- Email: [email protected]
Astrocytes secrete basal lamina after hemisection of rat spinal cord
Brain Research, 327 (1985) 135-141 Elsevier
135
BRE 10524
Astrocytes Secrete Basal Lamina After Hemisection of Rat Spinal Cord JERALD J. BERNSTEIN, REINA GETZ, MARY JEFFERSON and MURIEL KELEMEN
Veterans Administration Medical Center, Washington, DC 20422 and Department of Physiology and Neurosurgery, George Washington University School of Medicine, Washington, DC (U.S.A.) (Accepted May 15th, 1984)
Key words: regeneration - - basal lamina - - ultrastructure - - spinal cord - - injury - - histochemistry
Basal lamina is reconstructed over the lesioned surface of the spinal cord. The following experiment (90 rats) studies the ultrastructure of the formation of this membrane and the immunohistochemistry of laminin production (a major secreted component of basal lamina). After hemisection of the spinal cord at T6 animals were prepared for electron microscopy or antilaminin-biotin-avidin-peroxidase incubation. Three-5 days posthemisection, antilaminin reaction product was observed in astrocytes and their processes which faced the lesion, endothelia of blood vessels or pia. Ultrastructurally (3 days), basal lamina was polymerizing as small projections on the surface of astrocytic membranes facing the lesion, endothelia or pia. By 5 days the basal lamina was a single membrane, folded multiple sheets or in swirls. At 6-10 days the antilaminin reaction and the basal lamina (except for duplications) did not differ from normal. Reactive astrocytes secrete laminin for at least 3-5 days after hemisection and form basal lamina on the lesioned surface of the spinal cord after spinal cord hemisection.
INTRODUCTION The cellular responses of the spinal cord to lesion are varied and differ with hierarchy throughout the animal kingdom2,~o, 11.26. H o w e v e r , the f o r m a t i o n of a type of neuroglial scar is ubiquitous. L o w e r vertebrates such as fish and a m p h i b i a r e g e n e r a t e nerve fibers across the site of lesion and function is restored4, 5. These regenerating spinal axons in lower vertebrates grow through a neuroglial scar 3. Consequently, in fish and a m p h i b i a the neuroglial scar does not play a role in mechanically arresting the growth of centrally derived nerve fibers. In contrast, the concept of the neuroglial scar mechanically arresting the regenerating central spinal nerve fibers and sprouts has been accepted in higher vertebrates2,10,25, 26. If a drug or any t r e a t m e n t was utilized to induce r e g e n e r a t i o n of central nervous system axons or axonal sprouts after lesion, it was postulated that an otherwise i m p e n e t r a b l e neuroglial scar was ' l o o s e n e d ' and allowed for nerve fiber growth. Neuropath010gically, the fact that human and o t h e r higher v e r t e b r a t e lesioned spinal cord
and brain f o r m e d neuroglial scars reinforced this concept of neuroglial scars as an i m p e n e t r a b l e interface between the central nervous system (CNS) and the periphery2,6,7. Recently, many experiments on transplantation of neuroglial scars into optic and peripheral nerve of lower and higher v e r t e b r a t e s tested this hypothesis 1. H o w e v e r , the results in relationship to in situ spinal cord lesion and the neuroglial scar have not been e l a b o r a t e d . In m a m m a l i a n spinal cord, after induction of nerve fiber or nerve fiber sprout growth by CNS lesion, the neuroglial scar (in situ) did not a p p e a r to be a mechanical barrier since the nerve fibers grew through the lesion site 2.6,10A9,26. Following spinal cord lesion in the rat, multiple neuroglial scars in the form of cellular sheets, form in the lesioned spinal cord stumps, at times millimeters from the lesioned surface2,25. A f t e r lesion the spinal cord at the interface with the cicatrix (scar in lesion site) is covered with basal lamina 2,t2,16. This basal lamina is morphologically indistinguishable from normal 2. Since basal lamina covers the brainl spinal cord and astrocytes at blood vessels, the n o r m a l interface of the central nervous system with the periph-
Correspondence: J. Bernstein, V. A. Medical Center (1510), 50 Irving Street, N.W., Washington, DC 20422, U.S.A. 0006-8993/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
136 ery is the basal lamina. Basal lamina in the CNS can be considered a containment, supportive and guiding membrane which functions during development 21, regeneration, and perhaps all during life. This membrane is restored (within 10 days) after spinal cord lesion2,12,t6,25. Basal lamina in other injured and regenerating systems have additional functions. For example, in the frog, basal lamina is the site of synapse formation for regenerating peripheral nerves on muscle even after all the cells of the muscle have been removed 14,21,22. If the concept that the spinal cord basal lamina covering the injured spinal cord prevents the growth of regenerating CNS nerve fibers by containment is valid, then the origin of this basal lamina and its formation would be an important factor in spinal cord regeneration. This concept is partially supported by the fact that centrally derived nerve fibers or their sprouts that regenerate into the cicatrix after lesion of the spinal cord are enclosed in basal lamina and penetrate the neuroglial scar 2,6,25. The following experiment in the rat studies the origin of the basal lamina lining the lesioned surface and lesioned pial surface of the hemisected rat spinal cord. The uitrastructure of the formation of basal lamina was studied over 10 postoperative days. The secretion of a major basal lamina component (laminin) was studied immunohistochemically to determine the cells of origin of this basal lamina component after lesion. MATERIAL AND METHODS Ninety adult, male, Sprague-Dawley rats (250300 g) were used in this series of experiments. Ten animals per operated group (and 10 normals) were utilized on days 1-7 and 10 days after hemisection of the spinal cord (T6). After Chloropent anesthesia (180 mg/kg) the skin was incised over the T6 vertebra and the musculature removed from the spinal column with a subsequent laminectomy. The spinal cord was hemisected with a scalpel and the wound closed. After the appropriate time postoperative, 5 animals were perfused with 5% glutaraldehyde in phosphate buffer (pH 7.2), dehydrated, and embedded in Spurr. Thick sections were stained with toluidine blue and thin sections were impregnated with uranyl
acetate and lead citrate. For immunohistochemistry the remaining 5 animals at day 1-7, and 10 days (5 normals) were utilized for laminin immunohistochemistry. Laminin antibody conjugated and not conjugated with peroxidase and laminin were obtained from E-Y Laboratories. Laminin antibody was also conjugated with peroxidase (Sigma VI) utilizing the biotin-avidin method (Vector Laboratories). The optimal dilutions for antilaminin-biotin-avidin-peroxidase were 468 mg/ml for 2.5 h or 78 mg/ml for 17 h. Controls were carried out with every procedure which included suppression of endemic peroxidase with methanol and incubation with laminin and laminin antibody followed by peroxidase for comparisons of specificity of response, The chromagen was diaminobenzidine (DAB). Tissue was embedded in epoxy resin. RESULTS Ultrastructure At 1 and 2 days after hemisection, the spinal cord at the lesion site was replete with degenerating myelin, necrotic cells, membrane profiles and macrophages. It was difficult to define a zone that delineated the spinal cord from the lesion. At 3 days the lesioned surface of the spinal cord was discernable. Polymerizing basal lamina was observed as discontinuous wisps of electron dense particles in a slightly electron dense matrix which projected from the surface of astrocytes and their processes (Fig. 1A, B). Basal lamina formed on the surface membranes of astrocytes or their processes which faced the lesion site, endothelial cells, vacuoles in the spinal cord and/or the pial surface. Membranes of the same cell process not facing these zones did not show condensations of basal lamina. In addition to basal lamina associated with the membrane, electron dense condensations of basal lamina like structures were observed at distances from the polymerizing layer of basal lamina associated with the cell surface. At 4 and 5 days after hemisection, the basal lamina was a relatively thick broad sheet with variegated light and dark electron dense particles in a slightly electron dense matrix. The basal lamina on the cell surface of astrocytes was observed in the same zones mentioned previously (Fig. 1C, D). There appeared
137
!i¸:!!
to be repetitions of the basal lamina with two or 3 polymerizing laminae (Fig. 2A). In zones where the astrocytic surfaces were facing the endothelial cells of blood vessels basal lamina was polymerizing in what morphologically appeared to be swirls. However, the basal lamina could have been a single folded sheet in instances of what appeared to be duplications or swirls. At 6 - 1 0 days after hemisection the basal lamina on the astrocytes and their processes had returned to normal limits (Fig. 2B, C). Basal lamina on the membranes of endothelial (Fig. 3A) and pial cells or astrocytes facing the lesion usually had only a single basal lamina membrane (Fig. 3B, C). However, basal lamina repetitions in the form of folds or more than one lamina were observed. This was particularly the case in the perivascular space where the basal lamina secreted by the endothelia and astrocytes was in swirls.
Immunohistochemistry Normal animals only showed antilaminin reaction product on basal lamina of endothelial and pial cells and basal lamina covering the spinal cord. At 1 and 2 days postoperative specific reaction was as observed in the normal controls. At 3 - 5 days posthemisection antilaminin reaction product was observed in light microscopy as a reddish to dark brown precipitate with black granules. The reaction product was observed in cells (Fig. 3D) and cell processes (Fig. 3E) that were in juxtaposition to
Fig. 1. A: astrocyte (A) at the pial surface (site of lesion) 3 days after hemisection of the spinal cord. Wisps of basal lamina (arrows) are polymerizing on the surface of the astrocyte (x 17,450, all sections impregnated with lead citrate and uranyl acetate). B: astrocytic cell process lining vacuoles (see Fig. 3E) in the spinal cord 3 days after spinal cord hemisection. Wisps of basal lamina (arrows) are polymerizing on membrane surfaces juxtaposing the vacuoles (x 23,650). C: astrocyte (A) on the lesioned surface of the spinal cord 4 days after hemisection. The basal lamina (arrows) forms a continuous layer over the surface of the astrocyte (x 7400). D: blood vessel (V) with epithelial cells (E) in the spinal cord 4 days after hemisection. The perivascular space (P) is lined with basal lamina. One basal lamina covers the epithelial cell (curved arrow) and another basal lamina (arrows) covers the surface of the astrocytic processes (A) near the lesioned surface of the hemisected spinal cord (x 15,250).
138 vacuoles, lesion site, endothelial and pial cells. Light microscopically the cells which had antilaminin reaction product were astrocytes: Ultrastructurally, antilaminin reaction product was observed in the cyto-
/~ iii~/¸¸ ! /~i~i~ ?~ ~
iI~ i
plasm and basal lamina of pial, endothelial cells astrocytes and their processes. A t 6 - 1 0 days the normal pattern of reaction product associated with basal lamina was restored. This was the endothelial, pial and brain surface basal lanaina (which includes the basal lamina on neuroglial end feet at blood vessels). There was no longer any visually detectable antilaminin reaction product in either astrocytes, endothelial, pial cells or their processes. DISCUSSION These data show that, within the limits of the technique e m p l o y e d , astrocytes secrete laminin for at least 3 - 5 days at the site of interface between the cicatrix (scar in site of lesion), blood vessels, vacuoles and at the injured pial surface of the spinal cord after lesion. Laminin is a precursor of basal lamina and is secreted with procollagen IV, h e p a r a n sulfate proteoglycan and other basal lamina componentslS. These constituents polymerize on the cell surface and form the characteristic basal lamina of the cell s.9, 13,15,21. The basal lamina formed by astrocytes after spinal cord lesion is morphologically indistinguishable from normal 2.~s. However, there are duplications (two or 3 basal laminae), folded sheets or swirls of basal lamina. Because of p r o b l e m s in interpretation of the electron micrographs the basal lamina may be a continuous folded sheet rather than separate basal laminae. The fact that laminin is detected from days 3 - 5 and not before or after that period of time does not mean that laminin and other basal lamina precursors are not produced by astrocytes before or after this period of time after lesion. Astrocytes should produce basal lamina precursors (laminin, procollagen IV or hepa-
Fig. 2. A: basal lamina duplications or folded sheet (arrow) on the surface of an astrocyte (A) containing lipid droplets (L) on the surface of the hemisection at the cicatrix 5 days after lesion (× 15,550). B: astrocytic cell process lining vacuoles (see Figs. 1B, 3E) with basal lamina covered surfaces (arrows) 10 days after hemisection of the spinal cord (× 14,000). C; basal lamina on the surface of astrocytic processes juxtaposing the cicatrix (C) which contains a degenerating (D) cell, 10 days after spinal cord hemisection. The surface of the same astrocytic process abutting the neuron (N) in the spinal cord does not have basal lamina (× 12,451!).
139
......ii ¸ <~i!i!i!ii!~i!i~!iiiiiiiiiii:ii!!i;:iii;!;?~,' ~
ran sulfate) at all/or some portions of their life cycle. Normally, basal lamina is found overlying astrocytes and astrocytic cell processes at the brain and spinal cord surface and at perivascular spaces2,~2.16.18.24.25. In addition, endothelial, pial and Schwann cells nor-
Fig. 3. A: blood vessel (V) in the spinal cord subjacent to the hemisection 10 days postoperative. The endothelial cells have basal lamina covered surfaces lining the perivascular space surface (P). The astrocytic cell processes have basal lamina (arrows) on the surfaces juxtaposing the perivascular space (P). B: astrocytic cell processes (A) covered with basal lamina (arrow) juxtaposing the cicatrix 10 days after hemisection. The spinal cord contains torpedos (T; ends of severed nerve fibers) and enlarged extracellular spaces not lined with basal lamina (x 12,450). C: astrocyte at the lesioned pial surface (10 days post hemisection) with basal lamina (arrows) on the cell surface toward the pia (x 6000). D: electron micrograph of antilaminin reaction product in an astrocyte in the spinal cord subjacent to the site of hemisection (3 days post hemisection) (x 4500). E: oil objective light micrograph of antilaminin reaction product in astrocytic cytoplasm (curved arrow) and astrocytic cell processes (arrows) (see Figs. 1B and 2C).
maUy secrete basal lamina 23. The normal and 6 - 1 0 day lesioned spinal cord immunohistochemical reaction showed the same specificity. A t 6 - 1 0 days after lesion and normally, only the basal lamina of endothelial, pial, and astrocytic cell surfaces contained
140 antilaminin reaction product. However, basal lamina must be renewed normally and astrocytes and their processes would be expected to secrete the precur-
ter spinal cord lesion is an important anatomical event that restores and protects the homeostatic processes of the CNS. However, if basal lamina is a
sors of basal lamina during their life cycle. For this
containment m e m b r a n e in the CNS of mammals
reason, the reactive astrocytic reaction following spinal cord injury should be viewed as an a u g m e n t a t i o n
(since all central nervous elements are inside basal
of the normal laminin secretion by normal astrocytes. Astrocytes from 5-day-old fetuses in primary culture only produce iaminin for a few days 17. The pres-
lamina and regenerating CNS axons or sprouts do not penetrate the basal lamina:.~,~s), the restitution of this m e m b r a n e could be detrimental to the regeneration of centrally derived nerve fibers across the site of
ent data are in agreement with these findings. The astrocytic cellular response in situ is a regenerative and reparative response of reactive astrocytes after in-
lesion in the injured spinal cord. This action of basal
jury. Watson 23,24 has postulated that astrocytes can be
this does not deter central nervous system regenera-
lamina may be specific for higher vertebrates since lower vertebrate neuroglia form basal lamina and tion in fish and amphibia 1'~.2~.
considered the endothelial cell of the CNS because of their reaction to lesion and the normal interaction of astrocytes (basal lamina production) with pial and endothelial cells. The present data support aspects of this hypothesis. The reformation of basal lamina af-
ACKNOWLEDGEMENTS
REFERENCES
12 Feringa, E. R., Kowalski, T. F. and Vahlsing, H. L., Basal lamina formation at the site of spinal cord transection, Ann. Neurol., 8 (1979) 148-154. 13 Garbi, C. and Wollman, S. H., Basal lamina formation in thyroid epithelia in separated follicles in suspension culture, J. Cell Biol., 94 (t982) 489-492. 14 Glickman, M. and Sanes, J., Differentiation of motor nerve terminals formed in the absence of muscle fibers, J. Neurocytol., 12 (1983) 661-671. 15 Kleinman, H. K., McGarvey, M. L., Liotta, L. A., Robey, P. G., Tryaavason, K. and Martin, G. R., Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from the EHS sarcoma, Biochemistry, 21 (1982) 6188-6193. 16 Kowalski, T. F., Vahlsing, H. L. and Feringa, E. R., Light microscopic, immunohistochemical localization of the piaglial basal lamina, J. Histochem. Cytochem.. 28 (1980) 347353. 17 Liesi, P., Dahl, D. and Vaheri, A., Laminin is produced by early rat astrocytes in primary culture, J. Cell Biol., 96 (1983) 920-924. 18 Peters, A., Palay, S. and De Webster, H., The Fine Structure of the Nervous System: Neurons and Supporting Cells, W. B. Saunders Company, Philadelphia, 1976, pp. 248-254. 19 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. 163-193. 20 Reier, P. J. and Webster, H. deF., Remyelination in regenerating optic nerves of Xenopus tadpoles, J. Neurocytol.. 3 (1974) 591-618. 21 Sanes, J., Roles of extracellular matrix in neural development, Ann. Rev. Physiol.. 45 (1983) 581-60//. 22 Sanes, J., Marshall, U. and McMahan, U., Rcinnervation
1 Aquayo, A.. Dickerson, R., lrecarten, J., Attiwell, M., Bray, G. and Richardson, P., Ensheathment and myelination of regenerating PNS fibers by transplanting optic nerve gila. Neurosci. Lett., 9 (1978) 92-104. 2 Bernstein, J. J., During glial scar formation and cavatation necrosis after injury, the spinal cord regenerates neuronal cell processes. In F. Sell (Ed.), Nerve, Organ, and Tissue Regeneration, Academic Press, New York, 1983, pp. 215-230. 3 Bernstein, J. J. and Bernstein, M. E., Effects of glial-ependymal scar and teflon arrest on the regenerative capacity of goldfish spinal cord, Exp. Neurol., 193 (1967) 25-32. 4 Bernstein, J. J. and Gelderd, J. B., Regeneration of the long tracts in the goldfish, Brain Research, 20 (1970) 33-38. 5 Bernstein, J. J. and Getderd, J. B., Synaptic reorganization following regeneration of goldfish spinal cord, Exp. Neurol., 41 (1973) 402-410. 6 Bernstein, J. J. and Wells, M. R., Puromycin induction of transient regeneration in mammalian spinal cord, Brain Research, 43 (1980) 21-38. 7 Blackwood, W., McMenemey, W. H., Meyer, A, and Norman, R. M., Greenfield's Neuropathology, Edward Arnold, London, 1969, pp. 45-65. 8 Brownell, A. G. and Slavkin, H. C., Role of basal lamina in tissue interactions, Renal Physiol., 3 (1980) 193-204. 9 Brownell, A. G., Bessem, C. C. and Slavkin, H. C., Possible functions of mesenchyme cell-derived fibronectin during formation of basal lamina, Proc. nat. Acad. Sci. (U.S.A.), 78 (1981) 3711-3715. 10 Clemente, C. D., Regeneration in the vertebrate central nervous system, Int. Rev. Neurobiol., 6 (1964) 257-301. 11 Cotman, C., Nieto-Sampedro, M. and Harris, E., Synapse replacement in the nervous system of adult vertebrates, Physiol. Rev., 61 (1981)684-784.
Supported by the Veterans Administration. The authors thank Mr. R. Maddox for his aid.
141 of muscle fiber basal lamina after removal of myofibers: Differentiation of regenerating axons at original synaptic sites, J. CellBiol., 78 (1978) 176-198. 23 Watson, W. E., Physiology of neuroglia, Physiol. Rev., 54 (1974) 245-271. 24 Watson, W. E., Cell Biology of Brain, Chapman and Hall, London, 1976, pp. 527.
25 Wells, M. R. and Bernstein, J. J., Scar formation and the barrier hypothesis in the failure of mammalian central nervous system regeneration. In J. Jane and O. Steward (Eds.), Spinal Cord Trauma, Raven Press, New York, in press. 26 Windle, W. F., Regeneration in the Central Nervous System, C. C. Thomas, Springfield, IL, 1955, pp. 311.
135
BRE 10524
Astrocytes Secrete Basal Lamina After Hemisection of Rat Spinal Cord JERALD J. BERNSTEIN, REINA GETZ, MARY JEFFERSON and MURIEL KELEMEN
Veterans Administration Medical Center, Washington, DC 20422 and Department of Physiology and Neurosurgery, George Washington University School of Medicine, Washington, DC (U.S.A.) (Accepted May 15th, 1984)
Key words: regeneration - - basal lamina - - ultrastructure - - spinal cord - - injury - - histochemistry
Basal lamina is reconstructed over the lesioned surface of the spinal cord. The following experiment (90 rats) studies the ultrastructure of the formation of this membrane and the immunohistochemistry of laminin production (a major secreted component of basal lamina). After hemisection of the spinal cord at T6 animals were prepared for electron microscopy or antilaminin-biotin-avidin-peroxidase incubation. Three-5 days posthemisection, antilaminin reaction product was observed in astrocytes and their processes which faced the lesion, endothelia of blood vessels or pia. Ultrastructurally (3 days), basal lamina was polymerizing as small projections on the surface of astrocytic membranes facing the lesion, endothelia or pia. By 5 days the basal lamina was a single membrane, folded multiple sheets or in swirls. At 6-10 days the antilaminin reaction and the basal lamina (except for duplications) did not differ from normal. Reactive astrocytes secrete laminin for at least 3-5 days after hemisection and form basal lamina on the lesioned surface of the spinal cord after spinal cord hemisection.
INTRODUCTION The cellular responses of the spinal cord to lesion are varied and differ with hierarchy throughout the animal kingdom2,~o, 11.26. H o w e v e r , the f o r m a t i o n of a type of neuroglial scar is ubiquitous. L o w e r vertebrates such as fish and a m p h i b i a r e g e n e r a t e nerve fibers across the site of lesion and function is restored4, 5. These regenerating spinal axons in lower vertebrates grow through a neuroglial scar 3. Consequently, in fish and a m p h i b i a the neuroglial scar does not play a role in mechanically arresting the growth of centrally derived nerve fibers. In contrast, the concept of the neuroglial scar mechanically arresting the regenerating central spinal nerve fibers and sprouts has been accepted in higher vertebrates2,10,25, 26. If a drug or any t r e a t m e n t was utilized to induce r e g e n e r a t i o n of central nervous system axons or axonal sprouts after lesion, it was postulated that an otherwise i m p e n e t r a b l e neuroglial scar was ' l o o s e n e d ' and allowed for nerve fiber growth. Neuropath010gically, the fact that human and o t h e r higher v e r t e b r a t e lesioned spinal cord
and brain f o r m e d neuroglial scars reinforced this concept of neuroglial scars as an i m p e n e t r a b l e interface between the central nervous system (CNS) and the periphery2,6,7. Recently, many experiments on transplantation of neuroglial scars into optic and peripheral nerve of lower and higher v e r t e b r a t e s tested this hypothesis 1. H o w e v e r , the results in relationship to in situ spinal cord lesion and the neuroglial scar have not been e l a b o r a t e d . In m a m m a l i a n spinal cord, after induction of nerve fiber or nerve fiber sprout growth by CNS lesion, the neuroglial scar (in situ) did not a p p e a r to be a mechanical barrier since the nerve fibers grew through the lesion site 2.6,10A9,26. Following spinal cord lesion in the rat, multiple neuroglial scars in the form of cellular sheets, form in the lesioned spinal cord stumps, at times millimeters from the lesioned surface2,25. A f t e r lesion the spinal cord at the interface with the cicatrix (scar in lesion site) is covered with basal lamina 2,t2,16. This basal lamina is morphologically indistinguishable from normal 2. Since basal lamina covers the brainl spinal cord and astrocytes at blood vessels, the n o r m a l interface of the central nervous system with the periph-
Correspondence: J. Bernstein, V. A. Medical Center (1510), 50 Irving Street, N.W., Washington, DC 20422, U.S.A. 0006-8993/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
136 ery is the basal lamina. Basal lamina in the CNS can be considered a containment, supportive and guiding membrane which functions during development 21, regeneration, and perhaps all during life. This membrane is restored (within 10 days) after spinal cord lesion2,12,t6,25. Basal lamina in other injured and regenerating systems have additional functions. For example, in the frog, basal lamina is the site of synapse formation for regenerating peripheral nerves on muscle even after all the cells of the muscle have been removed 14,21,22. If the concept that the spinal cord basal lamina covering the injured spinal cord prevents the growth of regenerating CNS nerve fibers by containment is valid, then the origin of this basal lamina and its formation would be an important factor in spinal cord regeneration. This concept is partially supported by the fact that centrally derived nerve fibers or their sprouts that regenerate into the cicatrix after lesion of the spinal cord are enclosed in basal lamina and penetrate the neuroglial scar 2,6,25. The following experiment in the rat studies the origin of the basal lamina lining the lesioned surface and lesioned pial surface of the hemisected rat spinal cord. The uitrastructure of the formation of basal lamina was studied over 10 postoperative days. The secretion of a major basal lamina component (laminin) was studied immunohistochemically to determine the cells of origin of this basal lamina component after lesion. MATERIAL AND METHODS Ninety adult, male, Sprague-Dawley rats (250300 g) were used in this series of experiments. Ten animals per operated group (and 10 normals) were utilized on days 1-7 and 10 days after hemisection of the spinal cord (T6). After Chloropent anesthesia (180 mg/kg) the skin was incised over the T6 vertebra and the musculature removed from the spinal column with a subsequent laminectomy. The spinal cord was hemisected with a scalpel and the wound closed. After the appropriate time postoperative, 5 animals were perfused with 5% glutaraldehyde in phosphate buffer (pH 7.2), dehydrated, and embedded in Spurr. Thick sections were stained with toluidine blue and thin sections were impregnated with uranyl
acetate and lead citrate. For immunohistochemistry the remaining 5 animals at day 1-7, and 10 days (5 normals) were utilized for laminin immunohistochemistry. Laminin antibody conjugated and not conjugated with peroxidase and laminin were obtained from E-Y Laboratories. Laminin antibody was also conjugated with peroxidase (Sigma VI) utilizing the biotin-avidin method (Vector Laboratories). The optimal dilutions for antilaminin-biotin-avidin-peroxidase were 468 mg/ml for 2.5 h or 78 mg/ml for 17 h. Controls were carried out with every procedure which included suppression of endemic peroxidase with methanol and incubation with laminin and laminin antibody followed by peroxidase for comparisons of specificity of response, The chromagen was diaminobenzidine (DAB). Tissue was embedded in epoxy resin. RESULTS Ultrastructure At 1 and 2 days after hemisection, the spinal cord at the lesion site was replete with degenerating myelin, necrotic cells, membrane profiles and macrophages. It was difficult to define a zone that delineated the spinal cord from the lesion. At 3 days the lesioned surface of the spinal cord was discernable. Polymerizing basal lamina was observed as discontinuous wisps of electron dense particles in a slightly electron dense matrix which projected from the surface of astrocytes and their processes (Fig. 1A, B). Basal lamina formed on the surface membranes of astrocytes or their processes which faced the lesion site, endothelial cells, vacuoles in the spinal cord and/or the pial surface. Membranes of the same cell process not facing these zones did not show condensations of basal lamina. In addition to basal lamina associated with the membrane, electron dense condensations of basal lamina like structures were observed at distances from the polymerizing layer of basal lamina associated with the cell surface. At 4 and 5 days after hemisection, the basal lamina was a relatively thick broad sheet with variegated light and dark electron dense particles in a slightly electron dense matrix. The basal lamina on the cell surface of astrocytes was observed in the same zones mentioned previously (Fig. 1C, D). There appeared
137
!i¸:!!
to be repetitions of the basal lamina with two or 3 polymerizing laminae (Fig. 2A). In zones where the astrocytic surfaces were facing the endothelial cells of blood vessels basal lamina was polymerizing in what morphologically appeared to be swirls. However, the basal lamina could have been a single folded sheet in instances of what appeared to be duplications or swirls. At 6 - 1 0 days after hemisection the basal lamina on the astrocytes and their processes had returned to normal limits (Fig. 2B, C). Basal lamina on the membranes of endothelial (Fig. 3A) and pial cells or astrocytes facing the lesion usually had only a single basal lamina membrane (Fig. 3B, C). However, basal lamina repetitions in the form of folds or more than one lamina were observed. This was particularly the case in the perivascular space where the basal lamina secreted by the endothelia and astrocytes was in swirls.
Immunohistochemistry Normal animals only showed antilaminin reaction product on basal lamina of endothelial and pial cells and basal lamina covering the spinal cord. At 1 and 2 days postoperative specific reaction was as observed in the normal controls. At 3 - 5 days posthemisection antilaminin reaction product was observed in light microscopy as a reddish to dark brown precipitate with black granules. The reaction product was observed in cells (Fig. 3D) and cell processes (Fig. 3E) that were in juxtaposition to
Fig. 1. A: astrocyte (A) at the pial surface (site of lesion) 3 days after hemisection of the spinal cord. Wisps of basal lamina (arrows) are polymerizing on the surface of the astrocyte (x 17,450, all sections impregnated with lead citrate and uranyl acetate). B: astrocytic cell process lining vacuoles (see Fig. 3E) in the spinal cord 3 days after spinal cord hemisection. Wisps of basal lamina (arrows) are polymerizing on membrane surfaces juxtaposing the vacuoles (x 23,650). C: astrocyte (A) on the lesioned surface of the spinal cord 4 days after hemisection. The basal lamina (arrows) forms a continuous layer over the surface of the astrocyte (x 7400). D: blood vessel (V) with epithelial cells (E) in the spinal cord 4 days after hemisection. The perivascular space (P) is lined with basal lamina. One basal lamina covers the epithelial cell (curved arrow) and another basal lamina (arrows) covers the surface of the astrocytic processes (A) near the lesioned surface of the hemisected spinal cord (x 15,250).
138 vacuoles, lesion site, endothelial and pial cells. Light microscopically the cells which had antilaminin reaction product were astrocytes: Ultrastructurally, antilaminin reaction product was observed in the cyto-
/~ iii~/¸¸ ! /~i~i~ ?~ ~
iI~ i
plasm and basal lamina of pial, endothelial cells astrocytes and their processes. A t 6 - 1 0 days the normal pattern of reaction product associated with basal lamina was restored. This was the endothelial, pial and brain surface basal lanaina (which includes the basal lamina on neuroglial end feet at blood vessels). There was no longer any visually detectable antilaminin reaction product in either astrocytes, endothelial, pial cells or their processes. DISCUSSION These data show that, within the limits of the technique e m p l o y e d , astrocytes secrete laminin for at least 3 - 5 days at the site of interface between the cicatrix (scar in site of lesion), blood vessels, vacuoles and at the injured pial surface of the spinal cord after lesion. Laminin is a precursor of basal lamina and is secreted with procollagen IV, h e p a r a n sulfate proteoglycan and other basal lamina componentslS. These constituents polymerize on the cell surface and form the characteristic basal lamina of the cell s.9, 13,15,21. The basal lamina formed by astrocytes after spinal cord lesion is morphologically indistinguishable from normal 2.~s. However, there are duplications (two or 3 basal laminae), folded sheets or swirls of basal lamina. Because of p r o b l e m s in interpretation of the electron micrographs the basal lamina may be a continuous folded sheet rather than separate basal laminae. The fact that laminin is detected from days 3 - 5 and not before or after that period of time does not mean that laminin and other basal lamina precursors are not produced by astrocytes before or after this period of time after lesion. Astrocytes should produce basal lamina precursors (laminin, procollagen IV or hepa-
Fig. 2. A: basal lamina duplications or folded sheet (arrow) on the surface of an astrocyte (A) containing lipid droplets (L) on the surface of the hemisection at the cicatrix 5 days after lesion (× 15,550). B: astrocytic cell process lining vacuoles (see Figs. 1B, 3E) with basal lamina covered surfaces (arrows) 10 days after hemisection of the spinal cord (× 14,000). C; basal lamina on the surface of astrocytic processes juxtaposing the cicatrix (C) which contains a degenerating (D) cell, 10 days after spinal cord hemisection. The surface of the same astrocytic process abutting the neuron (N) in the spinal cord does not have basal lamina (× 12,451!).
139
......ii ¸ <~i!i!i!ii!~i!i~!iiiiiiiiiii:ii!!i;:iii;!;?~,' ~
ran sulfate) at all/or some portions of their life cycle. Normally, basal lamina is found overlying astrocytes and astrocytic cell processes at the brain and spinal cord surface and at perivascular spaces2,~2.16.18.24.25. In addition, endothelial, pial and Schwann cells nor-
Fig. 3. A: blood vessel (V) in the spinal cord subjacent to the hemisection 10 days postoperative. The endothelial cells have basal lamina covered surfaces lining the perivascular space surface (P). The astrocytic cell processes have basal lamina (arrows) on the surfaces juxtaposing the perivascular space (P). B: astrocytic cell processes (A) covered with basal lamina (arrow) juxtaposing the cicatrix 10 days after hemisection. The spinal cord contains torpedos (T; ends of severed nerve fibers) and enlarged extracellular spaces not lined with basal lamina (x 12,450). C: astrocyte at the lesioned pial surface (10 days post hemisection) with basal lamina (arrows) on the cell surface toward the pia (x 6000). D: electron micrograph of antilaminin reaction product in an astrocyte in the spinal cord subjacent to the site of hemisection (3 days post hemisection) (x 4500). E: oil objective light micrograph of antilaminin reaction product in astrocytic cytoplasm (curved arrow) and astrocytic cell processes (arrows) (see Figs. 1B and 2C).
maUy secrete basal lamina 23. The normal and 6 - 1 0 day lesioned spinal cord immunohistochemical reaction showed the same specificity. A t 6 - 1 0 days after lesion and normally, only the basal lamina of endothelial, pial, and astrocytic cell surfaces contained
140 antilaminin reaction product. However, basal lamina must be renewed normally and astrocytes and their processes would be expected to secrete the precur-
ter spinal cord lesion is an important anatomical event that restores and protects the homeostatic processes of the CNS. However, if basal lamina is a
sors of basal lamina during their life cycle. For this
containment m e m b r a n e in the CNS of mammals
reason, the reactive astrocytic reaction following spinal cord injury should be viewed as an a u g m e n t a t i o n
(since all central nervous elements are inside basal
of the normal laminin secretion by normal astrocytes. Astrocytes from 5-day-old fetuses in primary culture only produce iaminin for a few days 17. The pres-
lamina and regenerating CNS axons or sprouts do not penetrate the basal lamina:.~,~s), the restitution of this m e m b r a n e could be detrimental to the regeneration of centrally derived nerve fibers across the site of
ent data are in agreement with these findings. The astrocytic cellular response in situ is a regenerative and reparative response of reactive astrocytes after in-
lesion in the injured spinal cord. This action of basal
jury. Watson 23,24 has postulated that astrocytes can be
this does not deter central nervous system regenera-
lamina may be specific for higher vertebrates since lower vertebrate neuroglia form basal lamina and tion in fish and amphibia 1'~.2~.
considered the endothelial cell of the CNS because of their reaction to lesion and the normal interaction of astrocytes (basal lamina production) with pial and endothelial cells. The present data support aspects of this hypothesis. The reformation of basal lamina af-
ACKNOWLEDGEMENTS
REFERENCES
12 Feringa, E. R., Kowalski, T. F. and Vahlsing, H. L., Basal lamina formation at the site of spinal cord transection, Ann. Neurol., 8 (1979) 148-154. 13 Garbi, C. and Wollman, S. H., Basal lamina formation in thyroid epithelia in separated follicles in suspension culture, J. Cell Biol., 94 (t982) 489-492. 14 Glickman, M. and Sanes, J., Differentiation of motor nerve terminals formed in the absence of muscle fibers, J. Neurocytol., 12 (1983) 661-671. 15 Kleinman, H. K., McGarvey, M. L., Liotta, L. A., Robey, P. G., Tryaavason, K. and Martin, G. R., Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from the EHS sarcoma, Biochemistry, 21 (1982) 6188-6193. 16 Kowalski, T. F., Vahlsing, H. L. and Feringa, E. R., Light microscopic, immunohistochemical localization of the piaglial basal lamina, J. Histochem. Cytochem.. 28 (1980) 347353. 17 Liesi, P., Dahl, D. and Vaheri, A., Laminin is produced by early rat astrocytes in primary culture, J. Cell Biol., 96 (1983) 920-924. 18 Peters, A., Palay, S. and De Webster, H., The Fine Structure of the Nervous System: Neurons and Supporting Cells, W. B. Saunders Company, Philadelphia, 1976, pp. 248-254. 19 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. 163-193. 20 Reier, P. J. and Webster, H. deF., Remyelination in regenerating optic nerves of Xenopus tadpoles, J. Neurocytol.. 3 (1974) 591-618. 21 Sanes, J., Roles of extracellular matrix in neural development, Ann. Rev. Physiol.. 45 (1983) 581-60//. 22 Sanes, J., Marshall, U. and McMahan, U., Rcinnervation
1 Aquayo, A.. Dickerson, R., lrecarten, J., Attiwell, M., Bray, G. and Richardson, P., Ensheathment and myelination of regenerating PNS fibers by transplanting optic nerve gila. Neurosci. Lett., 9 (1978) 92-104. 2 Bernstein, J. J., During glial scar formation and cavatation necrosis after injury, the spinal cord regenerates neuronal cell processes. In F. Sell (Ed.), Nerve, Organ, and Tissue Regeneration, Academic Press, New York, 1983, pp. 215-230. 3 Bernstein, J. J. and Bernstein, M. E., Effects of glial-ependymal scar and teflon arrest on the regenerative capacity of goldfish spinal cord, Exp. Neurol., 193 (1967) 25-32. 4 Bernstein, J. J. and Gelderd, J. B., Regeneration of the long tracts in the goldfish, Brain Research, 20 (1970) 33-38. 5 Bernstein, J. J. and Getderd, J. B., Synaptic reorganization following regeneration of goldfish spinal cord, Exp. Neurol., 41 (1973) 402-410. 6 Bernstein, J. J. and Wells, M. R., Puromycin induction of transient regeneration in mammalian spinal cord, Brain Research, 43 (1980) 21-38. 7 Blackwood, W., McMenemey, W. H., Meyer, A, and Norman, R. M., Greenfield's Neuropathology, Edward Arnold, London, 1969, pp. 45-65. 8 Brownell, A. G. and Slavkin, H. C., Role of basal lamina in tissue interactions, Renal Physiol., 3 (1980) 193-204. 9 Brownell, A. G., Bessem, C. C. and Slavkin, H. C., Possible functions of mesenchyme cell-derived fibronectin during formation of basal lamina, Proc. nat. Acad. Sci. (U.S.A.), 78 (1981) 3711-3715. 10 Clemente, C. D., Regeneration in the vertebrate central nervous system, Int. Rev. Neurobiol., 6 (1964) 257-301. 11 Cotman, C., Nieto-Sampedro, M. and Harris, E., Synapse replacement in the nervous system of adult vertebrates, Physiol. Rev., 61 (1981)684-784.
Supported by the Veterans Administration. The authors thank Mr. R. Maddox for his aid.
141 of muscle fiber basal lamina after removal of myofibers: Differentiation of regenerating axons at original synaptic sites, J. CellBiol., 78 (1978) 176-198. 23 Watson, W. E., Physiology of neuroglia, Physiol. Rev., 54 (1974) 245-271. 24 Watson, W. E., Cell Biology of Brain, Chapman and Hall, London, 1976, pp. 527.
25 Wells, M. R. and Bernstein, J. J., Scar formation and the barrier hypothesis in the failure of mammalian central nervous system regeneration. In J. Jane and O. Steward (Eds.), Spinal Cord Trauma, Raven Press, New York, in press. 26 Windle, W. F., Regeneration in the Central Nervous System, C. C. Thomas, Springfield, IL, 1955, pp. 311.