Truncated Tau with the Fynbinding domain and ... - Wiley Online Library

JOURNAL OF NEUROCHEMISTRY

| 2008 | 107 | 351–360

doi: 10.1111/j.1471-4159.2008.05600.x

,1

*Department of Medical Science, School of Veterinary Medicine, University of Wisconsin, Madison, Wisconsin, USA Department of Neurology, Northwestern University, Chicago, Illinois, USA

Abstract The mechanisms underlying developmental myelination have therapeutic potential following CNS injury and degeneration. We report that transplanted central glial (CG)-4 cells had a diminished myelinating capacity in myelin-deficient (md) rats when cells express a mutated form of Tau (Tau [688]), which binds Fyn but not the microtubules. In the brain of the md rats, Tau [688]-transfected CG-4 cells displayed a decrease in cellular process outgrowth and myelination; in the spinal cord the extent of myelination rostral and caudal to the injection site was decreased. In contrast, control Tau [605]-transfected CG-

4 cells formed long cellular processes and substantial areas of myelin both in the brain and spinal cord. In culture, Tau [688]transfected CG-4 cells displayed a decrease in cellular process outgrowth, and Fyn localized largely in the cell body, not the processes. Thus, Tau in oligodendrocytes plays a key role in myelination, and a functional Tau-Fyn interaction might have therapeutic potential during demyelination and myelin repair following CNS injury and degeneration. Keywords: demyelination, Fyn, microtubules, multiple sclerosis, oligodendrocytes, targeting, Tau. J. Neurochem. (2008) 107, 351–360.

Oligodendrocytes (OLGs), the myelin-forming cells of the central nervous system (CNS), rely on cellular process outgrowth to ensheathe and myelinate nerve axons. Developmental myelination requires that bipolar OLG progenitors mature into myelin-forming cells with extensive arborizations of their cell processes. Fyn, a member of the Src family of non-receptor protein tyrosine kinases, is important for the morphological maturation of OLG progenitor cells, and the inhibition and/or absence of Fyn alter(s) cellular process outgrowth and myelination (Umemori et al. 1994; Osterhout et al. 1999; Sperber et al. 2001; Wolf et al. 2001). Fyn associates with integrin (Klinghoffer et al. 1999; Colognato et al. 2004; Liang et al. 2004), and region-specific defects are observed in the Fyn- and laminin-2-minus mice. Fynminus mice have a severe myelin deficit in forebrain at all ages (Sperber et al. 2001), whereas cervical spinal cord has no decrease in myelin content, number of OLGs, or number of myelinated fibers (Sperber et al. 2001). Also, the laminin2-minus mice have myelin deficit in the brain not the spinal cord (Chun et al. 2003). Recently, Lee et al. (2006) showed that transgenic mice that express a dominant-negative beta1 Integrin protein (lacking the C-terminal tail) have hypomyelinated axons in spinal cords and optic nerves with a

significant increase in the number of unmyelinated axons within the spinal cord and optic nerves. In contrast, the corpus callosum has no myelin defects, whereas during remyelination of the corpus callosum the actual percentage of myelinated axons is reduced (Lee et al. 2006). These previous studies provided evidence that integrin-dependent pathways are important during myelination; however, the conditional ablation of the beta 1-integrin gene in oligodendroglial cells did not affect CNS myelination and remyelination (Benninger et al. 2006). Therefore, these results indicate that both integrin-dependent and -independent

Received May 22, 2008; revised manuscript received July 14, 2008; accepted July 21, 2008. Address correspondence and reprint requests to Patrizia LoPresti, Department of Neurology, Northwestern University, Tarry Building 13-715, 300 East Superior Street, Chicago, IL 60611, USA. E-mail: [email protected] 1 The present address of Abdelmadjid Belkadi is the Case School of Medicine, Department of Neurosciences and the Center for Translational Neuroscience, 2109 Adelbert Rd, Cleveland, OH 44106, USA. Abbreviations used: MAP, microtubule-associated-protein; OLG, oligodendrocytes; PFA, paraformaldehyde; PLP, proteolipid protein; SC, spinal cord.

2008 The Authors Journal Compilation 2008 International Society for Neurochemistry, J. Neurochem. (2008) 107, 351–360

351

352 | A. Belkadi and P. LoPresti

signals direct the making of myelin; however the interplay among multiple signaling pathways is still obscure. The cytoskeleton modulates events underlying cellular process outgrowth and the synthesis of myelin, so it is likely that the regulatory mechanisms of the cytoskeleton and myelination overlap. Recent evidence supports this possibility. In fact, cytoskeletal function is regulated by many factors, including the microtubule-associated proteins (MAPs) (Maccioni and Cambiazo 1995), and work by Lee et al. (2005) has shown that 2¢-3¢-cyclic nucleotide 3¢-phosphohydrolase (CNP) protein, a myelin-related protein, acts as a MAP, which further reinforces the importance of MAPs in myelinogenesis. Among MAPs present in OLGs (Fischer et al. 1990; Vouyiouklis and Brophy 1993; LoPresti et al. 1995; Muller et al. 1997; Shafit-Zagardo et al. 1999), the microtubule-associated protein Tau was thought to be linked with myelin formation (LoPresti et al. 1995). Tau is present in cultured neonatal and adult OLGs, and in OLGs in situ (LoPresti et al. 1995, 2001; LoPresti 2002). The microtubule-associated protein Tau, a developmentally regulated protein in the CNS, consists of two and six isoforms in neonatal and adult brain, respectively. These protein isoforms arise from alternatively spliced and post-translational modifications (Goedert et al. 1991; Mandelkow et al. 1995; LoPresti 2002). Tau isoforms with 3 and 4 repeats in the microtubule-binding region occur in neonatal and adult CNS, respectively (Kosik et al. 1989; LoPresti 2002). In addition, Tau isoforms with no, one, or two inserts of 31 amino acids are also present (Fig. 1) (Collet et al. 1997; LoPresti 2002). Novel Tau isoforms without the microtubule-binding domain have also been reported (Luo et al. 2004). Tau was found to associate with the cell membrane of neuronal cells (Brandt et al. 1995). Furthermore, Tau binds Fyn both in neuronal and oligodendroglial cells (Brandt et al. 1995; Klein et al. 2002), and Fyn phosphorylates Tau in neuronal cells (Lee et al. 2004). The main Fyn SH3 domain-binding PXXP motif (Pro, Lys, Ser, Pro) in adult rat Tau is at the residues 223-226 (Kosik et al. 1989; Lee et al. 1998). The lack of this motif reduces the amount of Fyn bound to Tau by over 90% (Lee et al. 1998). The findings that Tau interacts with Fyn (Lee et al. 1998) and that this interaction facilitates the outgrowth of OLG processes (Klein et al. 2002) reinforce a putative role for Tau in myelination. Tau has a role in the outgrowth of cell processes (LoPresti et al. 2001; Klein et al. 2002) and intracellular transport regulation (Ebneth et al. 1998; Dixit et al. 2008). Thus it is likely that Tau has a key function during myelin formation and repair. Previous work indicated that the Tau-Fyn interaction facilitates cellular process outgrowth in cultured oligodendrocytes (Klein et al. 2002), whereas the current study addresses the role of the Tau-Fyn interaction in in vivo myelination using the md rat, an animal model of demyelination. The md rats have a point mutation in exon III of the proteolipid protein (PLP) gene. No PLP can be detected in md

rats, and most axons in the CNS lack myelin (Dentinger et al. 1982). The mutation is inherited in an X-linked recessive mode (Boison and Stoffel 1989), and affected males develop a severe axial body tremor often combined with seizures around day 10 and die approximately 3 weeks after birth (Boison and Stoffel 1989). Previous work showed that CG-4 cells proliferated, migrated, and formed myelin when transplanted in the md rat (Tontsch et al. 1994; Franklin et al. 1996; Espinosa de los Monteros et al. 1997). Here we test whether the myelinating capacity of Tau [688]-transfected CG-4 cells is altered when these cells are transplanted in the brain and spinal cords of the md rats. Experimental Tau-[688] inhibits the link of endogenous Tau-Fyn with the cytoskeleton. We report that Tau [688]-transfected CG-4 cells failed to extend cellular processes and formed limited areas of myelin in the brain. In addition we show that in the spinal cord myelination rostral and caudal to the injection site was reduced. Thus the current study identifies unexplored effects of a mutated form of Tau in oligodendrocytes and suggests that a functional TauFyn interaction in OLGs is required during developmental myelination.

Materials and methods Vectors The constructs used in this study are based on the adult rat Tau cDNA plasmid (Fig. 1), and the major Fyn SH3 binding site is at residues 223-226 (Fig. 1). We have maintained the terminology of

Fig. 1 Diagram indicates mutated Tau constructs used in this study. The truncated Tau constructs used in this study and by Klein et al. (2002) are based on the adult rat Tau with two amino inserts and four repeats in the microtubule-binding domain (Kosik et al. 1989). Klein et al. (2002) used truncated Tau [227] and [223]. Truncated Tau [227] binds Fyn, and the Fyn bound to Tau [227] is active (Klein et al. 2002). Control truncated Tau [223] does not bind Fyn. Tau [688] and [605] refer to the adult rat Tau cDNA plasmid (Brandt et al. 1995). The numbers in accordance to the Aa are also indicated, i.e., 233 Aa for Tau [688], 202 Aa for Tau [605] (Brandt et al. 1995). The truncated Tau constructs lack the microtubule-binding region. The PXXP (Pro, Lys, Ser, Pro) motif at residues 223-226 Aa of adult rat Tau is the major Fyn SH3 binding site (Lee et al. 1998). The lack of this motif reduces the amount of Fyn bound to Tau by over 90% (Lee et al. 1998). Grey boxes indicate four repeats in the microtubule-binding region. Blue boxes indicate two inserts of 31 amino acids each at the amino terminal end.


The Tau-Fyn interaction in myelin formation | 353

Tau [688] and [605] in accordance with the paper that originally described these constructs (Brandt et al. 1995) and we have also indicated the corresponding amino acid, i.e., Tau [233] for Tau [688] and Tau [202] for Tau [605] (Fig. 1). Tau [688] and [605] constructs, received from Dr. Gloria Lee (Brandt et al. 1995), cannot bind the microtubules because they lack the microtubule-binding domain (Fig. 1). Only Tau [688] binds Fyn since it has the main Fyn SH3 domain-binding PXXP motif (Pro, Lys, Ser, Pro) (Kosik et al. 1989; Lee et al. 1998). In contrast, Tau [605] does not bind Fyn since it lacks this domain. The experimental and control mutated Tau constructs used by Klein et al. (2002) are also indicated (Fig. 1). In addition, Tau [688] and Tau [605] constructs had the sequence DYKDDDDK (FLAG) fused to the amino-terminal end as an epitope tag (Brandt et al. 1995). Transfection and measurement of cellular processes in Tau-transfected cells CG-4 cells are non-transformed glial precursor cells isolated from the CNS of neonatal rat brains (Louis et al. 1992). CG-4 glial precursor cells were cultured in the presence of growth factors, whereas CG-4 cells differentiated into OLGs in the absence of growth factors (Louis et al. 1992). We received from Dr. Jean-Claude Louis (Amgen, CA) a very early passage to ensure that the cellular properties of these cells mimicked those of primary glial progenitor cells. Transfection of CG-4 cells was done as described by Chew et al. (2001) and drugresistant cells were isolated and tested for transfected Tau using FLAG immunoblots. FLAG immunoblot of Tau-transfected cell lysates (20-lg sample/lane) showed the apparent molecular weight of Tau [605] and Tau [608] in the expected range of 41 and 44 kDa, respectively (Fig. 2). Electrophoretic separation was done by sodium dodecyl sulfate polyacrylamide gel electrophoresis with 9% polyacrylamide. Transfer, immunoreaction, and detection were performed as previously described (LoPresti et al. 1995). Anti-FLAG monoclonal antibodies were obtained from Sigma. The experimental and control cells used in this study had similar expression of transfected Tau as judged by FLAG immunoblot. The comparison was made

Fig. 2 FLAG immunoblot of lysates of stably Tau-transfected CG-4 cells. The apparent molecular weight of Tau [605] and [688] was in the expected range of 41 and 44 kDa, respectively. Electrophoretic separation was done by sodium dodecyl sulfate polyacrylamide gel electrophoresis with 9% polyacrylamide. Transfer, immunoreaction, and detection were performed as previously described (LoPresti et al. 1995) (20-lg sample/lane). Anti-FLAG monoclonal antibodies were obtained from Sigma.

among cell lysates run in parallel (Fig. 2). To measure cellular processes, Tau-transfected cells were fixed on the second day of oligodendrocyte differentiation, and differential interference contrast (DIC) confocal images were taken using a Leica TCS SP2 Confocal Scanner DMRXE7 Microscope (Northwestern University Biological Imaging Facility). All major cellular branches of 32 Tau [688]transfected cells and of 28 Tau [605]-transfected cells were measured. Six fields per group were examined. The measurements were obtained from two independent experiments. Transplantation of CG-4 cells into the spinal cord and hemisphere of md rats Before transplantation, Tau-transfected CG-4 cells were detached from the Petri dishes using 0.1% trypsin and gently dissociated into a single cell with a fired bore rod Pasteur pipette. The cells were then rinsed in F-15 medium (GIBCO) and resuspended in DMEM (GIBCO) at a concentration of 75 000 cells per microliter. 18 md rats, housed at the University of Wisconsin-Madison, were used in these studies. The 7-day-old rats were divided into two groups. Each group received experimental or control Tau-transfected cells. The 7-day-old rat pups were anesthetized with isoflurane, and the skin was cleaned with 70% ethanol and Betadine. At the level of the spinal cord 12 rats (six per group) were used. The md rats were then placed in a stereo tactic apparatus and received a minimal skin incision at the thoraco-lumbar spine. The muscles were retracted and a laminectomy was performed at the thoraco-lumbar spine. 150 000 cells resuspended in 2 lL DMEM were injected in the spinal cord of the md rats. The same number of cells was used for both experimental and control Tau-transfected cells. A glass needle held by a micromanipulator was inserted into the dorsal column of newborn rats delivering 2 lL of cell suspension over a period of 10 min. The micropipette was withdrawn stepwise allowing cell diffusion. The muscles were sutured with a 5.0 filament (Ethicon, New Jersey) and the skin stitched using metallic sutures (Becton Dickinson, MD, USA). The animals were perfused 12 days later. Eight rats (four per group) were perfused with 4% paraformaldehyde (PFA) and four rats (two per group) with 2% PFA + 2% glutaraldehyde. At the level of the brain, six rats (three per group) were used. The CG-4 cells were injected at the corpus callosum of each hemisphere following the coordinates: 1 mm caudal to the bregma, 2 mm lateral to the midline and 1.5–2 mm deep in the parenchyma. Each injection site received 150 000 cells resuspended in 2 lL DMEM. The six animals (three per group) were perfused 12 days later with 4% PFA. Immunohistochemistry The rats were anesthetized with isoflurane and perfused transcardially with saline solution followed by 4% PFA in 0.1 M Sorensen’s phosphate buffer, pH 7.4. All the rats transplanted at the level of the spinal cord showed a white streak. The white streak formed by Tau [688]-transfected CG-4 cells was shorter than the one formed by Tau [605] transfected cells. Consistent length differences were observed between experimental and control transplanted rats. The myelinated segment of the cord was then trimmed and post-fixed in same fixative over night, then cryoprotected in 15% sucrose/0.1 M phosphate buffered saline. Twenty-micron thick free-floating sections were cut using a cryostat. The sections were washed several times in phosphate buffer saline with 0.3% Triton and incubated overnight at 4C with a rabbit polyclonal anti-PLP (gift from Dr. Ian R. Griffiths) at a dilution



of 1 : 100 000. The sections were then washed in PBSTand incubated with Alexa 494 donkey anti-rabbit (Invitrogen) at a dilution of 1 : 500 for 60 min at room temperature. Sections were also incubated with secondary antibody alone as a negative control. The rat brains were processed in the same way as the spinal cords. Briefly, the perfused brains were post-fixed in the same fixative over night and cryoprotected in sucrose. Twenty-micron thick free-floating coronal sections were cut using a cryostat and serially stored at 4C in 96-well plates (Becton Dickinson, New Jersey). To ensure that similar areas from experimental and control brains were sampled, the sections were picked according to the exact well number and analyzed further under a light microscope to identify comparable morphology. The sections were then processed for immunohistochemistry as described previously for the spinal cord. Antibodies Anti-Fyn rabbit polyclonal antibodies were obtained from Santa Cruz Biotechnology, Inc. and anti-alpha tyrosinated Tubulin from Accurate Scientific. Rabbit polyclonal anti-PLP antibody (gift from Dr. Ian R. Griffiths) was used at a dilution of 1 : 100 000 to stain hemispheres and spinal cord. Imaging of CNS tissues Images of brain and spinal cord tissues were visualized and captured using a Nikon microscope (Eclipse E800) equipped with a digital Spot camera (Diagnostic Instruments Inc., Sterling Heights, MI, USA) and MetaVue imaging software (Universal Imaging Corp., Downingtown, PA, USA). While capturing images, all parameters such as gain, contrast, brightness, and the positions and types of filters were set such that signals were not saturated (the brightest pixels < 255), and all images were captured using the same parameters to provide for adequate comparison. Ultrastructural analysis To investigate the ultrastructure of the myelin sheath formed by CG-4 cells, some trimmed cords were post-fixed in osmium tetroxide and embedded in Epon, as described previously (O’Connor et al. 2000). Some 1 lm sections were cut and stained with toluidine to identify the area of interest. Thin sections were cut with a Leica UCT ultramicrotome, mounted on grids, stained with uranyl acetate and lead citrate, and viewed on a JEOL 1220 electron microscope (Northwestern University Cell Imaging Facility). Immunofluorescence and confocal analysis Cells were processed for immunofluorescence as previously described (LoPresti et al. 2001). Images were acquired using a Zeiss 510 Meta laser scanning confocal microscope (Zeiss, Inc. Thornwood, NY, USA, Northwestern University Cell Imaging Facility). Data are presented as a zero degree projection of z series. First antibodies were omitted in negative controls.

Results Tau [688]-transfected CG-4 cells had shorter cellular processes and displayed Fyn largely in the cell soma To determine the role of the Tau-Fyn link with the cytoskeleton in myelination, Tau [688] and Tau [605]

constructs were stably transfected in CG-4 cells (see Materials and methods), equivalent to primary rat oligodendrocyte precursor cells. Experimental Tau [688] inhibits the link of endogenous Tau-Fyn with the cytoskeleton, whereas control Tau [605] does not. Plating of CG-4 cells in the absence of growth factors initiated oligodendrocyte differentiation, and Tau-transfected CG-4 cells were processed on the second day of OLG differentiation for cell process length. The length of the cellular processes was 24.70 ± 1.71 lm and 14.85 ± 0.69 lm in Tau [605]- and [688]transfected cells, respectively (p < 0.0001, non-parametric Kruskal-Wallis test). Tau [688]-transfected CG-4 cells displayed a 40% decrease in cellular process length. The degree of decrease in cellular process length was in the range of that found in murine OLGs (Klein et al. 2002). How Fyn distributed with respect to the microtubules in Tau-transfected cells was determined with coimmunostaining for Fyn and tyrosinated alpha-Tubulin. Tyrosinated Tubulin is present in immature microtubules, whereas detyrosinated Tubulin in mature ones (Warn et al. 1990; Ladrech and Lenoir 2002). Tyrosinated Tubulin has a critical role in the development of cellular extension, and it was previously shown that alpha-Tubulin binds Fyn in oligodendrocytes (Klein et al. 2002). Confocal imaging of immunostained cells revealed Fyn in the soma and processes of control Tau [605]-transfected cells (Fig. 3a) and largely in the soma of experimental Tau [688]-transfected cells (Fig. 3d). Immunostaining also showed Fyn in the cell nuclei. Superimposed images determined the presence of discrete Fyn-Tubulin positive structures in the soma and processes of control cells (Fig. 3c). In contrast, Fyn-Tubulin positive structures appeared to have a different pattern of intracellular distribution in experimental cells (Fig. 3f). The data of Fyn-Tubulin immunostaining suggest that the intracellular distribution of Fyn requires that the Tau-Fyn interaction connects with the cytoskeleton. Tau [688]-transfected CG-4 cells transplanted into the spinal cord of md rats had diminished myelination rostral and caudal to the injection site To address the role of the Tau-Fyn link with the cytoskeleton during myelin formation, we transplanted Tau [688]-transfected CG-4 cells into the md rats and used as control Tau [605]-transfected CG-4 cells. First, experimental and control cells were injected into the T8-9 spinal cord of 7-day-old md rats. At 12 days post-transplantation, immunocytochemistry for PLP identified myelin formed by the transplanted cells. The md rats lack PLP hence PLP can only originate from the transplanted cells. Crosssections of spinal cord showed PLP-positive cells in the dorsal columns. Low-magnification images (Fig. 4a and d) determined that injected cells survived and differentiated into PLP-positive cells; high-magnification exhibited that experimental and control cells similarly formed myelin in



(b)

(a) Fig. 3 Tau [688]-transfected CG-4 cells had shorter cellular processes and displayed Fyn largely in the cell soma. Immunolocalization of Fyn and tyrosinated alpha Tubulin in CG-4 cells that stably express Tau [605] or [688]. Confocal images of control (a–c) and experimental CG-4 cells (d–f). Zero degree projection of z series (a–f) showing Fyn (green), tyrosinated alpha Tubulin (red) and superimposed images (yellow). Control cells display Fyn in discrete entities in the soma and processes (a), whereas experimental cells display Fyn largely restricted to the cell body (d). It is also possible to appreciate Fyn in the nuclei. Superimposed images show differences in the intracellular distribution of FynTubulin positive structures in the soma and processes of control cells (c) versus experimental cells (f).

(d)

(e)

(f)

(c)

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 4 Tau [688]-transfected CG-4 cells formed myelin in the spinal cord of the md rats. Dorsal column cross-sections of the spinal cord of md rats showing distribution of Tau [688]– (a–c) and [605]– (d–f) transfected CG-4 cells transplanted in the dorsal spinal cord of md rat. Low-magnification (a, d) micrographs of PLP immunostaining show

the extent of myelin formed by grafted cells. High-magnification (b, c, e, f) micrographs show myelin formed in white (b, e) and gray matter (c, f). Note similarity in myelin formation and cellular process outgrowth in white (b, e) and gray (c, f) matter. Scale bar: a, d = 200 lm; b, c, e, and f = 50 lm.

both gray (Fig. 4c and f) and white (Fig. 4b and e) matter. Longitudinal sections of the spinal cord of the 12 transplanted rats exhibited a decrease of myelin formed by experimental versus control cells (Fig. 5), which perhaps reflects diminished migration from the injection site. In all six experimental animals, the length of the myelinated region differed from control (Fig. 5). Furthermore, myelin formed by Tau-transfected CG-4 cells

transplanted into the spinal cord of the md rats (Fig. 6c and d) showed its typical intraperiod line in both experimental and control groups (Fig. 6c and d). Therefore, Tau [688] did not prevent myelin compaction. In summary, when injected in the spinal cord, experimental cells extended cellular processes and formed a compact myelin, but the extent of myelination rostral and caudal to the injection site was decreased compared to control.



Fig. 5 The extent of myelination rostral and caudal to the injection site decreased in the spinal cord transplanted with Tau [688]-transfected CG-4 cells. In all six experimental animals the length of the myelinated region differed from control. Longitudinal section of the spinal cords of the md rats showing myelin (white streak) formed by experimental (a) and control (b) CG-4 cells. Myelin can be detected as a white streak since the spinal cord of the md rat is semitranslucent due to lack of myelin. White arrows indicate the site of cell injection. Circles delimit the areas of substantial myelin. Thus, control CG-4 cells formed a longer myelinated area than experimental cells. Scale bar = 1 mm.

Fig. 6 Tau [688]-transfected CG-4 cells were able to compact myelin. Low-magnification electron micrographs of myelinated axons in the spinal cord of the md rats after transplantation of CG-4 cells transfected with Tau [688] (a) or Tau [605] (b). Myelinated axons were present in both experimental and control cells. Bar = 2 lm. Myelin shows typical intraperiod line in electron micrographs of a myelinated axon in the spinal cord of the md rat 12 days following transplantation of experimental (c) and control (d) CG-4 cells. Scale bar: c, d = 0.1 lm.

Tau [688]-transfected CG-4 cells transplanted into the brain of md rats failed to extend cellular processes and formed limited areas of myelin We next transplanted Tau-transfected CG-4 cells into the brain of the md rats to investigate the role of the Tau-Fyn link with the cytoskeleton during myelination of the brain. The brain of all six md rats surviving the transplantation had a

successful transplantation by exhibiting a white streak along the corpus callosum. The newly formed myelin was then confirmed with immunohistochemistry. The difference between experimental and control cells was consistent among all the experiments [Correction added on 8th September 2008, after first online publication : sentence amended.]. Frozen sections of brain tissue stained with PLP showed that control cells (Fig. 7c) distributed along a longer distance than experimental cells (Fig. 7a); high-magnifications (b, d) of corresponding low-magnification panels (a, c) showed obvious differences in the extent of myelination. Further, highmagnification micrographs show that transplanted control cells had long processes and extensive myelin (Fig. 8c and d), whereas transplanted experimental cells had few, short processes and scarce myelin (Fig. 8a and b). In summary, in the brain of the md rats, experimental CG4 oligodendrocyte cells fail to extend cellular processes and form substantial areas of myelin.

Discussion Tau in oligodendrocytes plays a key role in myelination The cytoskeleton of myelin-forming cells is a key component in the development and function of these cells. Previous studies indicate that the Tau-Fyn interaction facilitates cellular processes outgrowth in cultured oligodendrocytes (Klein et al. 2002). Here we demonstrate that the Tau-Fyn interaction has a role in developmental myelination. CG-4 cells were stably transfected with control and experimental truncated Tau and transplanted in the brain and spinal cord of the md rats. In the brain, experimental cells have a severe defect in their ability to extend cellular processes and form myelin; in the spinal cord the cells extend cellular processes but the extent of myelination rostral and caudal to the injection site was decreased compared to control. Cultured experimental cells display shorter cellular process length. In addition experimental cells have Fyn largely in the cell body, whereas control cells distribute Fyn both in the cell body and the processes. Taken together these results demonstrate a direct effect of Tau in OLGs in in vivo myelination. During myelination Fyn activation largely occurs following integrin activation (Colognato et al. 2004), and the presence of mutated Tau [688] could affect Fyn-dependent signaling. Tau [688] would most likely inhibit the binding of the (endogenous) Tau-Fyn interaction with the cytoskeleton, which would result in a block of cytoskeleton-dependent signaling pathways. Previous studies demonstrate regionspecific variations in myelination in Fyn- and laminin-minus mice (Sperber et al. 2001; Chun et al. 2003) with the decrease of myelination being in the brain, not the spinal cord. Neither Fyn- nor laminin-minus mice had a decrease of myelination along the spinal cord. In contrast, Lee et al. (2006) have shown that mice with a beta 1 integrin gene



(a)

(b)

(c)

(d)

Fig. 7 Reduced myelin in the brain of the md rats transplanted with Tau [688]-transfected CG-4 cells. PLP immunostaining of experimental and control CG-4 cells transplanted in the brain. Experimental Tau [688]-transfected CG-4 cells formed a limited area of myelination (a) compared to control Tau [605]-transfected CG-4 cells (c). Higher-

magnification micrographs show that experimental cells formed patchy myelin (b), whereas control cells formed a considerable amount of myelin (d). White arrows indicate site of cell injection. Scale bar: a, c = 500 lm; b, d = 200 lm.

(a)

(b)

(c)

(d)

Fig. 8 Cellular process outgrowth failure in Tau [688]-transfected CG4 cells transplanted in the brain of md rats. PLP immunostaining of experimental and control CG-4 cells transplanted in the brain of md rats. High-magnification micrographs of PLP immunostaining show experimental Tau [688]-transfected cells have PLP largely in the cell

body and form limited areas of myelin (a, b). In contrast, control Tau [605]-transfected cells displayed PLP in the cell body and cellular processes, and formed larger myelinated areas than experimental cells (c, d). Arrows show lack of cellular processes (a, b). Scale bar: a, c ¼ 75 lm; b, d ¼ 50 lm.

lacking the C-terminal tail had normal myelin formation in the corpus callosum, yet the myelin had a higher g-ratio in the spinal cord. Furthermore, after cuprizone-induced demyelination, myelin repair in the corpus callosum was normal; however, the percentage of non-myelinated axons increased.

Thus, the myelin defects observed with Tau [688] are unique to this in vivo system. The spectrum of changes in myelin formation cannot be assessed later on during development because the md rats die at approximately 21 days of age. Although Fyn knockout



mice do not have any alteration in the time of onset and/or rate of myelination (Sperber et al. 2001), in this study we cannot completely rule out a block of myelin formation in the brain, although it would not be consistent with the pattern of myelination, which is developmentally regulated. The data on the CG-4 cells transplanted into the brain are consistent with the work by Sperber et al. (2001) in the Fyn-minus mice. Both studies showed that at this CNS site, cellular process outgrowth fails. It was previously reported that in the brain and SC, OLGs undergo a differential regulation with respect to factors and signaling pathways, which regulate the development and function of these cells (Sperber et al. 2001; Chun et al. 2003). While CG-4 cells (derived from the brain) (Louis et al. 1992) are able to extend cellular processes in the SC, myelination rostral and caudal to the injection site decreased. Changes in cell migration were not previously reported in Fyn-minus mice or in the myelin-associated glycoprotein/Fyn double knockout mice (Biffiger et al. 2000; Sperber et al. 2001) [Correction added on 8th September 2008, after first online publication : sentence amended.]. We also report normal myelin structure, which is consistent with the work by Sperber et al. (2001) in the Fyn-minus SC. In contrast, Seiwa et al. (2000) showed that Fyn tyrosine kinase participates in the compact myelin sheath formation in the central nervous system of Fyn-deficient mice. Since specific differences among tracts within the spinal cord were reported in the myelin-associated glycoprotein /Fyn double knock out (Biffiger et al. 2000), future studies need to evaluate potential region-specific and developmental differences in myelination caused by Tau [688]. The finding that cells displayed diminished myelination along the spinal cord, rostral and caudal to the injection site supports the possibility of diminished migration along the spinal cord. A possible diminished migratory capacity of experimental cells is supported by the finding that these cells displayed in culture an apparent increase in cell-cell contact, since they were often found in groups on the second day of OLG differentiation. In contrast, control cells do not display a similar pattern of association (data not shown). It was previously shown that alphaVBeta1 integrin regulates migration of oligodendrocyte precursor cells (Milner et al. 1996), and previous studies have shown that increased Src activity correlated with the spread of glial malignancy (Alper and Bowden 2005). Future studies will need to determine the signaling pathways by which the Tau-Fyn interaction regulates migration. The absence of the microtubule-binding domain has been reported in some Tau isoforms present in the CNS, and these truncated forms localize to the soma, processes, and tips of neuronal cells (Luo et al. 2004), and consistent with this report we have detected FLAG-immunopositive staining in the soma, processes and tips of Tau-transfected CG-4 cells (data not shown). Truncated Tau has also been shown to associate with the membrane of cultured neurons and OLGs

(Brandt et al. 1995; Klein et al. 2002). The localization of Tau at the membrane is quite intriguing in light of the present observation of a regulation of myelination. Future studies will determine whether these truncated forms of Tau reach the myelin-forming sites in vivo. Tau in oligodendrocytes and CNS degeneration Previous studies have shown myelin alterations at the time of neurodegeneration in aged animals that over express Tau in OLGs (Higuchi et al. 2005). Similarly, Tau deposits present in glial cells have been associated with CNS degeneration (Lin et al. 2003). Tau-positive inclusions in OLGs and white matter pathology are prominent in progressive supranuclear palsy, Pick’s disease, and corticobasal degeneration (Lee et al. 2001). This study shows, for the first time, a direct effect of Tau present in OLGs on myelin formation. Myelin alterations could then be the primary mechanism by which mutated Tau in OLGs leads to axonal degeneration, which also has implications for multiple sclerosis where axonal degeneration is the key factor in clinical decline (Bjartmar et al. 2003). Future studies then need to determine whether during multiple sclerosis alterations of Tau in OLGs are responsible for myelin changes and consequent axonal degeneration. Together, these data establish that Tau in OLGs has a role in developmental myelination, and the lack of a functional Tau-Fyn interaction might be at the heart of some human CNS degenerative diseases.

Acknowledgements This work was supported in part by the National Multiple Sclerosis Society grants RG2958A1 and PP0864 (P.L.). The authors are grateful to Professor Ian Duncan for the generous support, and thank Dr. Teepu Siddique for insightful discussions, critiques, and generous facilities in the Neurogenetics Lab. P.L. thanks EMD Serono/Pfizer and Teva Neuroscience.

References Alper O. and Bowden E. T. (2005) Novel insights into c-Src. Curr. Pharm. Des. 11, 1119–1130. Benninger Y., Colognato H., Thurnherr T., Franklin R. J., Leone D. P., Atanasoski S., Nave K. A., Ffrench-Constant C., Suter U. and Relvas J. B. (2006) Beta1-integrin signaling mediates premyelinating oligodendrocyte survival but is not required for CNS myelination and remyelination. J. Neurosci. 26, 7665–7673. Biffiger K., Bartsch S., Montag D., Aguzzi A., Schachner M. and Bartsch U. (2000) Severe hypomyelination of the murine CNS in the absence of myelin-associated glycoprotein and Fyn tyrosine kinase. J. Neurosci. 20, 7430–7437. Bjartmar C., Wujek J. R. and Trapp B. D. (2003) Axonal loss in the pathology of MS: consequences for understanding the progressive phase of the disease. J. Neurol. Sci. 206, 165–171. Boison D. and Stoffel W. (1989) Myelin-deficient rat: a point mutation in exon III (A——C, Thr75——Pro) of the myelin proteolipid protein causes dysmyelination and oligodendrocyte death. EMBO J. 8, 3295–3302.



Brandt R., Leger J. and Lee G. (1995) Interaction of tau with the neural plasma membrane mediated by tau amino-terminal projection domain. J. Cell Biol. 13, 1327–1340. Chew L. J., Yuan X., Scherer S. E., Qie L., Huang F., Hayes W. P. and Gallo V. (2001) Characterization of the rat GRIK5 kainate receptor subunit gene promoter and its intragenic regions involved in neural cell specificity. J. Biol. Chem. 276, 42162–42171. Epub 2001 Aug 31. Chun S. J., Rasband M. N., Sidman R. L., Habib A. A. and Vartanian T. (2003) Integrin-linked kinase is required for laminin-2-induced oligodendrocyte cell spreading and CNS myelination. J. Cell Biol. 163, 397–408. Collet J., Fehrat L., Pollard H., Ribas de Pouplana L., Charton G., Bernard A., Moreau J., Ben-Ari Y. and Khrestchatisky M. (1997) Developmentally regulated alternative splicing of mRNAs encoding N-terminal tau variants in the rat hippocampus: structural and functional implications. Eur. J. Neurosci. 9, 2723–2733. Colognato H., Ramachandrappa S., Olsen I. M. and Ffrench-Constant C. (2004) Integrins direct Src family kinases to regulate distinct phases of oligodendrocyte development. J. Cell Biol. 167, 365–375. Dentinger M. P., Barron K. D. and Csiza C. K. (1982) Ultrastructure of the central nervous system in a myelin deficient rat. J. Neurocytol. 11, 671–691. Dixit R., Ross J. L., Goldman Y. E. and Holzbaur E. L. (2008) Differential regulation of dynein and kinesin motor proteins by Tau. Science 319, 1086–1089. Ebneth A., Godemann R., Stamer K., Illenberger S., Trinczek B. and Mandelkow E. (1998) Overexpression of tau protein inhibits kinesin-dependent trafficking of vesicles, mitochondria, and endoplasmic reticulum: implications for Alzheimer’s disease. J. Cell Biol. 143, 777–794. Espinosa de los Monteros A., Zhao P., Huang C., Pan T., Chang R., Nazarian R., Espejo D. and de Vellis J. (1997) Transplantation of CG4 oligodendrocyte progenitor cells in the myelin-deficient rat brain results in myelination of axons and enhanced oligodendroglial markers. J. Neurosci. Res. 50, 872–887. Fischer I., Konola J. and Cochary E. (1990) Microtubule associated protein (MAP1B) is present in cultured oligodendrocytes and co-localizes with tubulin. J. Neurosci. Res. 27, 112–124. Franklin R. J., Bayley S. A. and Blakemore W. F. (1996) Transplanted CG4 cells (an oligodendrocyte progenitor cell line) survive, migrate, and contribute to repair of areas of demyelination in X-irradiated and damaged spinal cord but not in normal spinal cord. Exp. Neurol. 137, 263–276. Goedert M., Crowther R. A. and Garner C. C. (1991) Molecular characterization of microtubule-associated proteins tau and MAP2. Trends Neurosci. 14, 193–199. Higuchi M., Zhang B., Forman M. S., Yoshiyama Y., Trojanowski J. Q. and Lee V. M. (2005) Axonal degeneration induced by targeted expression of mutant human tau in oligodendrocytes of transgenic mice that model glial tauopathies. J. Neurosci. 25, 9434–9443. Klein C., Kramer E. M., Cardine A. M., Schraven B., Brandt R. and Trotter J. (2002) Process outgrowth of oligodendrocytes is promoted by interaction of Fyn kinase with the cytoskeletal protein tau. J. Neurosci. 22, 698–707. Klinghoffer R. A., Sachsenmaier C., Cooper J. A. and Soriano P. (1999) Src family kinases are required for integrin but not PDGFR signal transduction. EMBO J. 18, 2459–2471. Kosik K. S., Orecchio L. D., Bakalis S. and Neve R. L. (1989) Developmentally regulated expression of specific tau sequences. Neuron 2, 1389–1397. Ladrech S. and Lenoir M. (2002) Changes in MAP2 and tyrosinated alpha-tubulin expression in cochlear inner hair cells after amikacin treatment in the rat. J. Comp. Neurol. 451, 70–78.

Lee G., Newman S. T., Gard D. L., Band H. and Panchamoorthy G. (1998) Tau interacts with Src-family non-receptor tyrosine kinases. J. Cell Sci. 111, 3167–3177. Lee V. M., Goedert M. and Trojanowski J. Q. (2001) Neurodegenerative tauopathies. Annu. Rev. Neurosci. 24, 1121–1159. Lee G., Thangavel R., Sharma V. M., Litersky J. M., Bhaskar K., Fang S. M., Do L. H., Andreadis A., Van Hoesen G. and Ksiezak-Reding H. (2004) Phosphorylation of tau by Fyn: implications for Alzheimer’s disease. J. Neurosci. 24, 2304–2312. Lee J., Gravel M., Zhang R., Thibault P. and Braun P. E. (2005) Process outgrowth in oligodendrocytes is mediated by CNP, a novel microtubule assembly myelin protein. J. Cell Biol. 170, 661–673. Lee K. K., de Repentigny Y., Saulnier R., Rippstein P., Macklin W. B. and Kothary R. (2006) Dominant-negative beta1 integrin mice have region-specific myelin defects accompanied by alterations in MAPK activity. Glia 53, 836–844. Liang X., Draghi N. A. and Resh M. D. (2004) Signaling from integrins to Fyn to Rho family GTPases regulates morphologic differentiation of oligodendrocytes. J. Neurosci. 24, 7140–7149. Lin W. L., Lewis J., Yen S. H., Hutton M. and Dickson D. W. (2003) Filamentous tau in oligodendrocytes and astrocytes of transgenic mice expressing the human tau isoform with the P301L mutation. Am. J. Pathol. 162, 213–218. LoPresti P., Szuchet S., Papasozomenos S. C., Zinkowski R. P. and Binder L. I. (1995) Functional implications for the microtubuleassociated protein tau: localization in oligodendrocytes. Proc. Natl Acad. Sci. USA 92, 10369–10373. LoPresti P., Numa N. A. and DeVries G. H. (2001) Neu differentiation factor alpha and beta induced upregulation of tau protein and mRNA in cultured neonatal oligodendrocytes. Glia 35, 147–155. LoPresti P. (2002) Regulation and differential expression of tau mRNA isoforms as oligodendrocytes mature in vivo: implications for myelination. Glia 37, 250–257. Louis J. C., Magal E., Muir D., Manthorpe M. and Varon S. (1992) CG4, a new bipotential glial cell line from rat brain, is capable of differentiating in vitro into either mature oligodendrocytes or type2 astrocytes. J. Neurosci. Res. 31, 193–204. Luo M. H., Tse S. W., Memmott J. and Andreadis A. (2004) Novel isoforms of tau that lack the microtubule-binding domain. J. Neurochem. 90, 340–351. Maccioni R. B. and Cambiazo V. (1995) Role of microtubule-associated proteins in the control of microtubule assembly. Physiol. Rev. 75, 835–864. Mandelkow E., Song Y. H., Schweers O., Marx A. and Mandelkow E. M. (1995) On the structure of microtubules, tau, and paired helical filaments. Neurobiol. Aging 16, 347–354. Milner R., Edwards G., Streuli C. and Ffrench-Constant C. (1996) A role in migration for the alpha V beta 1 integrin expressed on oligodendrocyte precursors. J. Neurosci. 16, 7240–7252. Muller R., Heinrich M., Heck S., Blohm D. and Richter-Landsberg C. (1997) Expression of microtubule-associated proteins MAP2 and tau in cultured rat brain oligodendrocytes. Cell Tissue Res. 288, 239–249. O’Connor L. T., Goetz B. D., Couve E., Song J. and Duncan I. D. (2000) Intracellular distribution of myelin protein gene products is altered in oligodendrocytes of the taiep rat. Mol. Cell. Neurosci. 16, 396–407. Osterhout D. J., Wolven A., Wolf R. M., Resh M. D. and Chao M. V. (1999) Morphological differentiation of oligodendrocytes requires activation of Fyn tyrosine kinase. J. Cell Biol. 145, 1209–1218. Seiwa C., Sugiyama I., Yagi T., Iguchi T. and Asou H. (2000) Fyn tyrosine kinase participates in the compact myelin sheath formation in the central nervous system. Neurosci. Res. 37, 21–31.



Shafit-Zagardo B., Kress Y., Zhao M. L. and Lee S. C. (1999) A novel microtubule-associated protein-2 expressed in oligodendrocytes in multiple sclerosis lesions. J. Neurochem. 73, 2531–2537. Sperber B. R., Boyle-Walsh E. A., Engleka M. J., Gadue P., Peterson A. C., Stein P. L., Scherer S. S. and McMorris F. A. (2001) A unique role for Fyn in CNS myelination. J. Neurosci. 21, 2039–2047. Tontsch U., Archer D. R., Dubois-Dalcq M. and Duncan I. D. (1994) Transplantation of an oligodendrocyte cell line leading to extensive myelination. Proc. Natl Acad. Sci. USA 91, 11616–11620. Umemori H., Sato S., Yagi T., Aizawa S. and Yamamoto T. (1994) Initial events of myelination involve Fyn tyrosine kinase signaling. Nature 367, 572–576.

Vouyiouklis D. A. and Brophy P. J. (1993) Microtubule-associated protein MAP1B expression precedes the morphological differentiation of oligodendrocytes. J. Neurosci. Res. 35, 257–267. Warn R. M., Harrison A., Planques V., Robert-Nicoud N. and Wehland J. (1990) Distribution of microtubules containing post-translationally modified alpha-tubulin during Drosophila embryogenesis. Cell Motil. Cytoskeleton 17, 34–45. Wolf R. M., Wilkes J. J., Chao M. V. and Resh M. D. (2001) Tyrosine phosphorylation of p190 RhoGAP by Fyn regulates oligodendrocyte differentiation. J. Neurobiol. 49, 62–78.
