Diversity among reactive astrocytes: Proximal reactive astrocytes in lacerated spinal cord preferentially react with monoclonal antibody J1–31

Brain Research B&W

Vol. 30, pp. 395-404, Printed in the USA. All rights reserved.

1993 Copyright

036 I -9230/93 $6.00 t .OO 0 1993 Pergamon Press Ltd.

Diversity Among Reactive Astrocytes: Proximal Reactive Astroc~es in Lacerated Spinal Cord Preferentially React With Monoclonal Antibody 51-3 1 S. K. MALHOTRA,“]

M. SVENSSONJ H. ALDSKOGIUS,t AND T. K. SHNITKAg

R. BHATNAGAR,$

G. D. DASS

*Department ofZoology, University of Alberta, Edmonton, Alberta, Canada T6G 2E9 ;f-Department of Anatomy, Karolinska Institutet, Stockholm, Sweden $Department of Biological Sciences, Purdue University, West LafaJpette, IN ~Department of pathology, University ~fA~berta, Edmont~~n, Alberta, Canada, T6G 237 MALHOTRA, S. K., M. SVENSSON, H. ALDSKOGIUS, R. BHATNAGAR, G. D. DAS AND T. K. SHNITKA. Diversity among reactive astrocytes: Proximal reactive astrocytes in lacerated spinal cord pr&rentiailv react with monoclonal antibody JI31. BRAIN RES BULL 30(3/4) 395-404, 1993.-An Astrocyte-specific antigen recognized by monoclonal antibody J l-3 I is a more intense marker for proximal reactive astrocytes in lacerated rat spinal cord than is glial hbrillary acidic protein (GFAP). Thus, MAb 51-31 recognizes reactive astrocytes in the immediate vicinity of the lesion, whereas reactive astrocytes that are located at a distance from the lesion are not detected by immunofluorescent staining. These findings are relevant to the biochemical heterogeneity manifested respectively by reactive astrocytes located proximal and distal to a laceration-type injury of the spinal cord. and those that develop following axotomy with retrograde degeneration. Reactive astrocytes in the axotomy model are not stained with MAb J l-3 I, but are positive for GFAP. Lacerated spinal cord Immunocytochemistry

Hypoglossal nerve axotomy

Proximal and distal reactive astrocytes

REACTIVE astrocytes arise either from normal astrocytes which undergo hy~~rophy (37) or from astrocytic precursor cells which undergo hype~lasia followed by hy~~rophy (13,2 1). They manifest morphological and metabolic heterogeneity in the central nervous system (CNS), apparently due to the existence of different microenvironmental conditions in different lesions, and on these criteria, Duchesne et al. (12) recognized several subtypes of reactive astrocytes. Proximal reactive astrocytes and distal reactive astrocytes are the most common. These, respectively, represent reactive astrocytes in the immediate vicinity of the lesion and those which are further removed from the lesion. An increase in activity of succinate, glucose &-phosphate, lactate, and particularly of glutamate dehydrogenases can be seen in proximal but not in distal astrocytes (12). A third subtype of reactive astrocyte is the active astrocyte found only in human or experimentally induced epileptogenic lesions of the brain ( 12,39). Recently, several investigators also have drawn attention to differences between the reactive astrocytes that develop in relation to a laceration-type injury as opposed to the reactive astrocytes that develop as a consequence of axotomy (13,18,38).

’ To whom requests for reprints should be addressed.

395

Mab 31-31

Thus, while there is recognized diversity among reactive astrocytes in different categories of lesions in the CNS, the causes of such biochemical and functional heterogeneity are not yet known. It is conceivable that differences in various types of reactive astrocytes represent underlying differences in the normal astrocytes from which the reactive astrocytes arise. The more likely possibility is that different signals (factors) may be produced in specific categories of lesions that induce reactive astrocytes, e.g., laceration versus axotomy, as there is no direct traumatization of the CNS in axotomy, whereas laceration produces local trauma to the CNS and a release into the damaged tissue of blot-delved factors which may influence the development of the astrocytic reaction. Therefore, the development of reactive astrocytes may differ in different lesions of the CNS and could be influenced by differences in as yet unknown signals operative in triggering the formation of reactive astrocytes. The present article is a report of an astrocyte-specific antigen recognized by monoclonal antibody (MAb) J l-3 1 that is an intense marker of reactive astrocytes in the lacerated rat spinal cord in the region of the wound.

396

MAI. tioR’1 4 I. I’ (‘HAKAC-I

ERISTK‘S

OF MAh .I I-i I

MAb J l-31 (isotype IgGZb) (36) was produced by injecting Balh C female mice with homogeneized multiple sclerosis (MS) plaque tissue collected at autopsy and kept frozen (23). The reactive antigen for MAb J l-3 I. as determined by immunoprecipitation followed by SDS-PAGE (run under reducing conditions of gel clectrophoresis~, bands at a molecular weight ofapproximately 30 kDa (36). By imInun~~fluoresccnce microscopy the staining for MAb J 1-3 I overlapped with GFAP-positive cells (i.e., astrocytes and Miiller glia) in human and several other mammalian species (9-l I .32.34). By immunoelectron microscopy, MAb J I-3 I localization appeared in association with intermediate filaments (IF Fig. IA. B) characteristic of astrocytes. namely, GFAP-type filaments [Fig. 7A, B: (33)]. However. the antigen rccognizcd by MAb J 1-3 1 is distinct from GFAP, and this conclusion is based upon various cxperim~nts. c.g.. differential expression during development of the rat CNS (32). lack of inhibition of staining with MAh J l-3 I by prior incubation in antiserum to GFAP (24). and the expression of J l-3 I antigen in vitro in rat glioma cells (9L and c’6) which are apparently GFAP negative under our conditions of culture (15.25). Also, 51-31 antigen is distinct from vimentin. another IF protein known to be expressed in a wide variety of cells including glial cells (6.7.1 I). lnlmuno~uorescence staining for MAb J 1-3 1 has not been detected in non-CNS ceils. Thus. M.&b J l-3 I appears to preferentially immunostain astrocytes. However, in cryostat sections of human MS spinal cord. staining due to MAb J I-3 I was conspicuous in contrast to normal spinal cord. where httle or no staining could be detected. Therefore. it appeared that the enhanced staining in MS spinal cord could be due to the presence of reactive astrocvtes in MS plaques. A study on MS brain (autopsy samples) conhrmed the foregoing assumption. and showed enhanced staining with MAb J l-3 I in reactive astroeytcs in plaque regions in contrast to astroglia located in the adjacent apparently normal brain tissue (24). Due to the limited available supply of human CNS tissue. in a state of preservation suitable for immunocytochemical studies. a laceration-type lesion of the rat spinal cord has been used as a model system for the induction of reactive astrocytes (8.3 I ). L4CERATED

RAT SPINL

CORD MODEL

Surgically operated spinal cords (lumbar level L3-L4) of Long-Evans hooded rats (3-6 months old) were used as the source of reactive astrocytes. The details of the surgical procedure and animal care have been published previously (8,30,3 I). The surgical lesions extended for one segment in the longitudinal plane and through 50% of the transverse plane. These were shallow lesions and did not produce any paraplegic syndrome. There was some transient autonomic dysfunctions, however (8). Lesioned spinal cords and sham-operated spinal cords were collected at 4 days. 2 weeks. 3 months, and 6 months following surgery. Etherized rats were perfused with 4% paraformaldehyde in 0.2 M phosphate buffer, pH 7.2. and L3 and L4 segments of the spinal cord from lesioned and sham-operated rats were collected. These were further fixed in paraformaldehyde (24-48 h), washed in phosphate buffered saline (PBS). cryoprotected in 30% sucrose for the prepa~tion of cryostat sections. Approximately IO pm thick sections were mounted on coverglasses and stained for indirect immunotluorescence. Mouse MAb 3 1-3 1 and rabbit anticow GFAP antiserum were used as primary antibodies and sections were double-labelled with appropriate fluorochrome-

conjugated

Normal mouse serum (NMS) in place of primary antibodies. The sections were mounted in buffered glycerol containing p-phenylencdiamine (19) and examined by, cpifluorescence microscopy. The method for immunofluorcsccnce staining is given in Predy et al. (31) and Malhotra et al. (24). ln~munolabelling for electron microscopy was carried out using sections ofparaformaldehyde-fixed tissue cut on an Oxford Vibratome. The spinal cord tissues (fixed by perfusion followed by immersion) were sectioned at 20 to 30 pm and incubated in MAb J l-3 I ascites fluid ( 1500 dilution) and rabbit anticou GF,4P antiserum (Ddkopatts, I: 1000 dilution) overnight at 4°C’. The samples then were washed in PBS. and incubated with appropriate colloidal gold-c(~njugated (5 nm} secondary antibodies. NMS and NRS were used as controls in place of primary antibodies. The tissue was further hxed with X gluatar~~ldeh~dc followed by 2’5 OsOd prior to dehydration and embedding in Epon. Thin sections were examined in the electron microscope with or without staining with heavy metal salts. Immunoelectron microscopy using MAb J l-3 I also was carried out on normal rat cerebellum as a control for the localization of antibody binding sites. and normal

secondary

41

antibodies.

rabbit serum (NRS)

were used as controls

In essence. overlapping immunostaining for .I l-3 1 antigen and GFAP is discerned in reactive astrocytes in the region of the experimental lesion (Fig. 3A. B). However in the adjacent nonlesion region of the spinal cord, staining for J 1-3 I antigen is weak or nonexistent (Fig. 4A and Fig 5A), when compared with that seen after incubation with normal mouse serum as a control (Fig. 4B and Fig. 5B). In contrast. staining for GFAP jn readily detectable in astrocytes in the nonlesion region (Fig. 4C and Fig. XI’), examined in sections double labelied with the two antibodies. Thus. in the laceration-type experimental lesion. MAb J l-3 1 recognizes reactive astrocytes essentially in the vicinity of the lesion. whereas astrocytes (reactive or normal) that are distant from the lesion are not detected with MAb 51-31: however. positive staining with GFAP-antiserum is detectable in astroglia in such distant regions. The spread ofgliosis from hemisections of the rat spinal cord can occur in continuity, and/ or due to degeneration of ascending and descending fiber tracts (3.4) so that reactive astrocytes are seen at a distance of several spinal segments from the lesion, on the ipsilateral but not the contralateral side (3). The first appearance of clearly detectable staining with MAb J I-3 I can be observed as early as 4 days (the shortest time period examined) after spinal cord laceration. Thereafter, the intensity of staining is increased and optimum staining could be detected in specimens examined 2 weeks postsurgery. This staining persists for several months. and an intense reaction was detected even in two specimens collected 6 months after surgery. A semiquantitative interpretation by digital analysis of the immunofluorescence data obtained from specimens collected 16 days following surgery has been given by Predy et al. (31). The foregoing showed that in the region of the lesion, the fluorescence signal due to J I-3 I antigen was 20 times greater than in the nonlesion region, whereas the comparable signal for GFAP was threefold greater. Thus, J i-3 I antigen is a more intense marker of reactiveness of astrocytes than GFAP in lacerated rat spinal cord. and distinguishes proximal from distal reactive astrocytes. As mentioned previously. MAb Jl-31 also immunostains astrocytes in MS plaque regions more intensely than astroglia in the adjacent apparently normal white matter (24).

DIVERSITY

AMONG

REACTIVE

ASTROCYTES

FIG. I. lmmunogold label (5 nm) to localize J 1-3 I antigen (in rat cerebellum). The gold particles are concentrated in association with filaments (A) and are lacking in sections from cerebellum incubated with normal mouse serum as a control for MAb 51-3 I (B). Arrows indicate filaments. Calibration bar = 200 nm.

39x

MAI

ItORI

4

I 1

FIG. 2. lmmunogold label (5 nm) to localize gtial tibrillary acidic protein (in rat spinal cord) which is associated with filaments (A). B is a control incubated with normal rabbit serum instead of GFAP antiserum. Arrows indicate clustered filaments. Calibration bar A = 200 nm. B 7 350 nm.

DIVERSITY

AMONG REACTIVE

399

ASTROCYTES

FIG. 3. Double-labelled immunofluorescence preparation to show overlapping staining (arrows) with MAb J l-3 I (A) and GFAP antiserum (B) in reactive astrocytes in the region of the lesion in lacerated rat spinal cord (white matter). Calibration bar = 15 pm. COMPARISONOF LACERATION-INJURYWITH AXOTOMY MODEL

Femafe, Sprague-Dawley rats (about 3 months old, 200-250 g b.wt.) were anesthetized with chloral hydrate (35 mg/lOO g b.wt.) and subjected to transection of the right hypoglossal nerve where it passes the carotid artery. Following 4, 7, 14, or 42 days postoperative survival, the animals were reanesthetized and perfused via the left ventricle with saline (body temperature) followed by ice-cold phosphate buffered (0.15 M) 4% paraformaldehyde (w/v) and 14% saturated picric acid (v/v). The lower brain stem, including the level of the hypoglo~al nuclei, was removed and postfixed for 2-4 h. Tissue was stored at about 4°C in 0.15 M phosphate buffer containing 15-30% sucrose. Fourteen Frn frontal sections were cut on a cryostat and mounted on gelatinized slides. These were incubated in 1.5%normal horse serum and 0.3% Triton-X in phosphate buffer (0.15 M, 1 h, room temperature). This buffer was used in all of the following steps except during substrate incubations. Sections were incubated with monocfonal antibodies to glial fibrillary acidic protein (GFAP, Labsystems, 1: 1~0) or J l-3 1 (1:500) for 24 h at 4°C. The immunoreaction was visualized with indirect immunocytochemistry using the avidin-biotin method for fluorescent secondary antibodies. Some sections were incubated with biotinylated horse antimouse antibodies (Vector, 1:200, 1 h), washed 2 X 10 min, and incubated with Avidin-Biotin-Complex (ABCElite, Vector, 150, 1 h). After washing, peroxidase activity was demonstrated with diaminobenzidine in Tris-HCl-buffer (50 mg/ 100 ml, pH 7.4), containing 0.02% hydrogen peroxidase as substrate. The reaction product was intensified with nickel (1%) and cobalt [ 1%;cf., (I)]. After a final wash, the sections were dehydrated in alcohols to xylene and mounted in a nonaqueous medium (Diatexm). Other sections from the peripheral nerve injury

group were incubated with rhodamine-conjugated rabbit antimouse secondary antibodies (Dako, 1:20, 1 h). These slides were dehydrated and mounted in ~uoromont (Curt-@). Fluorescent material was examined with a Nikon microscope using epifluorescence. Control preparations were made without the incubations with primary antibodies. Results

The J 1-3 1 antibody labelled processes of nonneuronal cells surrounding the area postrema (Fig. 6A) as well as subpial astrocytes (Fig. 6B) in the brain stem. However, no immunoreactivity for J 1-3 1 was observed in the hypoglossal nucleus at any of the examined postoperative survival times (Fig. 6C and 6D). In contrast, GFAP-IR was present in this nucleus on both sides, but at markedly greater intensity ipsilateral to nerve transection (Fig. 6E and 6F). Thus, the important message of this paper is that MAb J l-3 1, in concert with GFAP-antibodies, demonstrates heterogeneity among GFAP-positive astrocytes, not only in relation to their geographic location to a surgical lesion, but also in different types of model systems. DISCUSSION

The differences in reaction with MAb Jl-3 1 exhibited by reactive astrocytes in the laceration-type lesion versus axotomy could conceivably be due to different signals or factors operative in the two model systems. Thus, the laceration-type injury results in a local destruction of neural and glial cells and their c,onnections. disruption of blood vessels, and an escape of the fluid and cellular components ofblood into the damaged tissue (27). This complex environment contains platelet-delved growth factor (PDGF). biologically active protein breakdown products (20) myelin breakdown products (5). and fibrin split products, which

FIG. 4. Double-labelled immunofluoresccnce preparation to show weak stainmg with MAh .I I-3 I (A) in comparison to intense stainmg with GFAP-antiserum (C) in astrocytes in the nonlesion region ofthe lacerated rat spinal cord. Arrows indicate similar locations in the two photographs. Sections B and D show double-labeiled preparations incubated with normal mouse serum (B) and normal rahhit serum (D) as respective controls 15 pm. for MAb J l--3 I and GFAP antiserum. Calibration hat

are capable of stimulating the proliferation of astrocytes. Soon after the injury, a healing response begins, which includes the production of autocrine and paracrine growth and trophic factors by activated astrocytes. microglia. and endothelial ceib

For example, within the first hour following cortical brain injury in rats, the amount ofgiia maturation factor{~MF) and acidic fibroblast growth factor (aFGF) in the wound increased by 7- and I 3-fold, respectively, compared to tissue ad-

(X,28.29}.

DIVERSITY

AMONG

REACTIVE

ASTRWYTES

FIG. 5. Double-labelled immunofluorescence preparation to show weak staining with MAb J 1-3 I (A) and intense staining with GFAP antiserum (C) in astrocytes in grey matter distant from the lesion site. Arrows indicate similar locations for comparison of staining with the two antibodies. B and D are normal sera controls for MAb J l-3 1 (B) and GFAP antiserum (D). Calibration bar = I5 pm.

jacent to the wound (29). Other growth factors that are involved

in the cascade of biochemical events leading to the production of reactive astrocytes likely include interleukin-1 (17) and tumor necrosis factor (14.40). In early lesions of multiple sclerosis, there

is also demyetination, necrosis, local disruption of the bloodbrain barrier, and activation of microglia (223); Merrill (26) has reviewed the role of growth factors and lymphokines in the evolution of the MS lesion.

FIG. 6. (A-D) J l-31 inlmunolabelling of the brain stem. I week after transection of the right surround the area postrema and extend towards the central canal (A). Suhpial astrocytes aid (B). However. no specific immunoreactivity is observed in the hvponlossal nucleus (C. D). .. _ immunolabelling in the hypoglossal nucleus I week after transection of the right hypoglossal noreactivity on the side of the operation (E, F). Bar 50 pm.

402

hypoglossal ncrvc. Immunoreactix processes their processes are likewise positive for J l-3 I Bar I()() tim (A). 50 urn (B-D). (E-F) GFAP nerve. Note the increased intensity of immu-

DIVERSITY

AMONG

REACTIVE

403

ASTROCYTES

On the other hand, axotomy induces a remote retrograde astrocytic response in the region of the affected motor nucleus in brain (18,38) or in spinal cord (16). The vasculature remains intact and blood components do not enter the tissue. Cellular necrosis is minimal. Mediators, not yet identified, probably are released from damaged neurons (38). Damaged neurons produce interferon-y, which could activate microglia in the region; the latter, in turn, release interleukin-1 and other lymphokines (2,26,35). Indeed, there is a beginning awareness that the reactive astrocytic response which follows a cerebral laceration or stab wound is essentially different from the remote astrocytic response which follows axotomy and retrograde degeneration (13. I8,38). The findings reported herein, concerning differences in the coexpression and en-

hancement of expression of GFAP and J l-3 1 marker proteins in reactive astrocytes respectively situated at the site of a laceration-type injury, or at a distance from such a lesion, or developing remotely following axotomy and retrograde degeneration, provide further objective support for the foregoing interpretation, and demonstrate the usefulness of MAb Jl3 1 in defining a subpopulation of reactive astrocytes. ACKNOWLEDGEMENTS

We are grateful to Mr. R. Predy for carrying out immunofluorescence staining on lacerated rat spinal cords, and to Mrs. K. Price for invaluable help in the preparation of the manuscript. This research work has been funded by grants awarded by the NSERC of Canada (OGP5021) and N.I.H. (Grant No. NS-08817).

REFERENCES 1. Adams, J. C. Heavy metal intensification of DAB-based reaction

product. J. Histochem. Cytochem. 29:775; 198 1. 2. Aldskogius. H.; Svensson, M. Neuronal and glial cell responses to axon injury. In: Malhotra. S. K, ed. Advances in structural biology, vol. 2. Greenwich, CT: Jai Press (in press). 3. Barrett. C. P.; Guth, L.; Donati. E. J.: Krikorian, J. G. Astroghal reaction in the gray matter of lumbar segments after midthoracic transection of the adult rat spinal cord. Exp. Neurol. 73:365-377; 1981. 4. Barrett, C. P.: Donati, E. J.; Guth, L. Differences between adult and neonatal rats in their astroglial response to spinal injury. Exp. Neurol. 841374-385: 1984. 5. Bologa, L. Myelin basic protein stimulates the proliferation of astrocytes: Possible explanation for multiple sclerosis plaque formation. Brain Res. 346: 199-203: 1985. 6. Calvo. J. L.: Carbonell, A. L.; Boya, J. Coexpression of vimentin and glial fibrillary acidic protein in astrocytes of the adult rat optic nerve. Brain Res. 532:355-357; 1990. 7. Cardone, B.; Roots. B. I. Comparative immunohistochemical study of filament proteins (glial fibrillary acidic protein) and vimentin in goldfish, octopus, and snail. Glia 3: 180- 192; 1990. 8. Das. G. D. Neural transplantation in spinal cord under differing conditions of lesions and their functional significance. In: Das, G. D.; Wallace, R. G., eds. Neural transplantation and regeneration. Proceedings in life sciences. New York: Springer-Verlag; 1986:l61. 9. de Vente. J.: Wietsma, J. J.: Bol. J. G. J. M.: Schipper, J.; Malhotra, S. K.; Steinbusch, H. W. M. Potassium increases cGMP in glial cells in rat hippocampus slices: A combined biochemical and immunocytochemical study. Neurosci. Res. Commun. 4:25-3 1; 1989. 10. Diefenbach, T. J.; Elbrink, J.; Malhotra, S. K. Characterization of intermediate filament proteins in astroglia of hamster cerebellum, Cytobios 59: 193-206: 1989. II. Diefenbach, T. J.; Elbrink, J.; Malhotra, S. K. Glial fibrillary acidic protein, 51-31 antigen and vimentin in adult hamster brain: An immunohistochmeical study. Cytobios 65:39-53; 1991. 12. Duchesne, P. Y.; Gerebtzoff, M. A.: Brotchi, J. Four types of reactive astrocytes. Bibl. Anat. 19:313-3 16: 1981, 13. Eng, L. F. Regulation of glial intermediate filaments in astrogliosis. In: Norenberg, M. D.; Hertz, L.; Schousboe, A., eds. The biochemical pathology of astrocytes. New York: Alan R. Liss; 1988:79-90. 14. Gadient, R. A.; Cron, K. C.; Otten, U. Interleukin-1 p and tumor necrosis factor-a synergistically stimulate nerve growth factor (NGF) release from cultured rat astrocytes. Neurosci. Lett. Il7:335-340; 1990. 15. Ghahary, A.: Bhatnagar, R.; Price, K.; Forsyth, N. L.; Shen, Y. J.; Tredget, T. E.; Malhotra, S. K. Rat glioma cell lines C6 and 9L synthesize type 1 collagen in vitro. Brain Res. Bull. 28:47-56; 199 1. 16. Gilmore, S. A.: Sims, T. J.; Leiting, J. E. Astrocvtic reactions in spinal gray matter following sciatic axotomy. Glia j:342-349; 1990. 17. Giulian, D.; Lackman, L. D. Interleukin-1 stimulation of astroglial proliferation after brain injury. Science 228:497-499; 1985.

18. Haj6s. F.; Kalman. M.; Zilles, K.; Schleicher, A.; Sotonyi, P. Remote astrocytic response as demonstrated by glial librillary acidic protein immunohistochemistry in the visual cortex of dorsal lateral geniculate nucleus lesioned rats. Glia 3:30 l-3 10; 1990. 19. Johnson, G. D.; de C Noguerie Aranjo. G. M. A simple method of reducing the fading of immunofluorescence during microscopy. J. Immunol. Methods 43:349-350; 1981. 20. Klatzo, I. Neuropathological aspects ofbrain oedema. J. Neuropathol. Exp. Neurol. 26:1-14: 1967. 21. Latov, N.; Nilaver, G.; Zimmerman, E. A.; Johnson, W. G.; Silverman, A.-J.: Defendini. R.; Cote, L. Fibrillary astrocytes proliferate in response to brain injury. Dev. Biol. 72:381-384; 1979. 22. Lumsden. C. E. The neuropathology of multiple sclerosis. In: Vinken, P. J.; Bruyn, G. W., eds. Handbook of clinical neurology. vol. 9; Multiple sclerosis and other demyelinating diseases. New York: American Elsevier: 1970: 2 17-309. 23. Malhotra, S. K.: Wong, F.; Cumming, P.; Ross, S. D.; Shnitka, T. K.; Manickavel, V.; Warren, K. G.; Jeffrey, V. A monoclonal antibody for cytoskeletal antigenic determinant(s) distinguishable from glial fibrillary acidic protein in astrocytes. Microbios. Lett. 26: 151-157; 1984. 24. Malhotra, S. K.: Predy, R.; Johnson, E. S.; Singh, R.; Leeuw, K. Novel astrocytic protein in multiple sclerosis plaques. J. Neurosci. Res. 22:36-49; 1989. 25. Malhotra, S. K.; Singh, R.; Bhatnagar, R.; Price, K.; Spencer; Feinstein. D. Selective expression of 51-31 antigen in glioma 9L cells. 21st Annual Meeting, American Society of Neurochemistry, March 1990. Abstract. 26. Merrill, J. E. Macroglia: Neural cells responsive to lymphokines and growth factors. Immunol. Today 8:146-150; 1987. 27. Nieto-Sampedro, M.: Cotman, C. W. Growth factor induction and temporal order in central nervous system repair. In: Cotman, C. W., ed. Synaptic plasticity. New York: Guilford; 1985:407-455. 28. Nieto-Sampedro. M.; Saneto, R. P.; de Vellis, J.; Cotman, C. W. The control of glial populations in brain: Changes in astrocyte mitogenic and morphogenic factors in response to injury. Brain Res. 343:320-328: 1985. 29. Nieto-Sampedro. M.; Lim, R.: Hicklin, D. J.; Cotman, C. W. Early release of glia maturation factor and acidic fibroblast growth factor after rat brain injury. Neurosci. Lett. 83:361-362: 1988. 30 Predy, R.: Malhotra, S. K. Reactive astrocytes in lesioned rat spinal cord: Effect of neural transplants. Brain Res. Bull. 22:8 l-87; i989 31 Predy. R.; Malhotra. S. K.; Das. G. D. Enhanced exuression of a protein antigen (J l-3 I antigen, 30 kilodaltons) by reactive astrocytes in lacerated spinal cord. J. Neurosci. Res. 19:397-404; 1988. 32 Predy, R.: Singh, D.; Bhatnagar, R.; Singh, R.; Malhotra, S. K. A new protein (J l-3 I antigen, 30 kD) is expressed by astrocytes, Miller glia and ependyma. Biosci. Rep. 7:49 l-502; 1987. 33. Schachner, M.: Hedley-Whyte, E. T.; Hsu, D. W.; Schoonmaker, G.; Bignami, A. Ultrastructrual localization of glial tibrillary acidic protein in mouse cerebellum by immunoperoxidase labeling. J. Cell Biol. 75167-73: 1977.

404

34. Schriider, H.: Malhotra. S. K. Characterization of rodent pineal astrocytes by immunofluorescence microscopy using a monoclonal antibody (51-31). Cell Tissue Res. 248:607-610: 1987. 35. SelmaJ, K. N.: Faroog, M.; Norton, T.: Raine. C. S.: Brosman. C. F. Proliferation ofastrocytes in vitro in response to cytokines. J. Immunol. 144: I29- 135; 1990. 36. Singh. R.; Singh, B.: Malhotra. S. K. A new marker protein for astrocytes. Biosci. Rep. 6:73-80: 1986. 37. Takamiya. Y.: Kohsaka. S.: Toya, S.: Otani, M.: .Tsukada, Y. Immunohistochemical studies on the proliferation of reactive astrocytes and the expression of cytoskeletal proteins following brain injury in rats. Dev. Brain Res. 38:201-210: 1988.

MAI.liOKI

3x

12 I,

I 41

I ctrlalf. W.: Graeber. M. B.: l&by. M. A.: Kreut/berg. 0. W. Incrcased glial fibrillary acidic protein synthesis in astrocytes during retrograde reaction of the rat facial nucleus. Glia I :90-95: 19x8. 39. Tiffany-Castiglioni. E.: C‘astiglioni. A. J.. Jr. Astr-ocytcs in epilepsy. In: Fedoroff, S.: Vernadakis. A.. eds. t\strocytes. vol. 3; Cell biology and pathology of astrocytes. Yew York: Academrc Press: 19X6:40 I 424. 40. Vilcek. J.; Lee. T. H. 1.umor necrosis factor. Nevv Insights Into the molecular mechanrsms of its multiple actions. J. Hiol. Chem. 266: 7313-7316; 1991.