Distant microglial and astroglial activation secondary to experimental spinal cord lesion

Ativação microglial e astroglial à distância secundárias a lesão da medula espinhal

DISTANT MICROGLIAL AND ASTROGLIAL ACTIVATION SECONDARY TO EXPERIMENTAL SPINAL CORD LESION

Ricardo José de Almeida Leme¹, Gerson Chadi¹

ABSTRACT - This paper analysed whether glial responses following a spinal cord lesion is restricted to a scar formation close to the wound or they might be also related to widespread paracrine trophic events in the entire cord. Spinal cord hemitransection was performed in adult rats at the thoracic level. Seven days and three months later the spinal cords were removed and submitted to immunohistochemistry of glial fibrillary acidic protein (GFAP) and OX42, markers for astrocytes and microglia, as well as of basic fibroblast growth factor (bFGF), an astroglial neurotrophic factor. Computer assisted image analysis was employed in the quantification of the immunoreactivity changes. At the lesion site an increased number of GFAP positive astrocytes and OX42 positive phagocytic cells characterized a dense scar formation by seven days, which was further augmented after three months. Morphometric analysis of the area and microdensitometric analysis of the intensity of the GFAP and OX42 immunoreactivities showed reactive astrocytes and microglia in the entire spinal cord white and gray matters 7 days and 3 months after surgery. Double immunofluorescence demonstrated increased bFGF immunostaining in reactive astrocytes. The results indicated that glial reaction close to an injury site of the spinal cord is related to wounding and repair events. Although gliosis constitutes a barrier to axonal regeneration, glial activation far from the lesion may contribute to neuronal trophism and plasticity in the lesioned spinal cord favoring neuronal maintenance and fiber outgrowth.

KEY WORDS: astrocyte, microglia, image analysis, immunohistochemistry.

Ativação microglial e astroglial à distância secundárias a lesão da medula espinhal

RESUMO - Este trabalho analisou se a resposta glial após a lesão da medula espinhal é restrita à formação da cicatriz no local do trauma ou pode também estar relacionada a efeitos tróficos parácrinos em toda a medula espinhal. Hemitransecção da medula espinhal de ratos adultos foi realizada em nível torácico baixo. Após 7 dias e 3 meses as medulas espinhais dos animais foram removidas e submetidas à imuno-histoquímica para a visualização da proteína fibrilar glial ácida GFAP (marcador do astrócito), do OX42 (marcador da microglia), e do fator de crescimento básico derivado do fibroblasto (bFGF), produzido pelo astrócito. Para a quantificação das alterações dos perfis imunoreativos utilizou-se análise de imagem computadorizada. No local da lesão observou-se aumento no número de astrócitos GFAP positivos bem como de células fagocíticas OX42 positivas, caracterizando uma formação cicatricial densa que aumentou aos 3 meses da lesão. A análise morfométrica da área e microdensitométrica da intensidade das imunoreatividades do OX42 e da GFAP mostrou microglia e astrócitos reativos nas substâncias cinzenta e branca em toda a medula espinal 7 dias e 3 meses após cirurgia. Marcação imunofluorescente com 2 cromógenos demonstrou o aumento da expressão do bFGF nos astrócitos reativos. Estes resultados mostram que a ativação glial no local da lesão da medula espinhal é necessária ao processo de cicatrização e reparo do tecido nervoso. Embora a gliose constitua uma barreira à regeneração axonal, a ativação glial em níveis distantes da lesão pode contribuir para o trofismo e plasticidade neuronal favorecendo a manutenção neuronal e o crescimento de fibras nervosas.

PALAVRAS-CHAVE: astrócito, microglia, análise de imagem, imuno-histoquímica.

Central nervous system (CNS) injury leads to an inflammatory response being microglia, blood-born macrophages and astrocytes the main protagonists. Astroglial and microglial responses depend on the nature of the injury, the microenvironment at the lesion site as well as the distance from the lesion¹. Microglia and astrocytes become activated right after the CNS lesion, undergoing remarkable changes in cell shape and functionality². Those cells proliferate and increase in size. Although activated microglia can show increase in the cytoplasmic processes or retraction of them, the latter called ameboid microglia, depending on the state of activation, astrocytes produce a thick network of processes that surrounds the lesioned area³. However, this can not explain the glial reaction that spreads far from the injury and in the areas of the lesioned CNS that are not involved in the primary lesion⁴. It is largely known that reactive microglia in the vicinity of the lesion is involved in the elimination of debris and in association with astrocytes participate in the wounding and repair events, thus contributing to scar formation³.

Considerable evidences are available to support the concept that astrocytes and microglia interact to trigger actions that support neuronal survival. It is known that activated microglia also secrete products that regulate astroglial growth and proliferation, being considered as a principal source of astroglia-stimulating peptides^3,5. Furthermore, activated astroglia is able to release proteins like basic fibroblast growth factor (bFGF) that can promote neuronal trophic support and plasticity which may be essential in regions of the lesioned nervous tissue far from injury^6-8.

In this study, quantitative image analysis in combination with immunohistochemistry was employed to show the astroglial and microglial activations and also the expression of glial bFGF in the lesioned rat spinal cord close to and far from injury site. The results are discussed in the view of the properties of reactive glial cells mediated repair and spinal cord trophic support in order to prevent secondary neuronal degeneration.

METHOD

1 - Spinal cord lesion

Adult male Wistar rats [body weight (b.w.) 220-250 g] from the Institute of Biomedical Sciences (São Paulo, Brazil) were kept under controlled temperature, humidity and light conditions. The animals received chloral hydrate anaesthesia (Merck, 0.6 mg/100 mg b.w.) and were placed in a spinal cord unit of a stereotaxic apparatus (Kopf). By means of an adjustable wire knife (Kopf) the rats had their spinal cord lesioned in the right side (n=7) or were submitted to sham operation (n=4). Laminectomy was performed at 11^th thoracic vertebrae with delicate twisers and the guide of the wire knife was placed in a vertical plane close to the lateral surface of the low thoracic level of the spinal cord. This level was chosen so that cranial and caudal unlesioned segments of the spinal cord could be analysed. The knife, that was previously turned medially, was then extended 1.5 mm and the guide was lifted 4.0 mm to transect the spinal cord. The knife was then withdrawn, the guide removed and the skin was sutured. Sham operation consisted of only laminectomy.

The animals were sacrificed 7 days or 3 months after surgery accordingly immunohistochemistry techniques⁷. The cords were reduced in small pieces, orientated in the rostro-caudal axis and frozen.

2 - Immunohistochemical procedures

2.1 - One-color immunoperoxidase experiments for the demonstration of GFAP and OX42 immunoreactivities

Adjacent serial 25µm thick transversal frozen sections were obtained with a cryostat (Leica, CM 3000, Germany) from the entire spinal cord according to Paxinos and Watson atlas⁹ (Fig 1). Thaw-mounted sections were sampled systematically during sectioning. One hundred series in a rostro-caudal order including every 100^th section were used for immunohistochemistry accordingly to described by Chadi and collaborators¹⁰. Immunoreactivity was detected by the avidin-biotin peroxidase technique as previously described by Chadi and collaborators¹¹. Sections from two series were then incubated for 48 hours at 4ºC with one of the following antisera: a rabbit polyclonal antisera against glial fibrillary acidic protein (GFAP) (Dakopatts, Denmark) diluted 1:1200 and a mouse monoclonal OX42 antibody (Harlan, UK) diluted 1:1000. Sections were then incubated with biotinylated either goat anti-rabbit or horse anti-mouse immunoglobulins diluted 1:200 (Vector, USA) for 2 hours and with an avidin-biotin peroxidase complex (both diluted 1:100, Vectastain, Vector, for 90 minutes). Immunoreactivity was visualized using 3-3'-diaminobenzidine tetrahydrochloride (DAB, Sigma) as a chromogen and H₂O₂ (0,05%, v/v, Sigma). The GFAP antibody recognizes the major protein of the cytoskeleton of astrocytes and the OX42 antisera identifies the complement CR3 receptor and was here used as a marker of microglia/macrophages cells.

2.2 - Two-color immunofluorescence experiments

The two-color immunofluorescence method was employed in a series of sections for simultaneous detection of GFAP and bFGF immunoreactivities as previously described by our group⁷. Briefly, sections were incubated overnight at 4ºC with a mixture of the rabbit polyclonal bFGF (1:600) and a mouse monoclonal GFAP (1:100) (Boehringer Mannheim, Germany) antisera. Immunoreactivities were visualized with fluorescein isothiocyanate-conjugated donkey anti-rabbit and Texas Red-conjugated donkey anti-mouse immunoglobulins (Jackson, U.S.A.) to demonstrate the bFGF and GFAP antibodies, respectively. The sections were examined in an Olympus AX70 epifluorescence microscope (U.S.A.).

3 - Image analysis

The semiquantitative morphometric and microdensitometric analyse of the GFAP and OX42 immunoreactivities were made on both sides of the spinal cord in two transversal sections per rat at rostro-caudal levels (cervical (C4), low thoracic (T12) and lumbar (L3) segments). The low thoracic corresponded to the level of the lesioned site. The image analysis procedures implemented on a Kontron-Zeiss KS400 image analyzer (Germany) have been described previously¹¹. Briefly, a television camera from the microscope (x40 objective) acquired the image. The fields (Fig 1) were selected within the funiculum lateralis (white matter) and in the anterior horn of the gray matter, bilaterally. After shading correction, a discrimination procedure was performed according to the following: the mean gray value (MGV) and s.e.m. of gray matter or white matter in areas of the spinal cord devoid of specific labeling (background, bg) was measured. Gray values darker than bgMGV ¾ 3 s.e.m. were considered as belonging to specific labeling and thus discriminated. The specific (sp) MGV was then defined as the difference between the bgMGV value and the MGV of discriminated profiles. The microdensitometric analysis was employed in the detection of the spMGV of the GFAP and OX42 immunoreactivities. This parameter reflects the amount per cell of the measured immunoreactivities. The glass value was left constant at 200 MGV. The procedure was repeated for each section to correct every measurement of specific labeling for its background value. The size of the sampled field was 2,48 x 10⁴µm².

In the morphometric evaluation, the areas of the GFAP and OX42 immunoreactive discriminated profiles, including cytoplasm and process, were measured and expressed as area per unit area (area / µm²). The size of the sampled field was described above.

4 - Statistical analysis

The values of area and the spMGV of the GFAP and OX42 immunoreactive discriminated profiles found in all sample fields of the 2 sections per level bilaterally were correlated after introducing them in a X,Y axis, respectively. The values were then plotted. The plotted values of the cervical, low thoracic and lumbar levels of the sham, 7 day and 3 month groups were compared by means of an analysis of variance with multivariate tests of significance (Tukey) for the statistical evaluation of the effects in the groups.

RESULTS

Analysis of OX42 immunoreactivity

In the white (Fig 2A,D) and gray (not shown) matters of the sham operated rats the OX42 immunoreactivity was found in the small cells, which sent thin and less ramified processes. The OX42 immunoreactive profiles were homogeneously distributed throughout regions of the entire spinal cord white and gray matters of the sham operated rats. In the hemitransected rats, OX42 immunoreactive profiles showed an enlarged cytoplasm and thick processes throughout spinal cord white matter bilaterally (Fig 2B,C,E,F). The OX42 immunoreactivity was markedly increased in the white matter regions close to the wound (Fig 2B,C). A massive appearance of round shape OX42 immunoreactive profiles without any processes were found bordering the lesion site of the 7 day rats (Fig 2B). A similar pattern of distribution of OX42 immunoreactivity was found in the gray matter of the lesioned rats, however, round shape immunoreactive profiles without processes were not observed (data not shown).

The multivariate test indicated significant differences of the spMGV/area plotted values, as determined by microdensitometric and morphometric image analyses, of the OX42 immunoreactive discriminated profiles in the sampled fields of the white and gray matters of cervical, low thoracic and lumbar levels of the spinal cord of the sham operated, 7 day and 3 month groups (Fig 3).

Seven days after the hemitransection, the spMGV of the OX42 immunoreactive discriminated profiles was increased in the sampled fields of the white and gray matters at cervical, low thoracic and lumbar levels compared to sham operated rats (Figs 2, 3; Table 1); after 3 months the values of the white matter were slightly further elevated at cervical and lumbar levels but were massively diminished at low thoracic level, and those of the gray matter were in fact decreased in the three studied levels compared to the 7 day hemitransected rats (Figs 2, 3; Table 1).

The area of the OX42 immunoreactive discriminated profiles in the sampled fields of the white matter was elevated at low thoracic level of the 7 day lesioned rats and also in all studied levels of the 3 month hemitransected group, however the time-induced increases in the area of the OX42 immunoreactivity in the white matter regions seemed to be less pronounced (p = 0,051) at low thoracic level close to the lesion (Figs 2, 3; Table 1). Furthermore, the substantial increases in the area of the OX42 immunoreactive profiles were found in the sampled fields of the gray matter of the lesion level and bellow to it which were only further attenuated after 3 months since the values after the later post lesion period were different from the sham operated rats (Figs 2, 3; Table 1).

Analysis of GFAP immunoreactivity

The analysis of GFAP immunoreactivity in the white matter of the spinal cord of the sham operated rats demonstrated the GFAP immunoreactive profiles similar to fibrous astrocytes (Fig 2G,J). The GFAP immunoreactive profiles were sparsely and homogeneously distributed throughout the white matter. These cells were elongated and had a small cytoplasm which accumulated low amount of GFAP immunoreactivity and possessed thin processes. Primary and secondary branches of these profiles were seen. The GFAP immunoreactivity was increased in spinal cord white matter of the lesioned group bilaterally 7 days and 3 months after the surgery (Fig 2H,I,K,L). The most remarkable increases of the GFAP immunoreactivity occurred in white matter regions close to the wound, which tended to decrease distally from the lesion (Fig 2G-L).

GFAP immunoreactive profiles similar to protoplasmic astrocytes were found homogeneously distributed in the gray matter of different levels of the spinal cord of the sham operated rats. Likewise in the white matter these profiles showed thin cytoplasm and ramified processes. The GFAP immunoreactivity was also increased bilaterally in all rostro-caudal levels of the lesioned rats, remarkably in the gray matter close to the injury. The increased GFAP immunoreactivity observed in the white and gray matters of the lesioned rats was due to an augmented number of GFAP immunoreactive profiles which showed larger cytoplasm and more ramified and thick processes (Fig 2H,I,K,L).

The multivariate test indicated significant differences of the spMGV/area plotted values, as determined by microdensitometric and morphometric image analyses, of the GFAP immunoreactive discriminated profiles in the sampled fields of the white and gray matters of cervical, low thoracic and lumbar levels of the spinal cord of the sham operated, 7 day lesioned and 3 month lesioned groups.

The spMGV of the GFAP immunoreactive discriminated profiles was increased in the white and gray matter sampled fields of all three studied levels of the 7 day and 3 month lesioned groups compared to sham operated rats (Figs 2, ⁴; table 1); three months after lesion the values in fact were decreased in the white matter of cervical and lumbar levels as well as in the gray matter of all studied levels but were augmented in the white matter of the low thoracic level compared to the 7 day lesioned rats (Figs 2, ⁴; Table 1).

The area of the GFAP immunoreactive discriminated profiles was increased in the white matter sampled fields of low thoracic and lumbar levels of the 7 day and 3 month lesioned groups (Figs 2, ⁴; Table 1) compared to the sham operated rats, however further significance towards increase was only found at the lesion level of the 3 month group (Figs 2, ⁴; Table 1). Furthermore, increases in the area of the GFAP immunoreactivity were found in the gray matter at low thoracic level 7 days and 3 months after lesion which seemed to be partially counteracted (p = 0,055) with the progression of the time post surgery (^{Fig 4}; Table 1).

Analysis of the two-color immunofluorescence technique for simultaneous detection of bFGF and GFAP immuoreactivity

The analysis of the coexistence of bFGF and GFAP immunoreactivities by means of two-color immunofluorescence technique, employing two different fluorophores, demonstrated the presence of a weak to moderate bFGF immunoreactivity in the nuclei of the vast majority of the GFAP immunoreactive astroglial profiles in all regions of the spinal cord of the sham-operated rats (Fig 5). In the white and gray matters of the 7 days lesioned rats, a moderate to strong bFGF immunoreactivity was observed within the nuclei of the GFAP immunoreactive astroglial profiles which in turn showed increased cytoplasm size and processes (Fig 5). The intensity of the bFGF immunoreactivity increased within the nuclei of the GFAP immunoreactive cells, which also accumulated an increased amount of cytoplasmic GFAP immunoreactivity remarkably close to the lesion site. Furthermore strong bFGF immunoreactivity was observed within the nuclei of reactive astroglial labeled cells only close to the lesion site of the 3 month lesioned rats.

Sections incubated with the bFGF antibody pre-absorbed with human recombinant bFGF showed no immunoreactivity. Furthermore, sections incubated with the solvent of the primary and secondary antibody solutions or the avidin-biotin solution did not show specific labeling (data not shown).

DISCUSSION

Neuronal and non neuronal cells in the vicinity of a spinal cord lesion undergo morphological and functional changes which are related to the local neuroimmune inflammatory events that mediate primarily the normal nervous tissue repair⁸. It is postulated that the size of a secondary neurodegeneration after a spinal cord injury depends on the magnitude of the inflammatory events¹. Astrocytes and microglia deserve special attention regarding their role during and after the inflammatory process in the lesioned spinal cord mainly in promoting the glial scar formation in contrast to their growth promoting abilities.

The present paper has demonstrated substantial increases in the OX42 immunoreactive spMGV/area plotted values which are quantitative morphological evidence of an intense microglial activation at the lesion site and at distant levels induced by a hemitransection at low thoracic level in the short and long term periods of analysis. The results are in agreement with previous findings of a distal and proximal activation of the microglial population, after a photochemically induced spinal cord lesion, however only a qualitative analysis was performed¹². Furthermore, at distant levels from the lesion, the degree of activation found by Koshinaga and Whittemore (1995) was not as evident as in our findings, probably reflecting the different mechanisms of lesion inducing it. Furthermore, previous study employing contusive models has also observed a rostro-caudal distribution of activated glial cells, however, the extent of the described microglial activation was not as prominent as in the transection model used in the present paper¹³.

It is known that reactive microglia undergo proliferation and hypertrophy close to the lesion and in the degenerated neuronal pathways^2,14. A more widespread reaction has been also shown after a mechanic trauma¹⁵, ischemia¹⁶, microbiological infection¹⁷, immunological disorders¹⁸ and axotomy¹⁹.

Microglial cells become ameboid in shape close to the site of a 7 day hemitransection employed in our work that is probably related to phagocytosis which is important in the removal of debris at the wound¹⁴. On the other hand, ameboid microglia was not found either close to injury 3 months after surgery or in other spinal cord levels in the studied periods. In these regions only activated microglia showing increased size of the cytoplasm and process and of levels of OX42 immunoreactivity were observed as previously reported². The spatial widespread and long lasting activation of microglial cells and also the morphological differences of these reactive cells close and far from the injury further suggest other functions of reactive microglia after a CNS lesion. Furthermore, the role of activated microglia in paracrine trophic events in the lesioned spinal cord was already postulated²⁰. It is possible that the microglial reaction seen in the spinal cord, particularly close to the lesion site, exerts growth promoting properties besides its phagocytic effect. The widespread microglial reaction detected in our work by the increase in microglial cell area and spMGV in white and gray matters is in line with these observations since axonal fiber regeneration can be promoted by microglia or their conditioned media⁵.

An intense and long lasting astroglial activation was described in the present paper in the entire spinal cord in response to a spinal cord hemitransection. The results are in line with previous reports that have demonstrated the subsequent astroglial activation after experimental spinal cord or brain lesions, particularly close to the lesion site^6,21,22 as well as accompanying the lesioned pathway²³.

The astrocytic reaction in the 7 day group, detected by the analysis of variance, was substantially intense at the lesion site in both gray and white matters even though it was seen to be less prominent by 3 months after hemitransection. The correlation between the area and spMGV of the GFAP immunoreactivity in the white matter at the lesion site by 3 months indicates that the glial scar is not a static formation but it is a continuous process of hypertrophy in that period post-surgery²⁴. The morphometric and microdensitometric image analyse described at the lesion level following spinal cord hemitransection are in line with previous descriptions regarding the astroglial scar bordering a CNS injury, which consists of tight packed hyperfilamentous astrocytes, with many of their processes apposed to one another with a limited extracellular space, and with many gap and tight junctions⁶. It is well known that the astroglial infiltration at the lesion site contributes to the formation of a physical barrier to fiber growth, furthermore paradoxical effects in respect to secretion of astroglial substances with inhibitory and trophic activities would restrict undesired fiber growth and prevent fibers from degeneration⁶. In normal CNS, a continuous lining of astrocytes processes covered by a basal lamina is interposed between nervous tissue and meninges (external glial membrane) and between nervous tissue and blood vessels (perivascular glial membrane)²². Furthermore the glial scar observed in this experiment can be seen as an attempt of the lesioned CNS to regain its integrity. In spite of the fact that the gliotyc scar that occurs after CNS lesions may interfere with growth of lesioned fibers, it has been postulated that reactive glial cells in the lesioned pathways may favor neuronal plasticity and trophism, possibly by protecting the tissue from secondary degeneration²⁵.

The morphometric and microdensitometric analyse of GFAP used in this work showed the astroglial activation that can took place in the white and gray matters in all studied levels of the spinal cord in regions far from the lesion site. These features suggested the occurrence of an astroglial activation in the entire organ. This paper is in line with the classical description showing morphological and functional changes in the glial cell populations following CNS lesions or manipulations²⁶ and that reactive astrocytes were also observed in pathways not primarily related to injury²⁷, indicating that the signals triggering astroglial reaction get through the site of injury. In that point of view the present paper contributes with a detailed description of the timing of the astroglial and microglial activation in the white and gray matters of the entire lesioned spinal cord near and far from the wound. It has also to be emphasized the similar spatial and temporal pattern of microglial and astroglial activation following the hemitransection performed in the present work which underlines a possible interaction between these two reactive glial cell populations in the trophic events in the lesioned CNS⁸.

It was also found in the present work that reactive astrocytes found in the lesioned site of the spinal cord and in distal and proximal areas to the lesion can produce an increased amount of bFGF, what is in full agreement with previous publications that showed upregulation of that astroglial neurotrophic factor after brain lesions^28,29. Furthermore, our results are in line with the paper of Koshinaga and Whitemore that showed changes in the intracellular bFGF in the reactive astrocytes located around the lesion area and in Wallerian degenerating fiber tracts¹².

It has been extensively described the potent neurotrophic actions of bFGF on neuronal populations in normal and lesioned brains^28,30. Trophic actions of bFGF have been observed following neuronal cell body lesions as well as fiber tract injury²⁸.

It seems possible that the increased synthesis of bFGF by reactive astrocytes in the white and gray matters of lesioned spinal cord may favor the maintenance of projecting neurons and a long term regeneration of the entire organ.

Thus, the present analysis speculates that the activation of astrocytes triggers via bFGF, paracrine trophic actions that may be related to wounding and repair events at the lesion site. Far from injury astroglial activation might help maintain gray matter neurons which may favor fibers outgrowth once adequate repair is performed. A paracrine astroglial trophic event may have taken place in the entire spinal cord white matter, a phenomenon that also lasts for a long period after the lesion, which indicates a continuous need for a trophic support by the lesioned and also unlesioned fiber. These observations are also in line with the finding that administration of neurotrophic factors increases fiber regeneration in the lesioned spinal cord of the rat³¹.

Acknowledgements - We are grateful to Miss Patrícia R. Campos for technical assistance and to Mrs. Cleide Rosana Prisco for her expertise in the statistical analysis.

Received 12 February 2001, received in final form 25 April 2001. Accepted 2 May 2001.

Dr. Gerson Chadi - Departamento de Anatomia, Universidade de São Paulo - Av. Prof. Lineu Prestes, 2415 - 05508-900 São Paulo SP ¾ Brasil. FAX: [phone]. E-mail: [e-mail]

• 1. Davies SJ, Field PM, Raisman G. Regeneration of cut adult axons fails even in the presence of continuous aligned glial pathways. Exp Neurol 1996;142:203-216.
• 2. Dusart I, Schwab ME. Secondary cell death and the inflammatory reaction after dorsal hemisection of the rat spinal cord. Eur J Neurosci 1994;6:712-724.
• 3. Stichel CC, Muller HW. The CNS lesion scar: new vistas on an old regeneration barrier. Cell Tissue Res 1998;294:1-9.
• 4. Schmitt AB, Buss A, Breuer S, et al. Major histocompatibility complex class II expression by activated microglia caudal to lesions of descending tracts in the human spinal cord is not associated with a T cell response. Acta Neuropathol (Berl) 2000;100:528-536.
• 5. Elkabes S, DiCicco-Bloom EM, Black IB. Brain microglia/macrophages express neurotrophins that selectively regulate microglial proliferation and function. J Neurosci 1996;16:2508-2521.
• 6. Fawcett JW, Asher RA. The glial scar and central nervous system repair. Brain Res Bull 1999;49:377-391.
• 7. Gomide VC, Chadi G. The trophic factor S-100b and basic fibroblast growth factor are increased in the forebrain reactive astrocytes of adult callosotomized rat. Brain Res 1999;835:162-174.
• 8. Giulian D, Vaca K, Corpuz M. Brain glia release factors with opposing actions upon neuronal survival. J Neurosci 1993;13:29-37.
• 9. Paxinos G, Watson C. The rat brain: in stereotaxic coordinates. San Diego: Hartcourt Brace Jovanovich, 1986.
• 10. Chadi G, Moller A, Rosén L, et al. Protective actions of human recombinant basic fibroblast growth factor on MPTP-lesioned nigrostriatal dopamine neurons after intraventricular infusions. Exp Brain Res 1993;97:145-158.
• 11. Chadi G, Rosen L, Cintra A, et al. Corticosterone increases FGF-2 (bFGF) immunoreactivity in the substantia nigra of the rat. Neuroreport 1993;4:783-786.
• 12. Koshinaga M, Whittemore SR. The temporal and spatial activation of microglia fiber tracts undergoing anterograde and retrograde degeneration following spinal cord lesion. J Neurotrauma 1995;12:209-222.
• 13. Carlson SL, Parrish ME, Springer JE, Doty K, Dosset L. Acute inflammatory response in spinal cord following impact injury. Exp Neurol 1998;151:77-88.
• 14. Giulian D. Microglia and neuronal disfunction. In Kettenmann H, Ransom BR (eds.). Neuroglia. London: Oxford Univers Press, 1995:671-684.
• 15. Streit WJ, Graeber MB, Kretzberg GW. Functional plasticity of microglia: a review. Glia 1988;1:301-307.
• 16. Jorgensen MB, Finsen BR, Jensen MB, Castellano B, Diemer NH, Zimmer J. Microglial and astroglial reactions to ischemic and kainic acid-induced lesions of the adult rat hippocampus. Exp Neurol 1993;120:70-88.
• 17. Matsumoto Y, Ohmori K, Fujiwara M. Microglial and astroglial reactions to inflammatory lesions of experimental autoimmune encephalomyelitis in the rat central nervous system. J Neuroimmunol 1992;37:23-33.
• 18. Matsumoto Y, Fujiwara M. Absence of donor-type major histocompatibility complex class I antigen-bearing microglia in the rat central nervous system of radiation bone marrow chimeras. J Neuroimmunol 1987;17:71-82.
• 19. Graeber MB, Streit WJ, Kreutzberg GW. The microglial cytoskeleton: vimentin is localized within activated cells in situ. J Neurocytol 1988;17:573-580.
• 20. Zachary JF, Baszler TV, French RA, Kelley KW. Mouse Moloney leukemia virus infects microglia but not neurons even though it induces motor neuron disease. Mol Psychiatry 1997;2:104-106.
• 21. Fitch MT, Doller C, Combs CK, Landreth GE, Silver J. Cellular and molecular mechanisms of glial scarring and progressive cavitation: in vivo and in vitro analysis of inflammation-induced secondary injury after CNS trauma. J Neurosci 1999;19:8182-8198.
• 22. Bignami A, Dahl D. Gliosis. In Kettenmann H, Ransom BR (eds.). Neuroglia. London: Oxford Univers Press, 1995;843-858.
• 23. Smith GM, Miller RH, Silver J. Changing role of forebrain astrocytes during development, regenerative failure, and induced regeneration upon transplantation. J Comp Neurol 1986;251:23-43.
• 24. Barret CP, Donati EJ, Guth L. Differences between adult and neonatal rats in their astroglial response to spinal injury. Exp Neurol 1984;84:374-385.
• 25. Moonen G, Rogister B, Leprince P, et al. Neurono-glial interactions and neural plasticity. Prog Brain Res 1990;86:63-73.
• 26. Shao Y, McCarthy KD. Plasticity of astrocytes. Glia 1994;11:147-155.
• 27. Riva MA, Donati E, Tascedda F, Zolli M, Racagni G. Short and long term induction of bFGF gene expression in rat central nervous system following kainate injection. Neuroscience 1994;59:55-65.
• 28. Chadi G, Cao Y, Petterson RF, Fuxe K. Temporal and spatial increase of astroglial bFGF synthesis after 6-hydroxydopamine-induced degeneration of the nigrostriatal dopamine neurons. Neuroscience 1994;61:891-910.
• 29. Humpel C, Chadi G, Lippoldt A, Ganten D, Fuxe K, Olson L. Increase of basic fibroblast growth factor (bFGF, FGF-2) messenger RNA and protein following implantation of a microdialysis probe into rat hippocampus. Exp Brain Res 1994;98:229-237.
• 30. Chadi G, Moller A, Rosen L, et al. Protective actions of human recombinant basic fibroblast growth factor on MPTP-lesioned nigrostriatal dopamine neurons after intraventricular infusion. Exp Brain Res 1993;97:145-158.
• 31. Cheng H, Cao Y, Olson L. Spinal cord repair in adult paraplegic rats: partial restoration of hind limb function. Science 1996;273:510-513.

¹

Laboratory of Neuroregeneration, Department of Anatomy, Institute of Biomedical Sciences, University of São Paulo, São Paulo SP, Brazil.

http://www.icb.usp.br/~neuron. This work was supported by grants from FAPESP (98/13122-5; 99/01319-1; 99/1308-0).

Publication Dates

• Publication in this collection
  28 Sept 2001
• Date of issue
  Sept 2001

History

• Accepted
  02 May 2001
• Received
  12 Feb 2001
• Reviewed
  25 Apr 2001