. Mol Brain Res 2: 57-67.

Ledeen RW (1989) Biosynthesis, metabolism and biological effects of gangliosides. In: Margolis RU & Margolis RK (eds) Neurobiology of glycoconjugates. Plenum Press, New York, p. 43-83.

Lee G, Neve RL & Kosik KS (1989) The microtubule binding domain of tau protein. Neuron 2: 1615-1624.

Lemke G & Axel R (1985) Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40: 501-508.

Li C, Tropak MB, Gerlai R, Clapoff S, Abramow-Newerly W, Trapp B, Peterson A & Roder J (1994) Myelination in the absence of myelin-associated glycoprotein. Nature 369: 747-750.

Li Q & Ownby CL (1993) A rapid method for extraction of DNA from agarose gels using a syringe. BioTechniques 15: 976-978.

Lim K, Ho JX, Keeling K, Gilliland GL, Ji X, Ruker F & Carter DC (1994) Three-dimensional structure of Schistosoma japonicum glutathione S-transferase fused with a six-amino acid conserved neutralizing epitope of gp41 from HIV. Protein Sci 3: 2233-2244.

Lin LH, Van Eldik LJ, Osheroff N & Norden JJ (1994) Inhibition of protein kinase C- and casein kinase II-mediated phosphorylation of GAP-43 by S100 beta. Mol Brain Res 25: 297-304.

Linington C, Bradl M, Lassmann H, Brunner C & Vass K (1988) Augmentation of demyelination in rat acute allergic encephalomyelitis by circulating mouse monoclonal antibodies directed against myelin/oligodendrocyte glycoprotein. Am J Pathol 130: 443-454.

Linington C, Webb M & Woodhams PL (1984) A novel myelin-associated glycoprotein defined by a mouse monoclonal antibody. J Neuroimmunol 6: 387-396.

Link H, Baig S, Jiang YP, Olsson O, Hojeberg B, Kostulas V & Olsson T (1989) B cells and antibodies in MS. Res Immunol 140: 219-226.

Maccioni RB, Vera JC, Dominguez J & Avila J (1989) A discrete repeated sequence defines a tubulin binding domain on microtubule-associated protein tau. Arch Biochem Biophys 275: 568-579.

Mackay JP & Crossley M (1998) Zinc fingers are sticking together. Trends Biochem Sci 23: 1-4.

89

Mandelkow EM, Schweers O, Drewes G, Biernat J, Gustke N, Trinczek B & Mandelkow E (1996) Structure, microtubule interactions, and phosphorylation of tau protein. Ann NY Acad Sci 777: 96-106.

Martini R, Bollensen E & Schachner M (1988) Immunocytological localization of the major peripheral nervous system glycoprotein P0 and the L2/HNK-1 and L3 carbohydrate structures in developing and adult mouse sciatic nerve. Dev Biol 129: 330-338.

Martini R & Schachner M (1986) Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J Cell Biol 103: 2439-2448.

Mata M, Alessi D & Fink DJ (1990) S100 is preferentially distributed in myelin-forming Schwann cells. J Neurocytol 19: 432-442.

Mata M & Fink DJ (1988) Calmodulin distribution in peripheral nerve: An EM immunocytochemical study. Brain Res 475: 297-304.

Matthieu JM, Mottet S, Kraus-Ruppert R, Cohen SR & Webster HD (1981) Effects of colchicine on myelination of rabbit optic nerve: a biochemical study. Neurotoxicology 2: 451-461.

May AP, Robinson RC, Vinson M, Crocker PR & Jones EY (1998) Crystal structure of the N-terminal domain of sialoadhesin in complex with 3' sialyllactose at 1.85 Å resolution. Mol Cell 1: 719-728.

McGarry RC, Helfand SL, Quarles RH & Roder JC (1983) Recognition of myelin-associated glycoprotein by the monoclonal antibody HNK-1. Nature 306: 376-378.

McKerracher L & David S (1997) Role of MAG as an axon growth inhibitory protein for regeneration of injured neurons in the CNS. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 309-317.

McTigue MA, Williams DR & Tainer JA (1995) Crystal structures of a schistosomal drug and vaccine target: glutathione S-transferase from Schistosoma japonica and its complex with the leading antischistosomal drug praziquantel. J Mol Biol 246: 21-27.

Melcangi RC, Magnaghi V & Martini L (2000) Aging in peripheral nerves: Regulation of myelin protein genes by steroid hormones. Prog Neurobiol 60: 291-308.

Miescher GC, Lützelschwab R, Erne B, Ferracin F, Huber S & Steck AJ (1997) Reciprocal expression of myelin-associated glycoprotein splice variants in the adult human peripheral and central nervous systems. Mol Brain Res 52: 299-306.

Miescher GC & Steck AJ (1996) Paraproteinaemic neuropathies. Baillieres Clin Neurol 5: 219-232.

Mikol DD & Stefansson K (1988) A phosphatidylinositol-linked peanut agglutinin-binding glycoprotein in central nervous system myelin and on oligodendrocytes. J Cell Biol 106: 1273-1279.

Minuk J & Braun PE (1996) Differential intracellular sorting of the myelin-associated glycoprotein isoforms. J Neurosci Res 44: 411-420.

Mirsky R & Jessen KR (1996) Schwann cell development, differentiation and myelination. Curr Opin Neurobiol 6: 89-96.

Mirsky R, Winter J, Abney E, Pruss RM, Gavrilovic J & Raff MC (1980) Myelin-specific proteins and glycolipids in rat Schwann cells and oligodendrocytes in culture. J Cell Biol 84: 483-494.

Montag D, Giese KP, Bartsch U, Martini R, Lang Y, Bluthmann H, Karthigasan J, Kirschner DA, Wintergerst ES, Nave KA & et al. (1994) Mice deficient for the myelin-associated glycoprotein show subtle abnormalities in myelin. Neuron 13: 229-246.

Monuki ES, Kuhn R & Lemke G (1993) Repression of the myelin P0 gene by the POU transcription factor SCIP. Mech Dev 42: 15-32.

Monuki ES, Weinmaster G, Kuhn R & Lemke G (1989) SCIP: a glia POU domain gene regulated by cyclic AMP. Neuron 3: 783-793.

Moore BW (1965) A soluble protein characteristic of the nervous system. Biochem Biophys Res Commun 19: 739-744.

Moore BW, Perez VJ & Gehring M (1968) Assay and regional distribution of a soluble protein characteristic of the nervous system. J Neurochem 15: 265-272.

90

Morell P, Quarles RH & Norton WT (1994) Myelin formation, structure and biochemistry. In: Siegel GJ (ed) Basic neurochemistry: Molecular, cellular, and medical aspects. Raven Press, New York, p. 117-144.

Moscarello MA (1997) Myelin basic protein, the "executive" molecule of the myelin membrane. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 13-26.

Mozier NM, Walsh MP & Pearson JD (1991) Characterization of a novel zinc-binding site of protein kinase C inhibitor-1. FEBS Lett 279: 14-18.

Määttä JA, Kaldman MS, Sakoda S, Salmi AA & Hinkkanen AE (1998) Encephalitogenicity of myelin-associated oligodendrocytic basic protein and 2', 3'-cyclic nucleotide 3'-phosphodiesterase for BALB/c and SJL mice. Immunology 95: 383-388.

Möller JR (1996) Rapid conversion of myelin-associated glycoprotein to a soluble derivative in primates. Brain Res 741: 27-31.

Möller JR, Durr PG, Quarles RH & Duncan ID (1997) Biochemical analysis of myelin proteins in a novel neurological mutant: the taiep rat. J Neurochem 69: 773-779.

Möller JR, Yanagisawa K, Brady RO, Tourtellotte WW & Quarles RH (1987) Myelin-associated glycoprotein in multiple sclerosis lesions: a quantitative and qualitative analysis. Ann Neurol 22: 469-474.

Naef R & Suter U (1999) Impaired intracellular trafficking is a common disease mechanism of PMP22 point mutations in peripheral neuropathies. Neurobiol Disease 6: 1-14.

Nakano R, Fujita N, Sato S, Inuzuka T, Sakimura K, Ishiguro H, Mishina M & Miyatake T (1991) Structure of mouse myelin-associated glycoprotein gene. Biochem Biophys Res Commun 178: 282-290.

Narayanan V, Barbosa E, Reed R & Tennekoon G (1988) Characterization of a cloned cDNA encoding rabbit myelin P2 protein. J Biol Chem 263: 8332-8337.

Navaza J (1994) AMoRe: An automated package for molecular replacement. Acta Cryst A50: 157-163.

Needham LK & Schnaar RL (1993) Carbohydrate recognition in the peripheral nervous system: A calcium-dependent membrane binding site for HNK-1 reactive glycolipids potentially involved in Schwann cell adhesion. J Cell Biol 121: 397-408.

Niemann S, Sidman RL & Nave K-A (1998) Evidence against altered forms of MAG in the dysmyelinated mouse mutant claw paw. Mamm Gen 9: 903-904.

Nishikawa T, Matsui-Lee IS, Shiraishi N, Ishikawa T, Ohta Y & Nishikimi M (1997) Identification of S100b protein as copper-binding protein and its suppression of copper-induced cell damage. J Biol Chem 272: 23037-23041.

Nobile-Orazio E, Hays AP, Latov N, Perman G, Golier J, Shy ME & Freddo L (1984) Specificity of mouse and human monoclonal antibodies to myelin-associated glycoprotein. Neurology 34: 1336-1342.

Nogales E, Wolf SG & Downing KH (1998) Structure of the αβ tubulin dimer by electron crystallography. Nature 391: 199-203.

Nussbaum JL, Neskovic N & Mandel P (1969) A study of lipid components in brain of the 'Jimpy' mouse, a mutant with myelin deficiency. J Neurochem 16: 927-934.

Omlin FX, Webster HdeF, Palkovits CE & Cohen SR (1982) Immunocytochemical localization of basic protein in major dense line regions of central and peripheral myelin. J Cell Biol 95: 242-248.

Osterhout DJ, Wolven A, Wolf RM, Resh MD & Chao MV (1999) Morphological differentiation of oligodendrocytes requires activation of Fyn tyrosine kinase. J Cell Biol 145: 1209-1218.

Otwinowski Z & Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. In: Carter CW Jr & Sweet RM (eds) Methods Enzymol 276: 307-326. Academic Press, London.

Owens GC, Boyd CJ, Bunge RP & Salzer JL (1990) Expression of recombinant myelin-associated glycoprotein in primary Schwann cells promotes the initial investment of axons by myelinating Schwann cells. J Cell Biol 111: 1171-1182.

91

Owens GC & Bunge RP (1989) Evidence for an early role for myelin-associated glycoprotein in the process of myelination. Glia 2: 119-128.

Owens GC & Bunge RP (1991) Schwann cells infected with a recombinant retrovirus expressing myelin- associated glycoprotein antisense RNA do not form myelin. Neuron 7: 565-575.

Parmantier E, Cabon F, Braun C, D'Urso D, Muller HW & Zalc B (1995) Peripheral myelin protein-22 is expressed in rat and mouse brain and spinal cord motoneurons. Eur J Neurosci 7: 1080-1088.

Parton RG & Simons K (1995) Digging into caveolae. Science 269: 1398-1399. Patel J, Marangos PJ, Heydorn WE, Chang G, Verma A & Jacobowitz D (1983) S-100-mediated

inhibition of brain protein phosphorylation. J Neurochem 41: 1040-1045.Patel PI, Roa BB, Welcher AA, Schoener-Scott R, Trask BJ, Pentao L, Snipes GJ, Garcia CA,

Francke U, Shooter EM, Lupski JR & Suter U (1992) The gene for the peripheral myelin protein PMP-22 is a candidate for Charcot-Marie-Tooth disease type 1A. Nature Genet 1: 159-165.

Pedraza L, Frey AB, Hempstead BL, Colman DR & Salzer JL (1991) Differential expression of MAG isoforms during development. J Neurosci Res 29: 141-148.

Pedraza L, Owens GC, Green LA & Salzer JL (1990) The myelin-associated glycoproteins: membrane disposition, evidence of a novel disulfide linkage between immunoglobulin-like domains, and posttranslational palmitylation. J Cell Biol 111: 2651-2661.

Pedraza L, Spagnol G, Latov N & Salzer JL (1995) Biosynthesis and regulation of expression of the HNK-1 epitope on myelin-associated glycoprotein in a transfected cell model system. J Neurosci Res 40: 716-727.

Peitsch MC (1996) ProMod and Swiss-Model: Internet-based tools for automated comparative protein modelling. Biochem Soc Trans 24: 274-279.

Pereyra PM, Horvath E & Braun PE (1988) Triton X-100 extractions of central nervous system myelin indicate a possible role for the minor myelin proteins in the stability in lamellae. Neurochem Res 13: 583-595.

Peters A, Palay SL & Webster deFH (1970) The fine structure of the nervous system. The cells and their processes. McGraw-Hill, New York.

Peters A, Palay SL & Webster HdeF (1991) The fine structure of the nervous system. Oxford University Press, New York.

Peters A & Proskaur C (1969) The ratio between myelin segments and oligodendrocytes in the optic nerve of the adult rat. Anat Rec 163: 243A.

Peters N, Smith JS, Tachibana I, Lee HK, Pohl U, Portier BP, Louis DN & Jenkins RB (1999) The human glia maturation factor gamma gene: genomic structure and mutation analysis in gliomas with chromosome 19q loss. Neurogenetics 2: 163-166.

Petit C (1996) Genes responsible for human hereditary deafness: symphony of a thousand. Nature Genet 14: 385-391.

Pfeiffer SE, Warrington AE & Bansal R (1993) The oligodendrocyte and its many cellular processes. Trends Cell Biol 3: 191-197.

Pierschbacher MD & Ruoslahti E (1984) Cell attachment activity of fibronectin can be duplicated by small synthetic fragments of the molecule. Nature 309: 30-33.

Poduslo JF (1984) Regulation of myelination: Biosynthesis of the major myelin glycoprotein by Schwann cells in the presence and absence of myelin assembly. J Neurochem 42: 493-503.

Porath J, Carlsson J, Olsson I & Belfrage G (1975) Metal chelate affinity chromatography, a new approach to protein fractionation. Nature 258: 598-599.

Privat A, Jacque C, Bourre JM, Dupouey P & Baumann N (1979) Absence of the major dense line in myelin of the mutant mouse "shiverer". Neurosci Lett 12: 107-112.

Quarles RH (1989) Myelin-associated glycoprotein in demyelinating disorders. Crit Rev Neurobiol 5: 1-28.

Quarles R, Colman D, Salzer J & Trapp B (1992) Myelin-associated glycoprotein: Structure-function relationships and involvement in neurological diseases. In: Martenson R (ed) Myelin: Biology and chemistry. CRC, Boca Raton, p. 4413-4448.

Quarles RH, Everly JL & Brady RO (1972) Demonstration of a glycoprotein which is associated with a purified myelin fraction from rat brain. Biochem Biophys Res Commun 47: 491-497.

92

Quarles RH, Everly JL & Brady RO (1973) Evidence for the close association of a glycoprotein with myelin in rat brain. J Neurochem 21: 1177-1191.

Quarles RH, Farrer RG & Yim SH (1997) Structure and function of myelin, an extended and biochemically modified cell surface membrane. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 1-12.

Rambotti MG, Spreca A, Leoncini P, Estenoz M, Constantino-Ceccarini E, Giambanco I & Donato R (1990) detection of S-100b protein in Triton cytoskeletons: An immunocytochemical study on cultured Schwann cells. J Histochem Cytochem 38: 1583-1589.

Ranvier ML (1878) Leçons sur l'histologie du système nerveux. F Savy, Paris.Reifenberger J, Reifenberger G, Liu L, James CD, Wechsler W & Collins VP (1994) Molecular

genetic analysis of oligodendroglial tumors shows preferential allelic deletions on 19q and 1p. Am J Pathol 145: 1175-1190.

Riccio P, Giovannelli S, Bobba A, Romito E, Fasano A, Bleve-Zacheo T, Favilla R, Quagliariello E & Cavatorta P (1995) Specificity of zinc binding to myelin basic protein. Neurochem Res 20: 1107-1113.

Rice LM & Brünger AT (1994) Torsion angle dynamics: Reduced variable conformational sampling enhances crystallographic structure refinement. Proteins 19: 277-290.

Ridsdale RA, Beniac DR, Tompkins TA, Moscarello MA & Harauz G (1997) Three-dimensional structure of myelin basic protein. II. Molecular modeling and considerations of predicted structures in multiple sclerosis. J Biol Chem 272: 4269-4275.

Ritchie JM (1984) Voltage-gated cation and anion channels in mammalian Schwann cells and astrocytes. J Physiol 82: 248-257.

Roa BB, Warner LE, Garcia CA, Russo D, Lovelace R, Chance PF & Lupski JR (1996) Myelin protein zero (MPZ) gene mutations in nonduplication type 1 Charcot-Marie-Tooth disease. Hum Mutat 7: 36-45.

Roach A, Takahashi N, Pravtcheva D, Ruddle F & Hood L (1985) Chromosomal mapping of mouse myelin basic protein gene and structure and transcription of the partially deleted gene in shiverer mutant mice. Cell 42: 149-155.

Rodriguez M & Scheithauer B (1994) Ultrastructure of multiple sclerosis. Ultrastruct Pathol 18: 3-13.

Rosenberg JE, Lisle DK, Burwick JA, Ueki K, von Deimling A, Mohrenweiser HW & Louis DN (1996) Refined deletion mapping of the chromosome 19q glioma tumor suppressor gene to the D19S412-STD interval. Oncogene 13: 2483-2485.

Rosenbluth J (1980) Peripheral myelin in the mouse mutant Shiverer. J Comp Neurol 193: 729-739.

Sadoul R, Fahrig T, Bartsch U & Schachner M (1990) Binding properties of liposomes containing the myelin-associated glycoprotein MAG to neural cell cultures. J Neurosci Res 25: 1-13.

Salzer JL (1997) Clustering sodium channels at the node of Ranvier: close encounters of the axon-glia kind. Neuron 18:843-846.

Salzer JL & Bunge RP (1980) Studies of Schwann cell proliferation. I. An analysis in tissue culture of proliferation during development, Wallerian degeneration, and direct injury. J Cell Biol 84: 739-752.

Salzer JL, Holmes WP & Colman DR (1987) The amino acid sequences of the myelin-associated glycoproteins: homology to the immunoglobulin gene superfamily. J Cell Biol 104: 957-965.

Sanchez I, Hassinger L, Paskevich PA, Shine HD & Nixon RA (1996) Oligodendroglia regulate the regional expansion of axon caliber and local accumulation of neurofilaments during development independently of myelin formation. J Neurosci 16: 5095-5105.

Sandri C, van Buren JM & Akert K (1982) Membrane morphology of the vertebrate nervous system. Prog Brain Res 46: 201-265.

Sanger F, Nicklen S & Coulson AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74: 5463-5467.

Saqr HE, Pearl DK, Yates AJ, van Brocklyn J, Bremer EG & Yates AJ (1993) A review and predictive models of ganglioside uptake by biological membranes. J Neurochem 61: 395-411.

93

Sargiacomo M, Sudol M, Tang Z & Lisanti MP (1993) Signal transducing molecules and glycosyl-phosphatidylinositol-linked proteins form a caveolin-rich insoluble complex in MDCK cells. J Cell Biol 122: 789-807.

Sasagasako N, Toda K, Hollis M & Quarles RH (1996) Myelin gene expression in immortalized Schwann cells: relationship to cell density and proliferation. J Neurochem 66: 1432-1439.

Sato S, Yanagisawa K & Miyatake T (1984) Conversion of myelin-associated glycoprotein (MAG) to a smaller derivative by calcium activated neutral protease (CANP)-like enzyme in myelin and inhibition by E-64 analogue. Neurochem Res 9: 629-635.

Schachner M & Bartsch U (2000) Multiple functions of the myelin-associated glycoprotein MAG (siglec-4a) in formation and maintenance of myelin. Glia 29: 154-165.

Scherer SS (1996) Molecular specializations at nodes and paranodes in peripheral nerve. Microsc Res Tech 34: 452-461.

Scherer SS, Bone LJ, Deschênes SM, Fischbeck K & Balice-Gordon RJ (1997) The role of the gap junction protein connexin32 in the myelin sheath. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 83-102.

Scherer SS, Deschenes SM, Xu YT, Grinspan JB, Fischbeck KH & Paul DL (1995a) Connexin32 is a myelin-related protein in the PNS and CNS. J Neurosci 15: 8281-8294.

Scherer SS, Wang DY, Kuhn R, Lemke G, Wrabetz L & Kamholz J (1994) Axons regulate Schwann cell expression of the POU transcription factor SCIP. J Neurosci 14: 1930-1942.

Scherer SS, Xu YT, Bannerman PG, Sherman DL & Brophy PJ (1995b) Periaxin expression in myelinating Schwann cells: modulation by axon-glial interactions and polarized localization during development. Development 121: 4265-4273.

Schlessinger J (1988) Signal transduction by allosteric receptor oligomerization. Trends Biochem Sci 13: 443-447.

Schmitt FO, Bear RS & Palmer KJ (1941) X-ray diffraction studies on the structure of the nerve myelin sheath. J Cell Comp Physiol 18: 31-41.

Schwann T (1839) Mikroskopische Untersuchungen über der Übereinstimmung in der Struktur und dem Wachsthum der Thiere und Pflanzen. Sanderschen Buchhandlung, Berlin.

Schwann T (1847) Microscopical researches into the accordance in the structure and growth of animals and plants. The Sydenham Society, London.

Schwechheimer K, Gass P & Berlet HH (1992) Expression of oligodendroglia and Schwann cell markers in human nervous system tumors. An immunomorphological study and western blot analysis. Acta Neuropathol 83: 283-291.

Seboun E, Oksenberg JR, Rombos A, Usuku K, Goodkin DE, Lincoln RR, Wong M, Pham-Dinh D, Boesplug-Tanguy O, Carsique R, Fitoussi R, Gartioux C, Reyes C, Ribierre F, Faure S, Fizames C, Gyapay G, Weissenbach J, Dautigny A, Rimmler JB, Garcia ME, Pericak-Vance MA, Haines JL & Hauser SL (1999) Linkage analysis of candidate myelin genes in familial multiple sclerosis. Neurogenetics 2: 155-162.

Sedzik J, Kotake Y & Uyemura K (1999) Structural and biochemical analysis of peripheral myelin glycoproteins (abstract). J Neurochem 73: S157.

Seilheimer B, Persohn E & Schachner M (1989) Antibodies to the L1 adhesion molecule inhibit Schwann cell ensheathment of neurons in vitro. J Cell Biol 109: 3095-3103.

Selinfreund RH, Barger SW, Welsh MJ & van Eldik LJ (1990) Antisense inhibition of glial S100 beta production results in alterations in cell morphology, cytoskeletal organization, and cell proliferation. J Cell Biol 111: 2021-2028.

Shapiro L, Doyle JP, Hensley P, Colman DR & Hendrickson WA (1996) Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17: 435-449.

Sheikh KA, Sun J, Liu Y, Kawai H, Crawford TO, Proia RL, Griffin JW & Schnaar RL (1999) Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc Natl Acad Sci USA 96: 7532-7537.

Sheu FS, Azmitia EC, Marshak DR, Parker PJ & Routtenberg A (1994) Glial-derived S100b protein selectively inhibits recombinant beta protein kinase C (PKC) phosphorylation of neuron-specific protein F1/GAP43. Mol Brain Res 21: 62-66.

94

Sheu FS, Huang FL & Huang KP (1995) Differential responses of protein kinase C substrates (MARCKS, neuromodulin, and neurogranin) phosphorylation to calmodulin and S100. Arch Biochem Biophys 316: 335-342.

Simons K & Ikonen E (1997) Functional rafts in cell membranes. Nature 387: 569-572.Simons K & Wandinger-Ness A (1990) Polarized sorting in epithelia. Cell 62: 207-210.Smith DB & Johnson KS (1988) Single-step purification of polypeptides expressed in Escherichia

coli as fusions with glutathione S-transferase. Gene 67:31-40.Snipes GJ & Suter U (1995) Molecular anatomy and genetics of myelin proteins in the peripheral

nervous system. J Anat 186: 483-494.Snipes GJ, Suter U & Shooter EM (1993) Human peripheral myelin protein-22 carries the

L2/HNK-1 carbohydrate epitope. J Neurochem 61: 1961-1964.Soderstrom M, Link H, Sun JB, Fredrikson S, Wang ZY & Huang WX (1994) Autoimmune T cell

repertoire in optic neuritis and multiple sclerosis: T cells recognising multiple myelin proteins are accumulated in cerebrospinal fluid. J Neurol Neurosurg Psychiatry 57: 544-551.

Spagnol G, Williams M, Srinivasan J, Golier J, Bauer D, Lebo RV & Latov N (1989) Molecular cloning of human myelin-associated glycoprotein. J Neurosci Res 24: 137-142.

Spreca A, Rambotti MG, Rende M, Saccardi C, Aisa MC, Giambanco I & Donato R (1989) Immunocytochemical localization of S-100b protein in degenerating and regenerating rat sciatic nerves. J Histochem Cytochem 37: 441-446.

Stahl N, Harry J & Popko B (1990) Quantitative analysis of myelin protein gene expression during development in the rat sciatic nerve. Mol Brain Res 8: 209-212.

Stamenkovic I & Seed B (1990) The B-cell antigen CD22 mediates monocyte and erythrocyte adhesion. Nature 345: 74-77.

Stebbins JW, Jaffe H, Fales HM & Möller JR (1997) Determination of a native proteolytic site in myelin-associated glycoprotein. Biochemistry 36: 2221-2226.

Steinman L (1996) Multiple sclerosis: A coordinated immunological attack against myelin in the central nervous system. Cell 85: 299-302.

Steinmetz MO, Stoffler D, Hoenger A, Bremer A & Aebi U (1997) Actin: from cell biology to atomic detail. J Struct Biol 119: 295-320.

Stephens RE (1982) Equimolar heterodimers in microtubules. J Cell Biol 94: 263-270.Sternberger NH, Iyotoma Y, Kies MW & Webster HdeF (1978) Immunocytochemical method to

identify basic protein in myelin forming oligodendrocytes of newborn rat CNS. J Neurocytol 7: 251-263.

Sternberger NH, Quarles RH, Itoyama Y & Webster HD (1979) Myelin-associated glycoprotein demonstrated immunocytochemically in myelin and myelin-forming cells of developing rat. Proc Natl Acad Sci USA 76: 1510-1514.

Stevens B, Tanner S & Fields DR (1998) Control of myelination by specific patterns of neural impulses. J Neurosci 18: 9303-9311.

Stoffel W & Bosio A (1997) Myelin glycolipids and their functions. Curr Opin Neurobiol 7: 654-661.

Stys PK, Lehning E, Saubermann AJ & LoPachin RMJr (1997) Intracellular concentrations of major ions in rat myelinated axons and glia: Calculations based on electron probe X-ray microanalyses. J Neurochem 68: 1920-1928.

Su Y, Brooks DG, Li L, Lepercq J, Trofatter JA, Ravetch JV & Lebo RV (1993) Myelin protein zero gene mutated in Charcot-Marie-tooth type 1B patients. Proc Natl Acad Sci USA 90: 10856-10860.

Suter DM & Forscher P (1998) An emerging link between cytoskeletal dynamics and cell adhesion molecules in growth cone guidance. Curr Opin Neurobiol 8: 106-116.

Suter U (1997) Myelin: keeping nerves well wrapped up. Curr Biol 7: R21-23.Suter U, Moskow JJ, Welcher AA, Snipes GJ, Kosaras B, Sidman RL, Buchberg AM & Shooter

EM (1992a) A leucine-to-proline mutation in the putative first transmembrane domain of the 22-kDa peripheral myelin protein in the trembler-J mouse. Proc Natl Acad Sci USA 89: 4382-4386.

Suter U & Snipes GJ (1995) Biology and genetics of hereditary motor and sensory neuropathies. Annu Rev Neurosci 18: 45-75.

95

Suter U, Welcher AA, Ozcelik T, Snipes GJ, Kosaras B, Francke U, Billings-Gagliardi S, Sidman RL Shooter EM (1992b) Trembler mouse carries a point mutation in a myelin gene. Nature 356: 241-244.

Swiatek PJ & Gridley T (1993) Perinatal lethality and defects in hindbrain development in mice homozygous for a targeted mutation of the zinc finger gene Krox20. Genes Dev 7: 2071-2084.

Tang S, Shen YJ, DeBellard ME, Mukhopadhyay G, Salzer JL, Crocker PR & Filbin MT (1997) Myelin-associated glycoprotein interacts with neurons via a sialic acid binding site at ARG118 and a distinct neurite inhibition site. J Cell Biol 138: 1355-1366.

Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F & Higgins DG (1997) The CLUSTAL_X windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25: 4876-4882.

Tobler AR, Notterpek L, Naef R, Taylor V, Suter U & Shooter EM (1999) Transport of Trembler-J mutant peripheral myelin protein 22 is blocked in the intermediate compartment and affects the transport of the wild-type protein by direct interaction. J Neurosci 19: 2027-2036.

Toda K, Small JA, Goda S & Quarles RH (1994) Biochemical and cellular properties of three immortalized Schwann cell lines expressing different levels of the myelin-associated glycoprotein. J Neurochem 63: 1646-1657.

Tooth HH (1886) The peroneal type of progressive muscular atrophy. HK Lewis, London.Topilko P, Schneider-Maunoury S, Levi G, Baron-Van Evercooren A, Chennoufi AB, Seitanidou

T, Babinet C & Charnay P (1994) Krox-20 controls myelination in the peripheral nervous system. Nature 371: 796-799.

Trapp BD (1990) Myelin-associated glycoprotein: Location and potential functions. Ann NY Acad Sci 605: 29-43.

Trapp BD, Andrews SB, Cootauco C & Quarles R (1989b) The myelin-associated glycoprotein is enriched in multivesicular bodies and periaxonal membranes of actively myelinating oligodendrocytes. J Cell Biol 109: 2417-2426.

Trapp BD, Andrews SB, Wong A, O'Connell M & Griffin JW (1989a) Co-localization of the myelin-associated glycoprotein and the microfilament components, F-actin and spectrin, in Schwann cells of myelinated nerve fibres. J Neurocytol 18: 47-60.

Trapp BD, Dubois-Dalcq M & Quarles RH (1984a) Ultrastructural localization of P2 protein in actively myelinating rat Schwann cells. J Neurochem 43: 944-948.

Trapp BD, Hauer P & Lemke G (1988) Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells. J Neurosci 8: 3515-3521.

Trapp BD, Kidd GJ, Hauer P, Mulrenin E, Haney CA & Andrews SB (1995) Polarization of myelinating Schwann cell surface membranes: role of microtubules and the trans-Golgi network. J Neurosci 15: 1797-1807.

Trapp BD & Quarles RH (1982) Presence of the myelin-associated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J Cell Biol 92: 877-882.

Trapp BD, Quarles RH & Suzuki K (1984b) Immunocytochemical studies of quaking mice support a role for the myelin-associated glycoprotein in forming and maintaining the periaxonal space and periaxonal cytoplasmic collar of myelinating Schwann cells. J Cell Biol 99: 594-606.

Tropak MB, Johnson PW, Dunn RJ & Roder JC (1988) Differential splicing of MAG transcripts during CNS and PNS development. Brain Res 464: 143-155.

Tropak MB & Roder JC (1997) Modulation of MAG and SMP binding by sialylation (Abstract). In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 366-367.

Tsang D, Tsang YS, Ho WK & Wong RN (1997) Myelin basic protein is a zinc-binding protein in brain: possible role in myelin compaction. Neurochem Res 22: 811-819.

Ueki K, Ramaswamy S, Billings SJ, Mohrenweiser HW & Louis DN (1997) ANOVA, a putative astrocytic RNA-binding protein gene that maps to chromosome 19q13.3. Neurogenetics 1: 31-36.

Umemori H, Kadowaki Y, Hirosawa K, Yoshida Y, Hironaka K, Okano H & Yamamoto T (1999) Stimulation of myelin basic protein gene transcription by Fyn tyrosine kinase for myelination. J Neurosci 19: 1393-1397.

96

Umemori H, Sato S, Yagi T, Aizawa S & Yamamoto T (1994) Initial events of myelination involve Fyn tyrosine kinase signalling. Nature 367: 572-576.

Uyemura K, Horie K, Kitamura K, Suzuki M & Uehara S (1979) Developmental changes of myelin proteins in the chick peripheral nerve. J Neurochem 32: 779-788.

Uyemura K, Yoshimura K, Suzuki M & Kitamura K (1984) Lipid binding activities of the P2 protein in peripheral nerve myelin. Neurochem Res 9: 1509-1514.

Vabnick I, Novakovic SD, Levinson SR, Schachner M & Shrager P (1996) The clustering of axonal sodium channels during development of the peripheral nervous system. J Neurosci 16: 4914-4922.

Valentijn LJ, Baas F, Wolterman RA, Hoogendijk JE, Bosch NHA, Zorn I, Gabreels-Festen AAWM, deVisser M & Bolhuis PA (1992) Identical point mutations of PMP-22 in Trembler-J mouse and Charcot-Marie-Tooth disease type 1a. Nature Genet 2: 288-291.

Vallat JM, Sindou P, Garbay B, Preux PM, Anani T, Richard L & Diot M (1999) Expression of myelin proteins in the adult heterozygous Trembler mouse. Acta Neuropathol 98: 281-287.

Vallee BL & Falchuk KH (1993) The biological basis of zinc physiology. Physiol Rev 73: 79-118.van Rossum D, Kuhse J & Betz H (1999) Dynamic interaction between soluble tubulin and C-

terminal domains of N-methyl-D-aspartate receptor subunits. J Neurochem 72: 962-973.Vartanian T, Szuchet S & Dawson G (1992) Oligodendrocyte-substratum adhesion activates the

synthesis of specific lipid species involved in cell signaling. J Neurosci Res 32: 69-78.Vartanian T, Szuchet S, Dawson G & Campagnoli AT (1986) Oligodendrocyte adhesion activates

protein kinase C-mediated phosphorylation of myelin basic protein. Science 234: 1395-1398.Vinson CR, LaMarco KL, Johnson PF, Landschulz WH & McKnight SL (1988) In situ detection of

dequence-specific DNA binding activity specified by a recombinant bacteriophage. Genes Dev 2: 801-806.

Virchow R (1846) Über das granulierte Aussehen der Wandungen der Gehirnventrikel. Allg Z Psychiatr 3: 242-255.

Virchow R (1854) Über das ausgebreitete Vorkommen einer dem Nervenmark analogen Substanz in den tierischen Geweben. Virchows Archiv für pathologische Anatomie 6: 562.

von Heijne G (1989) Control of topology and mode of assembly of a polytopic membrane protein by positively charged residues. Nature 341: 456-458.

Voshol H, van Zuylen CWEM, Orberger G, Vliegenthart JFG & Schachner M (1996) Structure of the HNK-1 epitope on bovine peripheral myelin glycoprotein P0. J Biol Chem 271: 22957-22960.

Vuorinen VS & Röyttä M (1990) Taxol-induced neuropathy after nerve crush: long-term effects on Schwann and endoneurial cells. Acta Neuropathol 79: 653-662.

Vuorinen V, Röyttä M & Raine CS (1989) The long-term cellular response to taxol in peripheral nerve: Schwann cell and endoneurial cell changes. J Neurocytol 18: 785-794.

Walikonis RS & Poduslo JF (1997) Role of cAMP during the process of demyelination and remyelination in peripheral nerve. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 129-135.

Waxman SG, Ritchie JM (1993) Molecular dissection of the myelinated axon. Ann Neurol 33: 121-136.

Weber RG, Sabel M, Reifenberger J, Sommer C, Oberstraß J, Reifenberger G, Kiessling M & Cremer T (1996) Characterization of genomic alterations associated with glioma progression by comparative genomic hybridization. Oncogene 13: 983-994.

Webster HD (1971) The geometry of peripheral myelin sheaths during their formation and growth in rat sciatic nerves. J Cell Biol 48: 348-367.

Weinberg HJ & Spencer PS (1976) Studies on the control of myelinogenesis. II. Evidence for neuronal regulation of myelin production. Brain Res 113: 363-378.

Welcher AA, Suter U, De Leon M, Snipes GJ & Shooter EM (1991) A myelin protein is encoded by the homologue of a growth arrest-specific gene. proc Natl Acad Sci USA 88: 7195-7199.

Wiche G & Cole RD (1976) Reversible in vitro polymerization of tubulin from a cultured cell line (rat glial cell clone C6). Proc Natl Acad Sci USA 73: 1227-1231.

97

Wilson R & Brophy PJ (1989) Role for the oligodendrocyte cytoskeleton in myelination. J Neurosci Res 22: 439-448.

Windebank AJ, Wood P, Bunge RP & Dyck PJ (1985) Myelination determines the caliber of dorsal root ganglion neurons in culture. J Neurosci 5: 1563-1569.

Winzler RJ (1970) Carbohydrates in cell surfaces. Int Rev Cytol 29: 77-125.Wong MH & Filbin MT (1994) The cytoplasmic domain of the myelin P0 protein influences the

adhesive interactions of its extracellular domain. J Cell Biol 126: 1089-1097.Wong MH & Filbin MT (1996) Dominant-negative effect on adhesion by myelin Po protein

truncated in its cytoplasmic domain. J Cell Biol 134: 1531-1541.Wood JG & Engel EL (1976) Peripheral nerve glycoproteins and myelin fine structure during

development of rat sciatic nerve. J Neurocytol 5: 605-615.Wood PM & Bunge RP (1975) Evidence that sensory axons are mitogenic for Schwann cells.

Nature 342: 430-433.Wood P, Moya F, Eldridge C, Owens G, Ranscht B, Schachner M, Bunge M & Bunge R (1990)

Studies of the initiation of myelination by Schwann cells. Ann N Y Acad Sci 605: 1-14.Wright PE & Dyson HJ (1999) Intrinsically unstructured proteins: Re-assessing the protein

structure-function paradigm. J Mol Biol 293: 321-331.Yanagisawa K & Quarles RH (1986) Jimpy mice: quantitation of myelin-associated glycoprotein

and other proteins. J Neurochem 47: 322-325.Yanagisawa K, Quarles RH, Johnson D, Brady RO & Whitaker JN (1985) A derivative of myelin-

associated glycoprotein in cerebrospinal fluid of normal subjects and patients with neurological disease. Ann Neurol 18: 464-469.

Yang H, Xiao ZC, Becker B, Hillenbrand R, Rougon G & Schachner M (1999) Role for myelin-associated glycoprotein as a functional tenascin-R receptor. J Neurosci Res 55: 687-701.

Yang LJ, Zeller CB, Shaper NL, Kiso M, Hasegawa A, Shapiro RE & Schnaar RL (1996) Gangliosides are neuronal ligands for myelin-associated glycoprotein. Proc Natl Acad Sci USA 93: 814-818.

Yates AJ & Wherrett JR (1974) Changes in the sciatic nerve of the rabbit and its tissue constituents during development. J Neurochem 23: 993-1003.

Yim SH, Toda K, Goda S & Quarles RH (1995) Comparison of the phosphorylation of myelin-associated glycoprotein in cultured oligodendrocytes and Schwann cells. J Mol Neurosci 6: 63-74.

Yin X, Crawford TO, Griffin JW, Tu P, Lee VM, Li C, Roder J & Trapp BD (1998) Myelin-associated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J Neurosci 18: 1953-1962.

Yin X & Trapp BD (1997) Myelin-associated glycoprotein: Ligand and/or receptor. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 27-36.

Yong WH, Ueki K, Chou D, Reeves SA, von Deimling A, Gusella JF, Mohrenweiser HW, Buckler AJ & Louis DN (1995) Cloning of a highly conserved human protein serine-threonine phosphatase gene from the glioma candidate region on chromosome 19q13.3. Genomics 29: 533-536.

Yong VW & Oh LYS (1997) Protein kinase C regulates process formation by oligodendrocytes. In: Juurlink BHJ, Devon RM, Doucette JR, Nazarali AJ, Schreyer DJ & Verge VMK (eds) Cell biology and pathology of myelin: Evolving biological concepts and therapeutic approaches. Plenum Press, New York, p. 173-185.

Yoshimura T, Kobayashi T & Goto I (1992) Protein kinase C in rat brain myelin. Neurochem Res 17: 1021-1027.

Zabrenetzky V, Krygier-Brevart V & Spencer PS (1984) Cyclic AMP-stimulated protein kinase activity in rabbit peripheral myelin. Neurochem Res 9: 121-132.

Zeelen JP, Hiltunen JK, Ceska TA & Wierenga RK (1994) Crystallization experiments with 2-enoyl-CoA hydratase, using an automated "fast-screening" crystallization protocol. Acta Cryst D50: 443-447.

98

Zhang Z, Devarajan P, Dorfman AL, Morrow JS (1998) Structure of the ankyrin-binding domain of alpha-Na,K-ATPase. J Biol Chem 273: 18681-18684.

Zimmer DB & van Eldik LJ (1988) Levels and distribution of the calcium-modulated proteins S100 and calmodulin in rat C6 glioma cells. J Neurochem 50: 572-579.

Zorick TS & Lemke G (1996) Schwann cell differentiation. Curr Opin Cell Biol 8: 870-876.

99