Interactions between meningeal cells and astrocytes in vivo and in vitro

187

Developmental Brain Research, 59 (1991) 187-196 © 1991 Elsevier Science Publishers B.V. 0165-3806/91/$03.50 ADONIS 016538069151249H BRESD 51249

Interactions between meningeal cells and astrocytes in vivo and in vitro K. Abnet 2'3, J.W.

Fawcett 1 and S.B. D u n n e t t 3

t Physiological Laboratory, Downing Street, Cambridge (U.K.), 2Harvard Medical School, Cambridge, MA (U.S.A.) and SDepartment of Experimental Psychology, Downing Street, Cambridge (U. K.) (Accepted 8 January 1991)

Key words: Astrocyte; Meninges; Glia limitans; Glial scar

At the interface between the meninges and the central nervous system there is a characteristic structure known as the glia limitans, consisting of many fine interdigitating astrocyte processes which contain both GFAP and vimentin, and a basal lamina. A similar structure is set up after brain injury where meningeal cells invade the lesion. We have experimentally put astrocytes and meningeal cells in contact with one another, both in vivo and in vitro, to see whether this results in the formation of a glia limitans. Cultured meningeal cells were injected into the hippocampus of adult rats, and from 1 to 12 weeks later brains were stained for GFAP and vimentin. One week after injection there was a widespread astrocytic reaction stretching up to 2 mm from the injection, the cells being stained intensely for both GFAP and vimentin. Over the next 4-6 weeks this widespread reaction subsided, the only remaining vimentin stained astrocytes, apart from those at the normal glia limitans, being in contact with the injected meningeal cells, or with meningeal cells which had migrated into the injection needle track. In vitro a structure reminiscent of the glia limitans formed where patches of astrocytes abutted meningeal cells; the astrocytes formed a layer of fine interdigitating processes all running parallel to the interface between the two cell types, and there was heavy staining for laminin and fibronectin. We conclude that a glia limitans forms wherever astrocytes and meningeal cells come into contact.

INTRODUCTION T h e meninges are comprised of m e s o d e r m a l cells which cover the surfaces of the brain and spinal cord and are s e p a r a t e d from neuronal cells of the central nervous system (CNS) by the basal lamina covered glia limitans. W h e n the m a t u r e CNS is injured, for instance by a knife cut, meningeal cells invade the cut, and a glia limitans similar to that seen at the brain surface reforms on either side of the injury 3'5'8'9'18'19'34'4°. Since a glia limitans forms where meningeal cells and astrocytes are in contact, it is r e a s o n a b l e to hypothesize that some form of interaction b e t w e e n astrocytes and meningeal cells is responsible for its formation. Several lines of evidence s u p p o r t this hypothesis. A glia limitans has been found to form, w h e t h e r during n o r m a l d e v e l o p m e n t or in response to d a m a g e , only when meningeal cells are present. Thus, n e o n a t a l injections of 6 - O H D A can d e p l e t e meningeal cells selectively over the cerebellum, and in such animals the cerebellar glia limitans fails to form n o r m a l l y in these regions 29'36. In contrast to the adult situation, knife-cut lesions of the brain in neonates do not lead to the f o r m a t i o n of a new glia limitans, and this correlates with the fact that meningeal cells do not invade such lesions

until the animals are over 1 w e e k old 3'28. In vitro, astrocytes d e v e l o p increased n u m b e r s of gap junctions when they are co-cultured with meningeal cells 1. The glia limitans is m a d e up of fine interdigitating filament-rich astrocytic processes, which are connected by gap and d e s m o s o m e - l i k e junctions, of which there are m o r e in the glia limitans than in astrocytes elsewhere, and they are covered in distinctive rectangular arrays of i n t r a m e m b r a n o u s particles 7,21. W h e r e a s the glial limitans can only be unequivocally identified electron microscopically, it can readily be visualised at the light microscopic level by its characteristic p a t t e r n of i m m u n o s t a i n i n g with vimentin, G F A P and laminin antibodies 12'17"26'34'4°. In the present p a p e r we use these light microscopic criteria to evaluate directly the effect of contact b e t w e e n cultured meningeal cells and astrocytes on the f o r m a t i o n of glia limitans-like structures, both in vivo and in vitro. MATERIALS AND METHODS

Meningeal cell cultures The meninges were stripped from the brains of newborn rat pups. These were incubated in 0.01% collagenase for 30 min, followed by 0.1% trypsin for 20 min. 0.001% DNAse was added to the mixture, then the tissue spun down, the supernatant removed and replaced

Correspondence: J. Fawcett, Physiological Laboratory, Downing Street, Cambridge, CB2 3EG, U.K.

188 with 50 mg soyabean trypsin inhibitor, 300 mg BSA and 20 /~g DNAse per 100 ml HBBS, and the tissue sucked through a 21 gauge needle 4 times. The cell suspension was filtered through 70/zm mesh, and plated in 75 cm 2 flasks in DMEM with 10% FCS.

Mixed glial cultures The meninges were largely removed from the brains of newborn rat pups, and the remaining brain tissue finely diced with a razor blade. The tissue was incubated with 0.1% trypsin for 20 rain, then 0.001% DNase added. The tissue was spun down, triturated in HBBS with DNase, BSA and trypsin inhibitor (as above), and plated onto poly-D-lysine coated glass cover slips in DMEM with 10% FCS.

Injections of meningeal cells Cells were removed from flasks with trypsin and EDTA, washed in medium, then resuspended in a small quantity of medium, with a cell concentration of 2 × 10 7 cells/ml for implantation into the hippocampus of female Sprague-Dawley rats (approx. 200 g body weight). The host rats were anaesthetized with 0.6 ml equithesin anaesthesia, and mounted in a Kopf stereotaxic frame. The cell suspensions were injected via a 10/A glass microsyringe with 23 gauge needle stereotaxically placed into the hippocampus at

-4.3 mm behind bregma, 3.5 mm lateral to the midline and 2.8 mm below dura, with the incisor bar set 2.3 mm below the interaural line. Between 2 and 4 ~tl of cell suspension was injected at the rate of 1/A/min, and a further 3 rain allowed before retraction of the needle. For controls we followed exactly the same procedure, but we heated the meningeal cell suspension up to 60 °C to kill all the cells before injecting them.

Immunohistochemistry All host animals were perfused under terminal chloral hydrate anaesthesia with 4% paraformaldehyde in 0.1 M PO4 buffer, and the brains were removed and sectioned on a cryostat at 12/~m thickness. Sections were blocked with 5% goat serum, then incubated in rabbit polyclonal anti-GFAP (1:200, DAKO) and monoclonal anti-vimentin (1:20, DAKO). Secondaries were FITC anti-rabbit and biotinylated anti-mouse (1:100, CALTAG) followed by RITC streptavidin (1:100, Serotec). Cultures were stained by fixation in 4% paraformaldehyde for 20 min, followed by blocking in 5% goat serum, then reacted with monoclonal anti-GFAP (1:20, Boehringer), anti-tetanus toxin (1:50, gift from Dr. R.O. Thomson, Wellcome labs), anti-laminin (1:250, Collaborative Research Biochemicals), anti-fibronectin (1:100, Dako), Ran-2 (supernatant 1:5) or A2B5 (supernatant 1:5). 0.2%

Fig. 1. The meningeal surface of the cerebral cortex of a normal adult rat brain. A: a GFAP stain, and B: a vimentin stain of the same section. The superficial layer of meningeal cells is stained for vimentin, but not GFAE Under this is a layer of processes which contain both GFAP and vimentin, a few of these double stained processes penetrating up to 50/~m into the brain. Beneath this superficial layer of gtia limitans, there are many astrocytic processes stained only with GFAP. Bar = 50/zm

189 Triton X-100 was present for GFAP staining. For tetanus toxin staining, the cultures were incubated in the toxin (1:200, Calbiochem) for 20 mins before fixation. The secondary antibodies were the same as used for the brain sections

detectable amounts of vimentin (Fig. 2). Some of the larger blood vessels were surrounded by processes containing vimentin alone or vimentin and GFAP. A few astrocytes in the corpus callosum also contained vimentin.

RESULTS

GFAP and Vimentin staining of normal rat brain

GFAP and Vimentin staining of brains injected with meningeal cells

As described by previous authors, some though not all astrocytes throughout the brain were stained for G F A P 4. The intensity of staining was relatively low in most cells, although it was greater in the corpus callosum. Vimentin staining was restricted to the meninges, and to the astrocytic processes of the glia limitans immediately underlying them; these processes also stained strongly for GFAP, as described by previous authors 4'3°'35 (Fig. 1). The ependymal cell layer also stained for vimentin, but the astrocytic processes underlying it did not contain

We first examined brains 7 d after a localized injection of meningeal cells had been made into the hippocampus. By this time a widespread gliotic reaction was evident around the injection. This was apparent as intense G F A P and vimentin immunoreactivity of many astrocytes surrounding the injection site and needle track. Vimentin staining was only seen within approx. 450 /tm of the injection site, whereas intense G F A P staining typically extended further, for up to 1.5 m m (see Fig. 3). These findings agree with previous observations in which the

Fig. 2. The ventricular surface of the rat brain, in the lateral ventricle. The ependymal cells are stained intensely for vimentin (B) and less so for GFAP (A). There is no layer of astrocytic processes which stain for both GFAP and vimentin underlying the ependymal cells. Bar = 50/~m

190 glial response to penetrating injuries to the brain have been studied lk12"2°'34. Injected meningeal cells, seen both at the injection site and along the needle track, stained strongly with vimentin antibody, but were GFAP-negative. Brains were also examined at 4, 6, 8, 10 and 12 weeks after injection of meningeal cells. The main change over this period was that the widespread glial reaction surrounding the injection site and needle track gradually subsided, the astrocytes losing their vimentin staining, and becoming less intensely stained for GFAP (Fig. 4). However, the astrocytic processes immediately surrounding the meningeal cells in the injection site and needle track remained both vimentin and GFAP-positive. The injection site itself remained approximately the same size, and the meningeal cells continued to stain with vimentin antibody. By 12 weeks, the longest survival time we examined, the only vimentin-positive astrocytic processes were those forming the glia limitans under the meninges, those immediately surrounding the injected meningeal

cells, and a few cells in the corpus callosum and around blood vessels. Also, by this time, some blood vessels adjoining the injection had a thick layer of vimentinpositive GFAP-negative cells surrounding them (fig. 5), suggesting that some meningeal cells had migrated for up to 400/xm along blood vessel walls. We made control injections of heat killed meningeal cells into 4 animals, which were allowed to survive to 10 weeks. In 3 of these animals we were unable to find any vimentin staining at the injection site, and GFAP staining was similar to the uninjected side of the brain. However the needle track in the superficial part of the cortex was extensively invaded with meningeal cells which had presumably migrated in from the surface of the brain, and the appearance here was identical to animals injected with live meningeal cells, with vimentin staining of the meningeal cells and the immediately adjacent astrocytes. In one of our controls there were a few vimentin stained cells at the injection site, which must have migrated from the surface of the brain, or from a large blood vessel.

Fig. 3. A parasaginal section of the hippocampus of a rat which had received an injection of meningeal cells 7 d previously. A and C: GFAP stains, and B and D: stained for vimentin. (C) and (D) are taken from the same section as (A) and (B), but at twice the magnification. The injected meningeal cells stain intensely for vimentin, and there are reactive astrocytes stained for both GFAP and vimentin stretching for about 400/xm from the injection site. Bars = 50/xm

191

Mixed glia cultures Cultures became confluent after 6 to 10 days. Initially GFAP-containing cells were randomly interspersed with cells which contained no GFAP. These stained with antibodies to fibronectin, and were therefore most likely to be meningeal cells or fibroblasts. By 2 weeks the cultures were predominantly composed of astrocytes, as determined by phase contrast or antibody staining, but there were patches composed almost exclusively of meningeal cells embedded in them. The cultures also contained areas in which many ciliated cells were inter-

mixed with astrocytes, and these were presumably ependymal cells. By 3-4 weeks of culture, the pattern of cellular distribution had become consolidated and stable. The most systematic staining was conducted at this survival time, using antibodies specific for different cell types. In particular, the G F A P antiserum stains astrocytes, R A N - 2 stains the surfaces of both astrocytes and meningeal cells 2, A2B5 stains neurons and cells of the type 2 astrocyte oligodendrocyte lineage 27, tetanus toxin and antitoxin has a similar staining pattern to A2B5, and

Fig. 4. An injection site 4 weeks after injection. A: a GFAP stain, and B: a vimentin stain. There is still intense vimentin staining in the graft, but double stained reactive astrocytes only stretch for 100 ~m from the injection site. The GFAP staining around the graft is still strong. Bar = 1 0 0 ~um

192 l a m i n i n a n d f i b r o n e c t i n stain the extracelluar matrix

Clear b o u n d a r i e s were seen b e t w e e n regions of astro-

associated with m a n y different cell types. V i m e n t i n has

cytes, a n d patches of m e n i n g e a l cells. A t the interface

n o t b e e n f o u n d to be a particularly useful m a r k e r in

b e t w e e n the two cell types t h e r e was often a 'wall' of

c u l t u r e , since most c u l t u r e d cells are at least m o d e r a t e l y

m a n y fine e l o n g a t e d astrocytic processes, r u n n i n g paral-

immunoreactive.

lel to the a s t r o c y t e - g l i a l b o u n d a r y , r a t h e r similar in

Fig. 5. Meningeal cell injection sites 12 weeks after injection. A, C andE: GFAP stains, and B, D and F: vimentin stains of the same sections. The injected meningeal cells are still present, and stain brightly for vimentin. However, the vimentin staining hardly extends beyond the graft; astrocytes double stained for GFAP and vimentin are only seen immediately adjacent to the graft. However, as judged from the GFAP stain, the astrocytes within 50 -100/zm of the graft still look reactive. The only vimentin staining seen any distance from the graft is associated with blood vessels. This is probably due to cells which have migrated along blood vessels, since the layer of vimentin-containingcells around these vessels close to grafts is very much thicker than that around a normal blood vessel. Bars = 50/~m.

193

Fig. 6. The meningeal cell-astrocyte boundary in tissue culture. A and C: GFAP stains. The diagram is a tracing of (A). In (A) there is a patch of astrocytes to the top of the picture, labelled 'A' in the diagram, and a boundary between astrocytes and meningeal cells around it. At this boundary there is a 'wall' of parallel astrocyte processes, labelled 'W' in the diagram, with some 'whiskers' of astrocyte processes, labelled 'X' in the diagram, coming out under the meningeal cells. The meningeal cells extend right up to the 'wall', the'whiskers' running underneath them. B: a tetanus toxin stain, showing that cells of the oligodendrocyte lineage are found on the surface of the astrocytes, but not on the meningeal cells. No oligodendrocytes are found on the 'whiskers', since these run underneath meningeal cells, and their surface is therefore not exposed for the oligodendrocytes to adhere to. C: a triangular area of meningeal cells, with astrocytes on either side, and a 'wall' of astrocyte processes at each interface. Bar = 50/~m. a p p e a r a n c e to a glia limitans. Radiating out from this wall t h e r e was usually a 'fringe' of fine astrocyte processes, which lay u n d e r n e a t h the layer of meningeal cells (see Fig. 6). T h e meningeal cells stained strongly with fibronectin and with R A N - 2 , and also stained with laminin, the laminin staining being particularly strong at the b o u n d a r i e s b e t w e e n astrocytes and meningeal cells (Fig. 7). A few astrocytes were generally seen in the meningeal areas, and these typically t o o k the form of strands of elongated, highly G F A P - p o s i t i v e cells running u n d e r n e a t h the meningeal cell layer. Astrocytes in contact with meningeal cells did not have the flat polygonal a p p e a r a n c e of cells in contact only with o t h e r astrocytes. In regions of the culture which contained

ciliated e p e n d y m a l cells, the astrocyte m o r p h o l o g y was not recognisably different from regions which contained astrocytes alone. DISCUSSION These results d e m o n s t r a t e that when meningeal cells and astrocytes are a p p o s e d , the astrocytes set up a layer of processes at the interface which are similar to the glia limitans found at the n o r m a l m e n i n g e a l surface of the brain, and which contain the same mix of i n t e r m e d i a t e filaments, G F A P and vimentin. A similar response is seen both in vitro and in vivo. This astrocytic response is relatively specific to meningeal cells since no similar

194

Fig. 7. Patches of meningeal cells in a mixed glial culture stained for fibronectin (A) and laminin (B, C). The astrocytes on either side of the strip of meningeal cells which occupies the centre of each picture have some fine particulate staining for both molecules. The meningeal cells all stain intensely for fibronectin, but laminin staining is rather more concentrated at the astrocyte-meningeal cell interface. Bar = 50/~m.

structures form at locations where astrocytes are in contact with ependymal cells. In tissue culture, where a patch of astrocytes abuts a patch of meningeal cells, we found a 'wall' of astrocytic processes, which ran parallel to the interface between the cell types. In addition, individual astrocytes in contact with meningeal cells did not have the flat polygonal appearance typical of astrocytes which are in contact with their own type, but instead were nearly always elongated, with fine, highly-GFAP immunoreactive processes. Meningeal cell injections into the mature brain initially caused a widespread astrocytic reaction, visible as an increase in G F A P staining and appearance of vimentin

immunoreactivity around the injection. However by approximately 4 - 6 weeks after injection this generalised response largely disappeared, leaving a thin layer of vimentin-containing astrocytic processes at the interface between brain and meningeal cells. This stable long-term appearance of the interface between meningeal cells and astrocytes was very similar to the layer of astrocytic processes found at the normal meningeal interface. Our results are consistent with the hypothesis that there is a local interaction between astrocytes and meningeal cells which causes the astrocytes to m a k e a wall of fine processes, resulting in the formation of the glia limitans. This interaction is a local one, since only

195 astrocytes immediately adjacent to meningeal cells show these changes. Indeed, the response may require direct cell-cell contact, since the signal to the astrocytes does not extend through the medium in tissue culture. Moreover, we have been unable to induce similar morphological changes in astrocyte cultures using meningeal cell conditioned medium. It is possible that the stimulus for the formation of a glia limitans is a basal lamina, since wherever astrocytes contact a basal lamina there is a glia limitans, i.e. at the meningeal interface, at the interface with Schwann cells in the dorsal and ventral roots of the spinal cord, and there are vimentin-containing glial endfeet on the larger blood vessels, which are surrounded by mesenchymal cells similar to meningeal cells. On the other hand there is no basal lamina between ependymal cells and astrocytes, and neither is there a glia limitans ~6. Equally, when immature brain tissue or selected astrocytes are transplanted into the mature brain, in the absence of meningeal cells, no glia limitans is seen to form around the graft, although a short-term generalised gliotic reaction still Occurs 1°'2(~'22'37. Lyser, looking at embryonic chick spinal cord in organ culture, also found that where mesenchymal cells abut astrocytes a glia limitans is set up 25. The cytoskeletal changes which occur in astrocytes in contact with meningeal cells are the same as those in astrocytes which show the short-lived and widespread glial reaction following injury. However, the signal which causes the two types of glial reaction is likely to be different, since the type of short-lived widespread gliosis which we observed can be induced by a wide range of non-specific injuries to the brain, and begins before there has been much meningeal cell invasion of the cut, which does not occur until 4 d after lesioning 3'5'8"11"26"34. Candidates for the signal causing the generalised shortterm response include the peptides described by Giulian et al.~S, and factors such as FGF and P D G F 3"39. Meningeal cells are probably responsible for the initial formation of glia limitans around areas of brain injury, but their role in the formation of large persistent glial scars, such as occur after penetrating lesions of the CNS, is ambiguous. There are certainly similarities between the astrocytic processes which form the glia limitans and those that are found in scars: both contain many fine interweaving processes, both contain vimentin, both have REFERENCES 1 Anders, J.J. and Salopek, M., Meningeal cells increase in vitro astrocytic gap junctional communication as measured by fluorescence recovery after laser photobleaching, J. Neurocytol., 18 (1989) 257-264. 2 Bartlett, P.E, Noble, M.D., Pruss, R.M., Raft, M.C., Rattray,

laminin around them, and both contain orthogonal particle arrays 11'12'23'34"4°'41. However, since the effects of meningeal cells appear to operate over such short distances, a large glial scar would have to be full of meningeal cells if these were the primary cause. Spinal cord lesions, which are particularly prone to scarring, have been reported to contain numbers of infiltrating mesodermal cells 1s'32. However, gliosis can clearly occur in the absence of meningeal infiltration, for instance in areas of axonal degeneration and demyelination. There has been speculation for at least a century as to the role of the glial scar in preventing the regeneration of axons in the central nervous system 32"33. It is not clear that reactive astrocytic processes constitute any barrier to axon growth 31, although astrocytic processes forming a glia iimitans may do 24, and three dimensional cultures of mammalian astrocytes have been shown to be relatively non-permissive for the regeneration of axons 14. On the other hand, meningeal cells in monolayer tissue culture have been shown to be non-permissive to the growth of retinal axons although they permit the growth of axons from dorsal root ganglia 13. It seems probable, therefore, that the layer of meningeal cells which invades many CNS injuries might form a barrier for the growth of axons. This does not in any way provide a general explanation for the failure of axonal regeneration in the CNS, but it might be relevant to the problems involved in grafting tissue into the CNS. Krueger et al. 2° describe the formation of a glia limitans, accompanied by meningeal cells, around parts of embryonic cortical grafts, and provide some evidence that these meningeal interfaces are permanent, and therefore prevent the graft from connecting with the host in these regions. It is, however, presumably possible for such barriers to be breached, since a successful strategy for grafting embryonic tissue to the mature brain is to make a 'delayed cavity' some weeks before inserting the graft. This procedure is necessary for grafting solid pieces of embryonic tissue to sites in the adult brain that are not richly vascularised, but the cavity becomes lined with meningeal cells, and a glia limitans is formed. Nevertheless, the grafts not only survive in these cavities but axons can breach this layer, at least in places, and give rise to an extensive innervation of the host brain 6"38. This ability to dissolve scar tissue may be restricted to embryonic tissue.

S. and Williams, C.A., Rat neural antigen-2 (RAN-2): A cell surface antigen on astrocytes, ependymal cells, Muller cells and lepto-meninges defined by a monoclonal antibody, Brain Res., 204 (1981) 339-351. 3 Berry, M., Maxwell, W.L., Logan, A., Mathewson, A., McConnell, P., Ashurst, D.E. and Thomas, G.H., Deposition of scar tissue in the central nervous system, Acta Neurochirurgica,

196 Suppl. 32 (1983) 31-53. 4 Bignami, A. and Dahl, D., Differentiation of astrocytes in the cerebellar cortex and the pyramidal tracts of the newborn rat. An immunofluorescence study with antibodies to a protein specific to astrocytes, Brain Res., 49 (1973) 393-402. 5 Bignami, A. and Dahl, D., The astroglial response to stabbing. Immunofluorescent studies with antibodies to astrocyte specific protein (GFA) in mammalian and submammalian vertebrates, Neuropath. Appl. Neurobiol., 2 (1976) 99-111. 6 Bjorklund, A. and Stenevi, U., Intraeerebral neural implants: neuronal replacement and reconstruction of damaged circuitries, Ann. Rev. Neurosci., 7 (1984) 279-308. 7 Brightman, M.W. and Reese, T.S., Junctions between intimately apposed cell membranes in the vertebrate brain, J. Cell Biol., 40 (1%9) 648-679. 8 Carbonell, A.L. and Boya, J., Ultrastructural study on meningeal regeneration and meningo- glial relationships after cerebral stab wound in the adult rat, Brain. Res., 439 (1988) 337-344. 9 Cavanagh, J.B., The proliferation of astrocytes around a needle wound in the rat brain, J. Anat., 106 (1970) 471-487. 10 Connor, J.R. and Bernstein, J.J., Astrocytes in rat fetal cerebral cortical homografts following implantation into adult rat spinal cord, Brain Res., 409 (1987) 62-70. 11 Dahl, D., Bignami, A., Weber, K. and Osborn, M., Filament proteins in rat optic nerves undergoing Wallerian degeneration: Localization of vimentin, the fibroblastic 100-A filament protein, in normal and reactive astrocytes, Exp. Neurol., 73 (1981) 496-506. 12 Eng, L.E, Reier, EJ. and Houle, J.D., Astrocyte activation and fibrous gliosis: glial fibrillary acidic protein CNS tissue, Prog. Brain. Res., 71 (1987) 439-455. 13 Fawcett, J.W., Bakst, I. and Rokos, J., Interactions between glial cells and axons in vitro, Soc. Neurosci. Abstr., 13 (1987) 1483. 14 Fawcett, J.W., Housden, E., Smith-Thomas, L. and Meyer, R.L., The growth of axons in three dimensional astrocyte cultures, Dev. Biol., 135 (1989) 449-458. 15 Giulian, D., Allen, R.L., Baker, T.J. and Tomozawa, Y., Brain peptides and glial growth. I. Glia-promoting factors as regulators of gliogenesis in the developing and injured central nervous system, J. Cell Biol., 102 (1986) 803-811. 16 Gotow, T. and Hashimoto, EH., Fine structure of ependyma and intercellular junctions in the area postrema of the rat, Cell Tiss. Res., 201 (1979) 207-225. 17 Janeczko, K., Spatiotemporal patterns of the astrogiiai proliferation in rat brain injured at the postmitotic stage of postnatal development: a combined immunocytochemical and autoradiographic study, Brain Res., 485 (1989) 236-243. 18 Krikorian, J.G., Guth, L. and Donati, E.J., Origin of connective tissue scar in the transected rat spinal cord, Exp. Neurol., 72 (1981) 698-707. 19 Krueger, S., Sievers, J., Hansen, C., Sadler, M. and Berry, M., Three morphologically distinct types of interface develop between adult host and fetal brain transplants: implications for scar formation in the adult central nervous system, J. Comp. Neurol., 249 (1986) 103-116. 20 Krueger, S., Sievers, J., Hansen, C., Sadler, M. and Berry, M., Three morphologically distinct types of interface develop between adult host and fetal brain transplants: implications for scar formation in the adult central nervous system, J. Comp. Neurol., 249 (1986) 103-116. 21 Landis, D.M. and Reese, T.S., Arrays of particles in freezefractured astrocytic membranes, J. Cell Biol., 60 (1974) 316-320. 22 Lawrence, J.M., Huang, S.K. and Raisman, G., Vascular and astrocytic reactions during establishment of hippocampal transplants in adult host brain, Neuroscience, 12 (1984) 745-760.

23 Liesi, P., Kaakola, S., Dahl, D. and Vaheri, A., Laminin is induced in astrocytes of adult brain by injury, EMBO J., 3 (1983) 683-686. 24 Liuzzi, EJ. and Lasek, R.J., Astrocytes block axonal regeneration in mammals by activating the physiological stop pathway, Science, 237 (1987) 642-645. 25 Lyser, K.M., The differentiation of glial cells and gila limitans in organ cultures of chick spinal cord, In Vitro, 8 (1972) 77-84. 26 Mathewson, A.J. and Berry, M., Observations on the astrocyte response to a cerebral stab wound in adult rats, Brain. Res., 327 (1985) 61-69. 27 Miller, R.H., Ffrench-Constant, C. and Raft, M.C., The macrogiial cells of the rat optic nerve, Annu. Rev. Neurosci., 12 (1989) 517-534. 28 Moore, I.E., Buontempo, J.M. and Weiler, R.O., Response of fetal and neonatal rat brain to injury, Neuropathol. Appl. Neurobiol., 13 (1987) 219-228. 29 Pehlemann, EW., Sievers, J. and Berry, M., Meningeal cells are involved in foliation, lamination, and neurogenesis of the cerebellum: evidence from 6-hydroxydopamine induced destruction of meningeal cells, Dev. Biol., 110 (t985) 136-146. 30 Pixley, S.K.R. and de Vellis, J., Transition between immature radial glia and mature astrocytes studied with a monoclonal antibody to vimentin, Developmental Brain Research, 15 (1984) 201-209. 31 Reier, P.J., Penetration of grafted astrocytic scars by regenerating optic nerve axons in Xenopus tadpoles, Brain Res., 164 (1979) 61-68. 32 Reier, P.J. and Houle, J.D., The glial scar: its bearing on axonal elongation and transplantation approaches to CNS repair, Adv. Neurol., 47 (1988) 87-138. 33 Reier, P.J., Stensaas, L.J. and Guth, L., The astrocytic scar as an impediment to regeneration in the central nervous system~ In C.C. Kao, R.P. Bunge and P.J. Reier (Eds,), Spinal cord reconstruction, Raven Press, New York, 1983, pp. 163-195. 34 Schiffer, D., Giordani, M.T., Migheti, A., Giaecone, G., Pezotta, S. and Mauro, A., Glial fibrillary acidic protein and vimentin in the experimental glial reaction in the rat brain, Brain Res., 374 (1986) 110-118. 35 Schnitzer, J., Franke, W.W. and Schachner, M., Immunocytochemical demonstration of vimentin in astrocytes and ependymal cells of developing and adult mouse nervous system, J. Cell Biol., 90 (1981) 435-447. 36 Sievers, J. and Pehlemann, EW., Influences of meningeal cells on brain development. Findings and hypothesis, Naturwissenschaften., 73 (1986) 188-194. 37 Smith, G.M., Miller, R.H. and Silver, J., Changing role of forebrain astrocytes during development, regenerative failure, and induced regeneration upon transplantation, J. Comp. Neurol., 251 (1986) 23-43. 38 Stenevi, U., Kromer, L.F., Gage, EH. and Bj6rklund, A., Solid neural grafts in intracerebral transplantation cavities. In A. Bj6rklund and U. Stenevi (Eds.), Neural grafting in the mammalian CNS, Elsevier, Amsterdam, 1985, pp. 41-50. 39 Takamiya, Y., Kohsaka, S., Toya, S., Otani, M., Mikoshiba, K. and Tsukada, Y., Possible association of platelet-derived growth factor (PDGF) with the appearance of reactive astroeytes following brain injury in situ, Brain Res., 383 (1986) 305-309. 40 Takamiya, Y., Kohsaka, S., Toya, S., Otani, M. and Tsukada, Y., Immunohistochemical studies on the proliferation of reactive astrocytes and the expression of cytoskeletal proteins following brain injury in rats, Brain. Res., 466 (1988) 201-210. 41 Wolburg, H. and Kaestner, R., Is the architecture of astrocytic membrane crucial for axonai regeneration in the central nervous system, Naturwissenschaften., 71 (1984) 484-485.