Blackwell Science, LtdOxford, UKCGACongenital Anomalies0914-3505The Japanese Teratology Society, 2004XX 2004444181188Review ArticleProteoglycans and CNS injuryF. Matsui and A. Oohira

Congenital Anomalies 2004; 44, 181–188 181

REVIEW ARTICLE

Proteoglycans and injury of the central nervous system

Fumiko Matsui and Atsuhiko Oohira Department of Perinatology, Institute for Developmental Research, Aichi Human Service Center, Kasugai, Aichi, Japan

ABSTRACT is a family of glycopro- INTRODUCTION teins which carry covalently-linked glycosaminoglycan It is well known that axonal regeneration is unsuccessful in chains, such as and . the injured adult mammalian central nervous system (CNS). are believed to play important roles in Histochemical studies have shown that a glial scar is formed morphogenesis and maintenance of various tissues at the injury site and prevents axonal regeneration. Many including the central nervous system (CNS) through inhibitory molecules, such as myelin-associated glycopro- interactions with molecules and growth fac- tein, Nogo, Semaphorin, and chondroitin sulfate proteogly- tors. In the CNS, a significant amount of evidence has cans (CSPG), have been found in the glial scar (for reviews, been accumulated to show that proteoglycans function as see Fawcett & Asher 1999; Morgenstern et al. 2002; Prop- modulators in various cellular events not only in the erzi et al. 2003; Rhodes & Fawcett 2004). Of these mole- development, but also in the pathogenesis of neuronal cules, CSPG recently have come into the limelight in terms diseases and lesions. When the CNS is injured, several of injury repair of the CNS. chondroitin sulfate proteoglycans (CSPG) are up-regu- Proteoglycan consists of a core and glycosami- lated in glial scars formed around the lesion site. The glial noglycan chains, such as chondroitin sulfate and heparan scar also contains some molecules inhibitory to axonal sulfate (Bandtlow & Zimmermann 2000; Oohira et al. growth, such as myelin-associated , Nogo, 2000). The core protein also attaches N-linked and O-linked and Semaphorin. In vitro studies revealed that CSPG oligosaccharides. Some proteoglycans exist in the extracel- largely exert a repulsive effect on axonal regeneration, lular matrix as secretory molecules, and others exist at the and a signal from CSPG modulates the actin cytoskeleton cell surface as transmembrane or glycosylphosphatidylinos- of outgrowing neurites through the Rho/ROCK pathway. itol (GPI) -anchored molecules (Fig. 1). Their expressions These findings suggest that CSPG are responsible for are developmentally regulated (Bandtlow & Zimmermann unsuccessful axonal regeneration in glial scars. Various 2000), and the structure of their carbohydrate moieties, attempts to overcome the inhibitory effect of CSPG have namely both glycosaminoglycans and oligosaccharides, also been pursued in vivo. Digestion of chondroitin sulfate changes in a development-related manner (Shuo et al. 2004). chains by chondroitinase ABC, suppression of CSPG In many cases, both core and carbohydrate moieties core protein synthesis by , suppression of gly- are necessary for proteoglycans to interact with extracellular cosaminoglycan chain synthesis by a DNA enzyme, and matrix molecules, cell adhesion molecules, and growth fac- inhibition of the Rho/ROCK pathway with specific inhib- tors (Oohira et al. 2000). Since those interactions are sup- itors were all successful for increasing axonal regenera- posed to be important for morphogenesis of various tissues, tion. For a clinical application, the most effective proteoglycan seems to modulate its function through the combination of these treatments needs to be examined in structural changes of protein and carbohydrate moieties in the future. the process of development. In the CNS, proteoglycans are considered to play pivotal Key Words: central nervous system, neurocan, phosphacan, roles in cell-cell and cell-substratum interactions in the proteoglycans, regeneration development, maintenance, and aging of normal tissues (Bandtlow & Zimmermann 2000; Oohira et al. 2000). Addi- tionally, since expressions of some CSPG drastically change Correspondence: Fumiko Matsui, PhD, Department of Perinatology, Insti- in response to CNS injuries, they are believed to be involved tute for Developmental Research, Aichi Human Service Center, Kasugai, Aichi 480-0392, Japan. Email: [email protected] in the pathogenesis and/or repair processes of neuronal dam- Received August 12, 2004; revised and accepted August 23, 2004. ages (Fawcett & Asher 1999; Morgenstern et al. 2002;

182 F. Matsui and A. Oohira

NEUROCAN proteoglycans are predominantly present in the CNS, and (Secretory type) others are in non-neuronal tissues as well as neuronal tissues. chondroitin sulfate A few, such as a small chondroitin/dermatan sulfate pro- core protein teoglycan decorin, are distributed ubiquitously in animal

NEUROGLYCAN C bodies (Iozzo 1999). (Transmembrane type) The CNS consists of two different cell types; neuronal (GPI-anchored type) cells and glial cells. Both cell types can be divided further Extracellular into many subclasses with a particular morphology and func- Intracellular chondroitin tion. Some proteoglycans are exclusively expressed by one sulfate heparan GPI cell type, but others are expressed by both. A typical glial sulfate proteoglycan is NG2, a large transmembrane CSPG, which Cell is expressed by oligodendrocyte progenitor cells in the CNS Nucleus (Levine & Nishiyama 1996). Neurocan is reported to be synthesized and secreted mainly by neuronal cells in the Fig. 1 Schematic representation of three types of proteoglycan. normal CNS, although it becomes actively synthesized by Neurocan (Rauch et al. 1992), glypican (Stipp et al. astroglial cells in response to injuries of the CNS (for details, 1994), and neuroglycan C (Watanabe et al. 1995) are depicted representing a secretory proteoglycan, a glyco- see below). In contrast, phosphacan and receptor-type pro- sylphosphatidylinositol (GPI) -anchored proteoglycan, and tein tyrosine phosphatase z/b (RPTPz/b) are produced by a membrane-spanning proteoglycan, respectively. both glial and neuronal cells (Shintani et al. 1998). Each individual proteoglycan shows a particular spa- tiotemporal expression pattern in the CNS, and in many Properzi et al. 2003; Rhodes & Fawcett 2004). Very inter- cases interacts with other extracellular and/or intracellular estingly, depletion of chondroitin sulfate chains or core pro- molecules (Oohira et al. 2000). This suggests that it would teins of CSPG from the lesion site promotes axonal function as a ligand and/or receptor in particular phases in regeneration in the mature CNS. For example, chondroitin CNS development. In fact, it has recently become clearer sulfate degrading enzyme improved axonal regeneration and that neural proteoglycans are involved in various develop- brought about a functional recovery of spinal cord axons mental events of the CNS including cell proliferation, migra- when the enzyme was applied to a rat with a spinal cord tion, cellular differentiation, neurite elongation, pathfinding injury (Bradbury et al. 2002). of axons, and synaptogenesis through molecular interactions In this paper, we first reviewed the research related to with growth factors, extracellular matrix molecules, cell nervous tissue proteoglycans, and then dealt with the recent adhesion molecules, and cytoskeletal components (Bandtlow reports on the expression of proteoglycans, especially & Zimmermann 2000; Oohira et al. 2000; Yoneda & CSPG, in the injured CNS. We also described the inhibitory Couchman 2003). In addition, neural proteoglycans have mechanism of CSPG for axonal regeneration and various been shown to regulate the neuronal plasticity by forming attempts to promote neuronal regeneration by depleting perineuronal nets around synapses (Matsui et al. 1998; Piz- CSPG of the injured sites. zorusso et al. 2002). As a result, more clues are available for understanding how to recover neuronal plasticity even in the Overview of proteoglycans in the central nervous system mature CNS. Early studies on proteoglycans were mainly pursued using connective tissues as sources, especially the cartilage which Expression changes of chondroitin sulfate proteoglycans has a massive amount of extracellular matrix. Although the after central nervous system injury CNS does not have as much the amount of extracellular Many proteoglycan species have been shown to change their matrix compared to the cartilage, it contains multiple species expression levels around lesion sites in response to CNS of proteoglycan in the extracellular matrix and at the cell injuries. Typical examples of those proteoglycans are sum- surfaces of neural cells (Oohira et al. 1994a; Bandtlow & marized in Table 1 and are reviewed below. Zimmermann 2000). They can be classified into two groups: Moon et al. (2002) reported that, following axotomy of proteoglycans bearing chondroitin sulfate chains (CSPG) the nigrostriatal tract in adult rats, HSPG were predomi- and proteoglycans bearing heparan sulfate chins (HSPG). In nantly found within the lesion core, whereas CSPG and addition, there are a few proteoglycans bearing keratan sul- KSPG were predominantly found in the lesion surrounding fate chains (KSPG). Many CSPG, such as neurocan, it. Axons sprouted within the lesion core, but rarely grew and , exist in the extracellular matrix, and most into the lesion surrounding it. Therefore, this study suggests HSPG, such as syndecan family molecules and glypican that CSPG and KSPG act as barriers to regenerating axons. family molecules, exist at the cell surface. Some of these Expression changes of HSPG and KSPG in the CNS injury

Proteoglycans and CNS injury 183

Table 1 Major proteoglycans whose expressions significantly change after injury of the central nervous system

Core protein Increase or Name Accession no.† size‡ (kDa) Type of GAG decrease§ References BC027971 145 CS Ø≠ Tang et al. (2003) Neurocan AF026547 245 CS ≠ McKeon et al. (1999) Versican V2 U26555 400 CS ≠ Asher et al. (2002) Versican V2 U26555 400 CS Ø Tang et al. (2003) NG2 NM_031022 300 CS ≠ Jones et al. (2002) Phosphacan/RPTPz//b U09357 90–400 CS, KS ≠Ø McKeon et al. (1999) Glypican 1 BC051279 64 HS ≠ Hagino et al. (2003) Syndecan 1 BC008765 80 HS ≠ Iseki et al. (2002) Syndecan 3 BC013974 120 HS ≠ Iseki et al. (2002) †Human or rat sequences are shown; ‡core protein size estimated by sodium dodecylsulfate–polyacrylamide gel electro- phoresis; §≠Ø, increase or decrease after central nervous system injury. CS, chondroitin sulfate; GAG, glycosaminoglycan; HS, heparan sulfate; KS, ; RPTPz//b, receptor-type protein tyrosine phosphatase z/b.

and the biological meanings are discussed in detail in other Up-regulation of neurocan after central nervous sections below. system injury Up-regulation of CSPG, such as neurocan (see the next Neurocan is one of the major CSPG in the extracellular section) and NG2 (Jones et al. 2002; Morgenstern et al. matrix of the CNS. Since neurocan interacts with certain 2002; Tang et al. 2003), has been reported both in cerebral neural cell adhesion molecules, extracellular matrix mole- and spinal cord injuries. However, not all the molecular cules, and growth factors, it is thought to be involved in cell species of CSPG are up-regulated in the glial scar. The migration and cell proliferation (reviewed in Oohira et al. expressions of other CSPG are somewhat complicated. For 2000). The expression of neurocan is developmentally reg- example, some immunohistochemical and in situ hybridiza- ulated, reaching a maximum level at around birth (Oohira tion studies have reported up-regulation of phosphacan/ et al. 1994b). The proteolytic cleavage of neurocan (Fig. 2) RPTPz/b after brain injury (Barker et al. 1996; Snyder et al. is also developmentally regulated. In the juvenile brain, both 1996; Deller et al. 1997; Naffah-Mazzacoratti et al. 1999). full-length neurocan and its fragments are detectable. In the However, western blot analyzes revealed the decreased normal adult brain, however, full-length neurocan is not expression of phosphacan/RPTPz/b, both in the surgically detectable, although both neurocan fragments are still detect- injured cerebral cortex (Wu et al. 2000; Matsui et al. 2002) able (Matsui et al. 1994; Milev et al. 1998). and in the kainate-lesioned brains (Matsui et al. 2002). Recent studies have shown that neurocan expression is up- McKeon et al. (1999) reported that the phosphacan protein regulated in the surgically injured adult cerebral cortex. It is level in gliotic tissues decreased to 67% of the level of interesting that, in addition to neurocan fragments, full- uninjured cerebral cortex, although phosphacan messenger length neurocan was detected in the injured adult brains ribonucleic acid (mRNA) was slightly elevated. Morgenstern (McKeon et al. 1999; Asher et al. 2000; Matsui et al. 2002). et al. (2002) described that phosphacan protein initially The amount of full-length neurocan was increased from 2 to decreased and then was up-regulated at later stages 7 days after trauma and then began to decrease to the normal (7–14 days) following cortical injury. Similar biphasic level. In glial scars, reactive astrocytes and oligodendrocytes responses of phosphacan and brevican were reported in rat seem to synthesize neurocan in response to growth factors spinal cord scar tissue (Tang et al. 2003). In contrast, versi- and cytokines (Asher et al. 2000). Up-regulation of neurocan can V2 protein level was reduced in rat spinal cord scar tissue was also reported in entorhinal cortex lesion (Haas et al. (Tang et al. 2003) and was elevated in cortical injury (Asher 1999), in retinal ischemia (Inatani et al. 2000), in kainate- et al. 2002). injured brain (Fig. 3: Matsui et al. 2002; Okamoto et al. The expression pattern of a CSPG species after injury may 2003), and in spinal cord injury (Qi et al. 2003; Tang et al. vary depending on the injured area of the CNS, the age of 2003). These studies indicate that up-regulation of neurocan the injured animal, and how the CNS was injured. is common among various types of CNS injuries.

184 F. Matsui and A. Oohira

Signal Ig-like HA-binding Chondroitin EGF- - Complement peptide sulfate-attaching like like regulatory- like

Intact neurocan (245 kDa)

Neurocan-N Neurocan-C (130 kDa) (150 kDa)

Fig. 2 Domain structure of the neurocan core protein. Neurocan-N and neurocan-C are generated from intact neurocan by proteolytic cleavage at the central portion of the core protein. Ig, immunoglobulin.

Fig. 3 Localization of neurocan in the rat hippocampus 4 days after kainate-treatment. Sections were excised from the brain of rats with (B,D) and without (A,C) seizures, and were Nissl-stained (A,B) or immunostained with an anti-neurocan antibody MAb 1G2 (C,D). Immunoreactivity for neurocan is intensified around the CA1 region of the hippocampus (D) where neuronal cell death was obvious (B). Bar, 1 mm.

Effects of chondroitin sulfate proteoglycans on cerebellar or cortical axons avoided neurocan or phosphacan axonal regeneration stripes and elongated preferentially on L1 (Asher et al. Many in vitro assays have shown that CSPG largely exert a 2000). Neurite outgrowth from fetal rat retinal ganglion cells repulsive effect on axonal regeneration. For example, a stripe and from adult rat dorsal root ganglion (DRG) neurons was assay using early postnatal rat neurons has shown that either also impaired on phosphacan and neurocan substrata (Inatani

Proteoglycans and CNS injury 185 et al. 2001; Sango et al. 2003). Since chondroitinase ABC way may become a useful tool to promote the axonal regen- treatment of these CSPG partially, not fully, suppressed their eration of the CNS. inhibitory effects on neurite elongation, both chondroitin sulfate chains and core proteins are thought to be respon- Depletion of chondroitin sulfate proteoglycans promotes sible. Thus, CSPG are generally thought to be inhibitory to axonal regeneration axonal regeneration. However, two research groups have To overcome the inhibitory effect of CSPG on the axonal found that ascending sensory axons regenerate into the area regeneration, various technologies to deplete CSPG around where CSPG are expressed following a spinal cord injury injured sites are advancing in vitro and in vivo. Digestion of (Pasterkamp et al. 2001; Inman & Steward 2003). Therefore, chondroitin sulfate chains by chondroitinase ABC promoted CSPG seem to be permissive to certain types of neurons axonal regeneration in a rat nigrostriatal system (Moon et al. under some particular conditions. 2001). In a spinal cord injury, injection of chondroitinase On repulsive substrate constituted of nervous system- ABC not only promoted axonal regeneration but also derived CSPG, actin arrangement in the growth cone of DRG improved functions, such as postsynaptic activity and loco- neurons was normal (Snow et al. 2001). Growth cones motion (Bradbury et al. 2002). Suppression of CSPG core altered their trajectory to avoid CSPG without collapse. In protein syntheses or chondroitin sulfate syntheses at lesion contrast, aggrecan, a large aggregating CSPG, altered growth sites also promoted axon outgrowth in these locations. For cone morphology (Borisoff et al. 2003). Although aggrecan example, Davies et al. (2004) administered decorin, a chon- is not known to be up-regulated following CNS injury, in droitin/dermatan sulfate proteoglycan, to injured sites in the vitro experiments have shown that aggrecan substrate is adult rat spinal cord. Decorin is known to inhibit the activity repulsive to DRG neurons. Unlike the nervous tissue CSPG of transforming growth factor b (Yamaguchi et al. 1990) and mixture, aggrecan substrate promoted the formation of more epidermal growth factor receptor tyrosine kinase (Santra actin bundles in the central domains of growth cones com- et al. 2002), and these growth factors promote syntheses of pared to permissive laminin substrates. There is a possibility other proteoglycan species (Asher et al. 2000). As expected, that the difference of CSPG species used in these experi- decorin suppressed expressions of several CSPG, such as ments, namely, the nervous tissue CSPG mixture versus neurocan, brevican, phosphacan and NG2, and promoted aggrecan, caused the different morphology of growth cones. axon growth across adult rat spinal cord injuries. Another example to reduce expression of chondroitin sul- Signal transduction of proteoglycans in axonal fate chains is a DNA enzyme designed to target and degrade regeneration mRNA of an enzyme, xylosyltansfarase-1, which initiates Extracellularly secreted proteoglycans require cell surface glycosaminoglycan synthesis on core proteins. The DNA receptors to modulate neurite outgrowth. However, receptors enzyme, when infused around injured sites, reduced expres- specific to proteoglycans have not been identified. Since sion of chondroitin sulfate chains in the injured spinal cord, some CSPG are known to bind to cell adhesion molecules, and allowed DRG neurons to regenerate as expected such as NCAM and NgCAM (reviewed in Oohira et al. (Grimpe & Silver 2004). Lastly, suppression of intracellular 2000; Rauch et al. 2001), these adhesion molecules could be Rho/ROCK signal promoted axonal regeneration. As entrances for extracellular CSPG signals. The intracellular described above, the Rho/ROCK pathway mediates signals signals provoked in neuronal cells by CSPG have also been from CSPG and other inhibitory molecules, and PKC is unclear. However, recent studies demonstrated that the Rho/ involved in Rho-activation. Therefore, a PKC inhibitor, ROCK pathway was involved in the CSPG signal transduc- Gö6976, was infused to injured adult rat spinal cords to tion, for example, aggrecan substrate stimulated Rho, a fam- suppress the Rho/ROCK pathway. This treatment promoted ily of small GTPase, in chick DRG neurons. In addition, regeneration of dorsal column axons, but not corticospinal suppression of ROCK, a Rho-kinase, by Y-27632, a specific tract axons (Sivasankaran et al. 2004). All these trials sug- inhibitor of ROCK, reduced the inhibitory effect of aggrecan gest that both depletion of CSPG itself from the lesion site (Borisoff et al. 2003). A specific inhibitor of the Rho, C3 and suppression of CSPG-derived intracellular signals are transferase, also blocked CSPG to inhibit axon growth of promising potential treatments for overcoming CNS injuries. retina ganglion cells (Monnier et al. 2003). Furthermore, blocking activity of PKC attenuated the ability of CSPG to Expressions of heparan sulfate proteoglycans in the activate Rho and to inhibit neurite outgrowth (Sivasankaran injured CNS et al. 2004). The Rho/ROCK pathway is known to modulates In addition to CSPG, HSPG have been reported to be actin cytoskeleton in outgrowing neurite (for a review, see expressed in the injured brain. Distribution patterns of Amano et al. 2000) and several molecules inhibitory for mRNA for syndecan-1 and -3, members of a transmembrane neurite outgrowth also activate this pathway (Luo 2000). HSPG family, overlapped with those of fibroblast growth Therefore, compounds which can modulate this signal path- factor (FGF) receptor 1 and pleiotrophin, a heparin-binding 186 F. Matsui and A. Oohira growth factor, respectively, in the injured brain (Iseki et al. help overcome CNS injuries. To establish a more effective 2002). In vitro analysis using goldfish showed that hepariti- and reliable treatment applicable to human CNS injuries, a nase treatment of optic tectal membranes abolished their better understanding of physiological roles of the CSPG in promoting activity for neurite outgrowth from retinal the injury repair is needed. Our knowledge is very limited at explants (Su & Elam 2003). These results suggest that HSPG present about the cell surface target molecules and the intra- promote axon regeneration, presumably serving as corecep- cellular signaling pathway of the extracellular CSPG up- tors for certain heparin-binding growth factors, such as FGF regulated around lesion sites of the CNS. and pleiotrophin in the injured brain. However, glypican-1, a GPI-anchored HSPG, may participate in the inhibitory process (Hagino et al. 2003). Since glypican-1 serves as a ACKNOWLEDGMENTS receptor for a repellent protein, Slit, and glypican-1 mRNA We thank all of our colleagues and collaborators who have is coexpressed with Slit mRNA in the reactive astrocytes of contributed so much dedicated effort and thought to our injured brain, glypican-1 may cooperate with Slit protein to work. Research conducted at the authors’ laboratory was create a barrier against regenerating axons. supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Culture and Sports Expressions of keratan sulfate proteoglycans in the of Japan and from the Japan Society for the Promotion of injured central nervous system Science, and a grant from the Mizutani Foundation for Immunohistochemical studies have shown that the expres- Glycoscience. sion of keratan sulfate is up-regulated following nigrostriatal axotomy (Moon et al. 2002) and spinal cord injury (Krautstrunk et al. 2002) in adult rats. Up-regulation of kera- REFERENCES tan sulfate was also found in the cortical injury of neonatal rats (Geisert et al. 1996). Since phosphacan, one of the major Amano M, Fukata Y, Kaibuchi K (2000) Regulation and CSPG in the brain, carries keratan sulfate chains, it is sup- functions of Rho-associated kinase. Exp Cell Res 261: 44– posed to be a main proteoglycan which carries keratan sul- 51. fate chains in the CNS. Recently, Butler et al. (2004) found Asher RA, Morgenstern DA, Fidler PS et al. (2000) Neuro- that keratanase treatment of rat hippocampal slices in culture can is upregulated in injured brain and in cytokine-treated resulted in a disorganized regeneration of mossy fibers astrocytes. J Neurosci 20: 2427–2438. extending from granule cells. These studies suggest that Asher RA, Morgenstern DA, Shearer MC, Adcock KH, KSPG play various roles in axonal regeneration. However, Pesheva P, Fawcett JW (2002) Versican is upregulated in a quantitative analysis for keratan sulfate after injury and CNS injury and is a product of oligodendrocyte lineage identification of core proteins which carry keratan sulfate cells. J Neurosci 22: 2225–2236. chains at the injury sites remain to be examined. Bandtlow CE, Zimmermann DR (2000) Proteoglycans in the developing brain: New conceptual insights for old pro- teins. Physiol Rev 80: 1267–1290. CONCLUSION Barker RA, Dunnett SB, Faissner A, Fawcett JW (1996) The Proteoglycans have been shown to be involved in various time course of loss of dopaminergic neurons and the gli- cellular events, not only in the development but also in the otic reaction surrounding grafts of embryonic mesenceph- pathogenesis of the CNS, and they are now becoming more alon to the striatum. Exp Neurol 141: 79–93. central in the field of neuroscience. In the repair process of Borisoff JF, Chan CC, Hiebert GW et al. (2003) Suppression CNS injuries, several CSPG species are transiently up- of Rho-kinase activity promotes axonal growth on inhib- regulated around lesion sites, suggesting that they prevent itory CNS substrates. Mol Cell Neurosci 22: 405–416. the expansion of damaged areas and the aberrant projection Bradbury EJ, Moon LD, Popat RJ et al. (2002) Chondroiti- of nerve fibers. However, this transient accumulation of the nase ABC promotes functional recovery after spinal cord CSPG around lesion sites can also be considered responsible injury. Nature 416: 636–640. for the unsuccessful axonal regeneration in the CNS. Some Butler CD, Schnetz SA, Yu EY et al. (2004) Keratan sulfate attempts to deplete CS-glycosaminoglycan and/or CSPG proteoglycan phosphacan regulates mossy fiber outgrowth from injured areas, and to block a signal cascade down- and regeneration. J Neurosci 24: 462–473. stream of the extracellular CSPG are effective for promotion Davies JE, Tang X, Denning JW, Archibald SJ, Davies SJ of axonal regeneration and functional recovery in some (2004) Decorin suppresses neurocan, brevican, phos- injury models of experimental animals. We have considered phacan and NG2 expression and promotes axon growth for a long time that axonal regeneration is almost impossible across adult rat spinal cord injuries. Eur J Neurosci 19: in the CNS, but now promising clues are more available to 1226–1242. Proteoglycans and CNS injury 187

Deller T, Haas CA, Naumann T, Joester A, Faissner A, ated with immature cells. Perspect Dev Neurobiol 3: 245– Frotscher M (1997) Up-regulation of astrocyte-derived 259. tenascin-C correlates with neurite outgrowth in the rat Luo L (2000) Rho GTPases in neuronal morphogenesis. Nat dentate gyrus after unilateral entorhinal cortex lesion. Rev Neurosci 1: 173–180. Neuroscience 81: 829–846. Matsui F, Kawashima S, Shuo T et al. (2002) Transient Fawcett JW, Asher RA (1999) The glial scar and central expression of juvenile-type neurocan by reactive astro- nervous system repair. Brain Res Bull 49: 377–391. cytes in adult rat brains injured by kainate-induced sei- Geisert EE, Bidanset DJ, Del Mar N, Robson JA (1996) Up- zures as well as surgical incision. Neuroscience 112: 773– regulation of a keratan sulfate proteoglycan following cor- 781. tical injury in neonatal rats. Int J Dev Neurosci 14: 257– Matsui F, Nishizuka M, Yasuda Y, Aono S, Watanabe E, 267. Oohira A (1998) Occurrence of a N-terminal proteolytic Grimpe B, Silver J (2004) A novel DNA enzyme reduces fragment of neurocan, not a C-terminal half, in a perineu- glycosaminoglycan chains in the glial scar and allows ronal net in the adult rat cerebrum. Brain Res 790: 45–51. microtransplanted dorsal root ganglia axons to regenerate Matsui F, Watanabe E, Oohira A (1994) Immunological beyond lesions in the spinal cord. J Neurosci 24: 1393– identification of two proteoglycan fragments derived from 1397. neurocan, a brain-specific chondroitin sulfate proteogly- Haas CA, Rauch U, Thon N, Merten T, Deller T (1999) can. Neurochem Int 25: 425–431. Entorhinal cortex lesion in adult rats induces the expres- McKeon RJ, Jurynec MJ, Buck CR (1999) The chondroitin sion of the neuronal chondroitin sulfate proteoglycan neu- sulfate proteoglycans neurocan and phosphacan are rocan in reactive astrocytes. J Neurosci 19: 9953–9963. expressed by reactive astrocytes in the chronic CNS glial Hagino S, Iseki K, Mori T et al. (2003) Slit and glypican-1 scar. J Neurosci 19: 10778–10788. mRNAs are coexpressed in the reactive astrocytes of the Milev P, Maurel P, Chiba A et al. (1998) Differential regu- injured adult brain. Glia 42: 130–138. lation of expression of hyaluronan-binding proteoglycans Inatani M, Honjo M, Otori Y et al. (2001) Inhibitory effects in developing brain: Aggrecan, versican, neurocan, and of neurocan and phosphacan on neurite outgrowth from brevican. Biochem Biophys Res Commun 247: 207–212. retinal ganglion cells in culture. Invest Ophthalmol Vis Sci Monnier PP, Sierra A, Schwab JM, Henke-Fahle S, Mueller 42: 1930–1938. BK (2003) The Rho/ROCK pathway mediates neurite Inatani M, Tanihara H, Oohira A, Honjo M, Kido N, Honda growth-inhibitory activity associated with the chondroitin Y (2000) Upregulated expression of neurocan, a nervous sulfate proteoglycans of the CNS glial scar. Mol Cell tissue specific proteoglycan, in transient retinal ischemia. Neurosci 22: 319–330. Invest Ophthalmol Vis Sci 41: 2748–2754. Moon LD, Asher RA, Rhodes KE, Fawcett JW (2001) Inman DM, Steward O (2003) Ascending sensory, but not Regeneration of CNS axons back to their target following other long-tract axons, regenerate into the connective tis- treatment of adult rat brain with chondroitinase ABC. Nat sue matrix that forms at the site of a spinal cord injury in Neurosci 4: 465–466. mice. J Comp Neurol 462: 431–449. Moon LD, Asher RA, Rhodes KE, Fawcett JW (2002) Rela- Iozzo RV (1999) The biology of the small leucine-rich pro- tionship between sprouting axons, proteoglycans and glial teoglycans. Functional network of interactive proteins. cells following unilateral nigrostriatal axotomy in the J Biol Chem 274: 18843–18846. adult rat. Neuroscience 109: 101–117. Iseki K, Hagino S, Mori T et al. (2002) Increased syndecan Morgenstern DA, Asher RA, Fawcett JW (2002) Chon- expression by pleiotrophin and FGF receptor-expressing droitin sulphate proteoglycans in the CNS injury response. astrocytes in injured brain tissue. Glia 39: 1–9. Prog Brain Res 137: 313–332. Jones LL, Yamaguchi Y, Stallcup WB, Tuszynski MH Naffah-Mazzacoratti MG, Arganaraz GA, Porcionatto MA (2002) NG2 is a major chondroitin sulfate proteoglycan et al. (1999) Selective alterations of glycosaminoglycans produced after spinal cord injury and is expressed by synthesis and proteoglycan expression in rat cortex and macrophages and oligodendrocyte progenitors. J Neurosci hippocampus in pilocarpine-induced epilepsy. Brain Res 22: 2792–2803. Bull 50: 229–239. Krautstrunk M, Scholtes F, Martin D et al. (2002) Increased Okamoto M, Sakiyama J, Mori S et al. (2003) Kainic acid- expression of the putative axon growth-repulsive extracel- induced convulsions cause prolonged changes in the chon- lular matrix molecule, keratan sulphate proteoglycan, fol- droitin sulfate proteoglycans neurocan and phosphacan in lowing traumatic injury of the adult rat spinal cord. Acta the limbic structures. Exp Neurol 184: 179–195. Neuropathol (Berl) 104: 592–600. Oohira A, Katoh-Semba R, Watanabe E, Matsui F (1994a) Levine JM, Nishiyama A (1996) The NG2 chondroitin sul- Brain development and multiple molecular species of pro- fate proteoglycan: A multifunctional proteoglycan associ- teoglycan. Neurosci Res 20: 195–207. 188 F. Matsui and A. Oohira

Oohira A, Matsui F, Watanabe E, Kushima Y, Maeda N Shuo T, Aono S, Matsui F et al. (2004) Developmental (1994b) Developmentally regulated expression of a brain changes in the biochemical and immunological characters specific species of chondroitin sulfate proteoglycan, neu- of the carbohydrate moiety of neuroglycan C, a brain- rocan, identified with a monoclonal antibody 1G2 in the specific chondroitin sulfate proteoglycan. Glycoconj J 20: rat cerebrum. Neuroscience 60: 145–157. 267–278. Oohira A, Matsui F, Tokita Y, Yamauchi S, Aono S (2000) Sivasankaran R, Pei J, Wang KC et al. (2004) PKC mediates Molecular interactions of neural chondroitin sulfate pro- inhibitory effects of myelin and chondroitin sulfate pro- teoglycans in the brain development. Arch Biochem Bio- teoglycans on axonal regeneration. Nat Neurosci 7: 261– phys 374: 24–34. 268. Pasterkamp RJ, Anderson PN, Verhaagen J (2001) Periph- Snow DM, Mullins N, Hynds DL (2001) Nervous system- eral nerve injury fails to induce growth of lesioned ascend- derived chondroitin sulfate proteoglycans regulate growth ing dorsal column axons into spinal cord scar tissue cone morphology and inhibit neurite outgrowth: A light, expressing the axon repellent Semaphorin3A. Eur J Neu- epifluorescence, and electron microscopy study. Microsc rosci 13: 457–471. Res Tech 54: 273–286. Pizzorusso T, Medini P, Berardi N, Chierzi S, Fawcett JW, Snyder SE, Li J, Schauwecker PE, McNeill TH, Salton Maffei L (2002) Reactivation of ocular dominance plas- SRJ (1996) Comparison of RPTPz/b, phosphacan, and ticity in the adult visual cortex. Science 298: 1248–1251. trkB mRNA expression in the developing and adult rat Properzi F, Asher RA, Fawcett JW (2003) Chondroitin sul- nervous system and induction of RPTPz/b and phos- phate proteoglycans in the central nervous system: phacan mRNA following brain injury. Mol Brain Res Changes and synthesis after injury. Biochem Soc Trans 31: 40: 79–96. 335–336. Stipp CS, Litwack ED, Lander AD (1994) Cerebroglycan: Qi ML, Wakabayashi Y, Enomoto M, Shinomiya K (2003) An integral membrane heparan sulfate proteoglycan that Changes in neurocan expression in the distal spinal cord is unique to the developing nervous system and expressed stump following complete cord transection: A comparison specifically during neuronal differentiation. J Cell Biol between infant and adult rats. Neurosci Res 45: 181–188. 124: 149–160. Rauch U, Feng K, Zhou XH (2001) Neurocan: A brain Su YK, Elam JS (2003) Proteoglycan regulation of goldfish chondroitin sulfate proteoglycan. Cell Mol Life Sci 58: retinal explant growth on optic tectal membranes. Brain 1842–1856. Res Dev Brain Res 142: 169–175. Rauch U, Karthikeyan L, Maurel P, Margolis RU, Margolis Tang X, Davies JE, Davies SJ (2003) Changes in distribu- RK (1992) Cloning and primary structure of neurocan, a tion, cell associations, and protein expression levels of developmentally regulated, aggregating chondroitin sul- NG2, neurocan, phosphacan, brevican, versican V2, and fate proteoglycan of brain. J Biol Chem 267: 19536–19547. tenascin-C during acute to chronic maturation of spinal Rhodes KE, Fawcett JW (2004) Chondroitin sulphate pro- cord scar tissue. J Neurosci Res 71: 427–444. teoglycans: Preventing plasticity or protecting the CNS? Watanabe E, Maeda N, Matsui F, Kushima Y, Noda M, J Anat 204: 33–48. Oohira A (1995) Neuroglycan C, a novel membrane- Sango K, Oohira A, Ajiki K, Tokashiki A, Horie M, Kawano spanning chondroitin sulfate proteoglycan that is H (2003) Phosphacan and neurocan are repulsive substrata restricted to the brain. J Biol Chem 270: 26876–26882. for adhesion and neurite extension of adult rat dorsal root Wu YP, Siao CJ, Lu W et al. (2000) The tissue plasminogen ganglion neurons in vitro. Exp Neurol 182: 1–11. activator (tPA)/plasmin extracellular proteolytic system Santra M, Reed CC, Iozzo RV (2002) Decorin binds to a regulates seizure-induced hippocampal mossy fiber out- narrow region of the epidermal growth factor (EGF) growth through a proteoglycan substrate. J Cell Biol 148: receptor, partially overlapping but distinct from the EGF- 1295–1304. binding epitope. J Biol Chem 277: 35671–35681. Yamaguchi Y, Mann DM, Ruoslahti E (1990) Negative reg- Shintani T, Watanabe E, Maeda N, Noda M (1998) Neurons ulation of transforming growth factor-b by the proteogly- as well as astrocytes express proteoglycan-type protein can decorin. Nature 346: 281–284. tyrosine phosphatase z/RPTPb: Analysis of mice in which Yoneda A, Couchman JR (2003) Regulation of cytoskeletal the PTPz/RPTPb was replaced with the LacZ gene. organization by syndecan transmembrane proteoglycans. Neurosci Lett 247: 135–138. Matrix Biol 22: 25–33.