Neuron,
Vol. 5, 463-469,
October,
1990, Copyright
0
1990 by Cell Press
Macrophages Can Modify the Nonpermissive Nature of the Adult Mammalian Central Nervous System S. David, C. Bouchard, 0. Tsatas, and N. Giftochristos Centre for Research in Neuroscience McGill University Montreal General Hospital Research Montreal, Quebec Canada, H3G IA4
Institute
Summary Although astrocytic gliosis has been linked to failure of axonal regeneration in the adult mammalian CNS, its role is not fully established. We used an in vitro assay to investigate the role of reactive astrocytes and macrophages in influencing axonal growth in the lesioned adult rat optic nerve. Soon after optic nerve transection, the nonpermissive nature of the optic nerve is altered to a permissive state near the lesion. This may account for injury-induced axonal sprouting and may contribute to the failure of these sprouts to elongate beyond the site of the lesion in vivo. We provide evidence that this lesion-induced change in the axonal growth-promoting properties of the CNS near the lesion may be produced by mononuclear phagocytes. In addition, several months after optic nerve transection, the degenerated nerves, which consist mainly of astrocytes and lack myelin, i.e., astrocytic “scar” tissue, are a good substrate for neurite growth. Taken together, these results suggest that in this in vitro system, substantial inhibitory effects are not associated with regions of astrocytic gliosis and that the nonpermissive nature of the CNS white matter can be modified by macrophages. Introduction Neurons in the adult mammalian CNS have the intrinsic ability to regenerate their axons for long distances under appropriate experimental conditions (Richardson et al., 1980; David and Aguayo, 1981; Benfey and Aguayo, 1982; Friedman and Aguayo, 1985; So and Aguayo, 1985). The failure of CNS neurons to regenerate axons through adult CNS tissue in vivo appears to be a function of the CNS glial environment (Aguayo, 1985). Many of the earlier studies have suggested that astrocytes in the mammalian CNS might hinder or inhibit axonal regeneration (Reier et al., 1989). However, many of these results are open to other interpretations, and therefore the role of astrocytic gliosis in the failure of regeneration is not yet fully established. Astrocytic gliosis has been shown to be stimulated by macrophages (Giulian et al., 1989). It has been reported recently that a minor protein component of mammalian CNS myelin has a marked inhibitory effect on fibroblast spreading and neurite extension in vitro (Caroni and Schwab, 1988a, 1988b; Schwab and Caroni, 1988) and possibly in vivo (Schnell and Schwab,
1990). Other inhibitory molecules also produced by oligodendrocytes have been identified (Pesheva et al., 1989). In this study, we have used an in vitro assay to determine whether astrocytic gliosis also contributes to the failure of axonal growth and to examine the effects of macrophages on the nonpermissive properties of the adult mammalian CNS white matter. The rat optic nerve is a convenient region of the CNS to study the role of glia or glial components in regeneration, since it lacks neuronal cell bodies and is easily accessible for experimental manipulations. Injuries to the CNS results in axonal sprouting that is not sustained for long (Ramon y Cajal, 1928). We and others have shown that retinal ganglion cell axons undergo sprouting soon after an intracranial optic nerve transection (Richardson et al., 1982; Crafstein and Ingoglia, 1982; Giftochristos and David, 1988). It is not known whether the injury-induced axonal sprouting is accompanied by changes in the glial environment. In the present study, unfixed cryostat sections of transected adult rat optic nerve were used as a tissue culture substrate to study the changes in the ability of the injured CNS tissue to support axonal growth. This approach has been used previously to examine the differences between embryonic and adult CNS tissue (Carbonetto et al., 1987) and the differences between the gray and white matter of the adult mammalian CNS (Savio and Schwab, 1989). Results
and
Discussion
Changes in the Nonpermissive Properties of Adult Optic Nerve 5 Days after Transection Cell Aftachmenf Unfixed, longitudinal cryostat sections of transected adult rat optic nerve were obtained 5 days after transection and picked up on poly+-lysine-coated, round glass coverslips (15 mm). These tissue sections were prepared for tissue culture (Carbonetto et al., 1987) and seeded with PC12 cells. A significantly greater number of PC12 cells attached to the tissue sections near the site of the lesion compared with regions distal to the lesion (Figure 1). The PC12 cells attached to regions that contained astrocytes as determined immunohistochemically (Figures IA and IB). These results suggest that CNS tissue near the site of a lesion has greater adhesive properties than regions distal to a lesion. Neurite Growth To determine the ability of optic nerve tissue near the site of a lesion to support neurite growth (which we were unable to visualize for technical reasons using PC12 cells), we cultured unfixed cryostat sections of adult rat optic nerves obtained 5 days after transection with explants of embryonic days 8-10 (E8-EIO) chick dorsal root ganglia (DRGs). After 5 days in vitro, 30% of the DRG explants seeded near the site of the
Figure
I
0’
E
F
1. PC12 Cells
Plated
on
Unfixed
Cryostat
Sections
of Adult
Rat Optic
Nerve
5 Days after
an Optic
N
Nerve
i TA
TN
Transection
Sections have been double-labeled with anti-CFAP (A and C) and nuclear yellow (B and D). (A) Hypertrophied astrocytes near the site of transection. (B) Same section as in (A), showing the attachment of PC12 cells visualized with nuclear yellow. Note that a large number of pC12 cells are attached predominantly to regions that are CFAP+. (C)An area of a transected optic nerve distal to the site of transection. The anti-GFAP labeling of the astrocytic processes is finer and less intense compared with that of regions nearer the lesion. (D) Fewer pC12 cells are attached to these regions distal to the site of transection. (E) Camera lucida drawing showing the distribution of PC12 cells attached to a slngie cryostat section of a transected optic nerve 5 days after transection. The sectlon has been divided into two parts: the hatched area represents the portion of the nerve distal to the lesion (as in [Cl), in which the anti-CFAP labeling is finer and less Intense compared wtth regions of the nerve near the iesion, which contain hypertrophied astrocytes (as in [A]). (F) Histogram of the numbers of PC12 cells atiached to cryostat sections of the normal optic nerve (N), the transected optic nerve distal to the site of lesion (TA) 5 days after transection (hatched area in [El), and the region of the transected nerve near the lesion (TN). Note that a significantly greater number of PC12 cells attach to regions near the lesion (TN). Mean i SD of three experiments; The asterisk indicates P < 0.01; significance levels between the values of the transected and normal nerves ‘were estimated using Student’s t-test. Magnification: 144x (all panels).
optic nerve transection showed extensive neurite growth onto the tissue sections, whereas none of those plated distal to the lesion showed such growth (Figures 2C and 2D; Table 1). The latter growth was similar to that on sections of normal adult rat optic nerve (Carbonetto et al., 1987) (Figures 2A and 2B; Table 1). These results indicate that within a few days after lesioning, the normally nonpermissive CNS white matter is able to support cell attachment and axonal growth near the site of the lesion. There are several possible explanations for this
change in the growth-promoting properties of the injured CNS. One possibility is that reactive astrocytes near the lesion may be induced to express diffusable growth factors and/or neurite growth-promoting adhesion molecules. Since the glial cells in the frozen sections are not viable in our present assay and the tissue sections are washed extensively prior to culturing, it is unlikely that growth factors which may be produced in vivo by astrocytes are responsible for the observed neurite growth during the 5 days in culture. We have also shown previously that the expression of
Changes 465
in the Nonpermissive
CNS Environment
Figure 2. Neurite Growth from EB-El0 Chick DRGs Plated beside Cryostat Sections of Unfixed Optic Nerve The DRC is located at the top of each micrograph. (A) and (B) show the lack of neurite growth onto sections of normal rat optic nerve. In (A) the DRC is located away from the section; in LB) it lies against the optic nerve section.(C) In contrast to the normal nerve, neurites grow extensively onto sections of the optic nerve near the site of a transection, 5 days after lesioning. (D) However, neurite growth on the transected optic nerve, distal to the site of transection, is similar to that on the normal optic nerve. (E) Neurites grow extensively on sections of the normal adult rat optic nerve treated overnight with a nitrocellulose implant obtained 5 days after implanting into a cerebral cortical lesion. (F) Extensive neurite growth on to a section of optic nerve 4 months after transection. Magnification: 27x (A and B); 42x (C-F).
laminin and heparan sulphate proteoglycan are not correlated with the sprouting response in the injured adult rat optic nerve (Giftochristos and David, 1988). It is not known whether any of the other adhesion molecules are expressed differently by reactive astrocytes in close proximity to the lesion. Another possibility is that inhibitory molecules (Caroni and Schwab, 1988a; Pesheva et al., 1989), such as those associated with CNS myelin, may have been inactivated. It is unlikely that these effects are due to the absence of myelin, since CNS myelin is known to take several weeks or months to be cleared from the le-
Table
1. The
Percentage Normal
Growth Extensive Moderate Minimal
(36) 0 16.5 83.5
of Embryonic
Chick
DRGs
Extending
sioned CNS (Perry and Gordon, 1988; Stall et al., 1989). Furthermore, PC12 cells were found to be attached to regions near the lesion that stained positively with a polyclonal antiserum against myelin basic protein (data not shown). It is possible, however, that the inhibitory proteins associated with myelin are inactivated near the site of the lesion soon after lesioning. This would account not only for the axonal sprouting that occurs soon after the induction of CNS lesions, but also for the failure of these sprouts to elongate for long distances away from the lesion site. These inhibitory molecules may be inactivated by
Neurites
onto
Cryostat
Sections
5 Days Trans. Near Lesion t23Ja
5 Days Trans. Distal to Lesion (23)’
4-6 Months Trans. (29jb
Normal ON, NC Implant (20)’
Normal NC/RS Control
30 48 22
0 13 87
45 27.5 27.5
65 15 20
4.5 19.5 76
of Unfixed
Optic
Nerve
ON, (2V
Normal M&M 46 19 35
ON, (28)’
Normal ON, MDP Control
(34)’
8.5 23.5 68
Growth onto the sections of up to 20 neurites was considered minimal growth, between 20 and 50 was considered moderate and more than 50 was considered extensive growth. Numbers in parentheses indicate the number of DRCs examined. a Five days after optic nerve transection. 6 Four to 6 months after optic nerve transection. c Normal optic nerve sections incubated for 24 hr with nitrocellulose implants obtained 5 days after implanting into cerebral d Normal optic nerve sections incubated with nitrocellulose paper soaked in rat serum. e Normal optic nerve sections incubated for 48 hr with activated peritoneal macrophage-conditioned medium. ‘ Normal optic nerve sections incubated for 48 hr with N-acetylmuramyl-t-alanyl-o-isoglutamine.
growth,
cortex.
Figure 3. Longitudinal Optic Nerve 5 Days
Cryostat Section after Transection
(A) EDl’cells are found only near yellow labeling of the same tissue
of a Transected
Adult
change in the nonpermissive properties of the optic nerve 5 days after transection, we examined whether mononuclear phagocytes might be able to modify the nonpermissive properties of the adult rat optic nerve. This was done by treating cryostat sections of unfixed, normal rat optic nerve with mononuclear phagocytes from the iesioned CNS. These cells were obtained by implanting small pieces of nitrocellulose paper into the cerebral cortex of adult rats. After 5 days the nitrocellulose paper was removed and cultured for 24 hr with unfixed frozen sections of normal adult rat optic nerves. Cells emerged from the nitrocellulose filters within a few minutes after being placed in culture. The majority of these cells appeared to be mononuclear phagocytes, because they labeled positively with the monoclonal antibody ED1 (92.5%) and phagocytosed Latex beads in vitro (80.5%) (Figure 4). These values were obtained from counting a total of 3325 cells in two separate experiments. Qnly 2% of these cells were glial fibrillary acidic protein-positive (CFAP+) astrocytes, and all of these were EDI-. Norma! adult rat optic nerve sections incubated overnight with the nitrocellulose implants were capable of supporting excellent neurite growth from embryonic chick DRGs (Figure 2E; Table 1). Similar results were also obtained with medium conditioned by activated peritoneal macrophages (Table ?), thus suggesting that substances released by macrophages may be responsible for this effect. Some nonneuronal cells were associated with the DRG explants despite the addition of cytosine arabinoside. However, the presence of these nonneuronal cells was not sufficient to ensure the growth of neurites onto sections of untreated, normal optic nerve. For the reasons cited earlier, it is unlikely that the permissive properties of the treated ncrmal nerves are due to the increased expression of diffusable growth factors or adhesion molecules by the glial cells, since there are no viable cells in the tissue sections. This effect may therefore reflect the inactivation of inhibitory molecules in the optic nerve sections, or
Rat
the site of lesion. (B) Nuclear section. Magnification: 80x.
substances released by macrophages (Nathan, 1987). We therefore examined the distribution of mononuclear phagocytes in the transected adult rat optic nerve 5 days after a transection. Immunohistochemical Localization of Mononuclear Phagocytes The pattern of distribution of mononuclear phagocytes in the adult rat optic nerve 5 days after nerve transection was determined in longitudinal cryostat sections. Mononuclear phagocytes, which were identified with the monoclonal antibody ED1 (Dijkstra et al., 1985), were located only near the lesion site (Figure 3). This area corresponds to the region that supports cell attachment and neurite growth in vitro and axonal sprouting in vivo. Macrophages Change Optic Nerve Sections Because of the close tion of mononuclear
Figure (A) The
4. Cells majority
Obtained
the Nonpermissive Adult to a Permissive Substrate correlation between the distribuphagocytes and the localized
from
of the celis
Nitrocellulose are EDI+.
(5) Most
Implants of them
Placed also
in the
Lesioned
phagocytose
Latex
Brain beads
in vitro.
Magnification,
350x
(A); 480x
(B).
Changes 467
in the Nonpermissive
CNS Environment
the expression of growth-promoting molecules by the cells derived from the nitrocellulose implants and peritoneal macrophages. Although we cannot exclude the latter possibility, it appears to be unlikely because neurites were not seen to grow preferentially on top of the cells from the nitrocellulose implants, and the neurites did not show directional growth toward these cells when the majority of them were located away from the tissue section. Macrophages have also been shown to augment the production of nerve growth factor (NGF) by nonneuronal cells in the transected rat sciatic nerve (Heumann et al., 1987a, 1987b). Such an effect is unlikely to account for our results because NGF was added to all our DRG cultures, including the control cultures containing untreated, normal optic nerve. Predegenerated Adult Optic Nerve Sections Are a Permissive Substrate In another set of experiments we investigated whether astrocytic gliosis has an inhibitory effect on axonal growth. For these experiments we used adult rat optic nerves 4-6 months after transection. Such long posttransection survival times were used because of the slow removal of myelin debris in Wallerian degeneration in the CNS, which may be due to the poor recruitment of macrophages to regions distal to the site of the lesion (Perry and Gordon, 1988). These predegenerated nerves lacked intact myelin and were composed of an astrocytic “scar.” When unfixed cryostat sections of these nerves were cultured with E8-El0 chick DRGs, 45% of the ganglia grew neurites extensively onto these sections of degenerated optic nerve (Figure 2F; Table 1). Neurite growth on these sections occurred equally well near and distal to the site of transection. Small segments of these optic nerves that were prepared for electron microscopy contained very little myelin debris and were composed of densely packed astrocytic processes and their perikarya. It has been reported that such nerves contain fewer dark cells that resemble oligodendrocytes compared with normal nerves (Miller et al., 1986). These nerves represent the so called astrocytic scar tissue, which has been proposed in many studies to inhibit axonal regeneration in the mammalian CNS (reviewed by Reier et al., 1989). Many of these studies are open to other interpretations in the light of recent reports that inhibitory molecules are associated with CNS myelin (Caroni and Schwab, 1988a, 1988b). We now provide in vitro evidence that predegenerated optic nerves containing reactive astrocytes and devoid of intact myelin are an excellent substrate for axonal growth. These results suggest that significant inhibitory effects may not be associated with reactive astrocytes in the predegenerated optic nerve. Furthermore, the localized change from a nonpermissive to a permissive state near the lesion at 5 days posttransection becomes more widespread after several months, so as to include the entire length of the degenerated optic nerve.
Conclusions The results of these experiments indicate the following: -The lesioned CNS white matter consisting of astrocytic gliosis and devoid of myelin (4-6 months after lesioning) is capable of supporting neurite growth in vitro, thus suggesting that substantial inhibitory influences may not be present in regions of astrocytic gliosis. -The nonpermissive adult CNS white matter can be rendered permissive near the site of the lesion within 5 days after lesioning and may account for the sprouting seen in vivo soon after lesioning. -This transformation of the adult CNS white matter from a nonpermissive to a permissive state may be the result of the action of cells, such as macrophages, derived from the peripheral circulation. The failure of regeneration in the adult mammalian CNS may therefore be due, in part, to the inability of the sprouting axons to grow beyond the immediate vicinity of the lesion, as a result of the presence of inhibitory molecules such as those associated with CNS myelin. Many of the damaged neurons die during the time required for the removal of CNS,myelin from the degenerating CNS (Ramon y Cajal, 1928; Villegas-P&ez et al., 1988), and those neurons that survive may no longer be in a growth mode (Thanos and Vanselow, 1989). The identification of the substance(s) that is released by macrophages near the lesion and that can alter the nonpermissive nature of CNS tissue may allow one to attempt to modify the CNS glial environment in vivo at an early stage after CNS damage, when injured neurons may still be in a growth mode. Experimental Optic
Procedures
Nerve Transection
Under chloral hydrate anesthesia (42 mg/lOO g BW), the right optic nerve of female Sprague-Dawley rats (250 g) was transected intraorbitally or intracranially as described previously (Giftochristos and David, 1988).
Preparation
of Tissue Sections
Five days or 4-6 months were deeply anesthetized
cryostat
sections
after optic nerve transection, the rats and perfused with PBS. Longitudinal
(10 pm) of the unfixed
optic
nerves
were
picked
up on poly-clysine-coated glass coverslips. The tissue sections were then UV-irradiated for 15 min and washed extensively with Eagle’s minimum essential medium (Carbonetto et al., 1987).
Culturing
of PC12 Cells on Tissue Sections
PC12 cells were cultured in Petri dishes in RPM1 containing 5% fetal bovine serum, 10% horse serum, penicillin/streptomycin, and Fungizone. These cells were triturated, passed through a 23 gauge needle, and plated onto the frozen optic nerve sections at a density of about 2 x 104 cells per coverslip. Cultures were maintained in the above medium with NGF (25 rig/ml) at 37OC in a CO2 incubator. After 5 days, the cultures were fixed with 4% paraformaldehyde and labeled with a rabbit anti-glial fibrillary acidic protein antibody, followed by a goat anti-rabbit secdndary antibody conjugated to fluoresceine. Cultures were also labeled with 0.001% nuclear yellow.
DRC Explant DRGs
from
Cultures E8-10 chicks
were
placed
alongside
the frozen
sec-
tions on glass coverslips and cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum, MEM-vitamins, penicillin/streptomycin, and 25 rig/ml NCF. After overnight culture, the medium was changed to a serum-free, chemically defined medium (Bottenstein and Sato, 1978) containing 50 rig/ml NCF and 20 PM cytosine arabinoside (Carbonetto et al., 1987). After 5 days, the cultures were fixed with 4% paraformaldehyde in 0.1 M phosphate buffer (pH Z4) and stained with Coomassie blue. Preparation of Mononuclear Phagocytes from the Lesioned CNS and Treatment of Normal Rat Optic Nerve Sections Under chloral hydrate anesthesia (42 mg/lOO g BW), a small area of the skull over the parietal lobe was removed in 250 g female Sprague-Dawley rats. The dura was cut and retracted and a 2 mm long incision made in the cerebral cortex. A 2 x 1 mm piece of sterile nitrocellulose paper was inserted into this incision, and the wound was closed. Five days after implantation, the animals were sacrificed by cervical dislocation and the nitrocellulose implants were removed. The nitrocellulose implantswere rinsed in culture media and placed in wells containing coverslips with frozen sections of normal rat optic nerve in DMEM plus 10% fetal bovine serum. These cultures were maintained for 24 hr in a 5% CO2 incubator at 37X The nitrocellulose papers were removed, and the tissue sections were rinsed. Embryonic chick D~Gswere then plated as described above. In thecontrol group, the normal rat optic nerve sections were incubated for 24 hrwith pieces of nitrocellulose paper soaked in normal rat serum. Preparation of Activated Peritoneal Macrophages Sprague-Dawley rats were injected intraperitoneally with 5 ml of 3% thioglycolate. After 4 days, peritoneal cells were obtained by lavage with DMEM. The cells obtained were plated at a density of 2 x 106 cells per ml in Petri dishes. Unattached cells were removed after 18 hr, and the macrophages were activated by treatment with N-acetylmuramyl+alanyl-pisoglutamine (Sigma) (Phillips and Chedid, 1988) at 20 pgiml for 4 days. Immunocytochemistry Five days after an optic nerve transection, rats were deeply anesthetized with chloral hydrate and perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Longitudinal cryostat sections (10 pm) were picked up on gelatin-coated glass slides. These tissue sections were incubated, sequentially, with 3% ovalbumin (Sigma) overnight, with the monoclonal antibody ED1 (Serotec; 1:lOO) overnight, and with a goat anti-mouse secondary antibody conjugated to rhodamine (Cappel, 1: 100) for 1 hr. The tissue sections were then stained with 0.001% nuclear yellow for 30 5. Nitrocellulose implants were cultured overnight on poly+lysine-coated glass coverslips. Cells emerged from the nitrocellulose papers within a few minutes after being placed In culture. These cultures were fixed with ethanol-acetic acid (95:5) for 20 min at -4OC. The cultures were then incubated, sequentially for 30 min each, with the monoclonal antibody EDI, goat antimouse IgC conjugated to rhodamine, rabbit anti-GFAP, and goat anti-rabbit IgG conjugated to fluorecein. Some of these cultures were incubated with Latex beads for 2 hr at 37OC prior to fixation and immunocytochemistry. Acknowledgments
April
Aguayo, A. J. (1985). Axonai regeneration from Injured neurons in the adult mammalian central nervous system. In Synaptic Plasticity, C. W. Cotman, ed. (New York: Cullford Press), pp. 457-484. Benfey, M.. and Aguayo, A. 1. (1982). Extensive elongation axons from rat brain into peripheral nerve grafts. Nature 150-152. Bottenstein, J. E., and Sato, 6. H. (1978). Growth blastoma cell line In serum-free supplemented Natl. Acad. Sci. USA 76, 514-517.
20, 1990; revised
July 27, 1990.
of 296,
of a rat neuromedium. Proc.
Carbonetto, S., Evans, D., and Cochard, P. (1987). growth in culture on tissue substrata from central eral nervous systems. j. Neurosci. Z 610-620.
Nerve fibre and periph-
Caroni, P., and Schwab, M. E. (1988a). Two membrane protein fractions from iat central myelln with inhibitory properties ior neurite growth and fibroblast spreading. J. Cell Biol. 706, 1281-1288. Caroni, P., and Schwab, M. E. (1988b). Antibody against myellnassociated inhibitor of neurite growth neutralizes nonpermissive substrate properties of CNS white matter. Neuron I, 85-96. David, 5, and Aguayo, A. J. (1981). Axonal eiongation into PNS “bridges” after CNS injury In adult rats. Science 274, 931-933. Dijkstra, C. D., Doepp, E. A., Joling, P.. and Kraal, G. (1985). The heterogeneity of mononuclear phagocytes in iymphoid organs: distinct macrophage subpopulations in the rat recognized by monoclonal antibodies EDI, ED2 and ED3. Immunology 54, 589-599. Friedman, B., and Aguayo, A. I. (1985). Injured neurons In the olfactory bulb of adult rat form new axons along peripheral nerve grafts. J. Neurosci. 5, 1616-1625. Giftochristos, RI., and David, 5. (1988). Laminin and heparan su;phate proteoglycan in the lesloned adult mammaljan central nervous system and their possible relationship to axonal sprouting. J. Neurocytol. 17 385-397. Giulian, D., Chen, J., Ingeman, J. E., George, i. K., and Noponen, M. (1989). The role of mononuclear phagocytes in wound healing after traumatic injury to adult mammalian brain. J. Neurosci. 9, 4416-4429. Grafstein, B., and Ingoglia, of the optic nerve in adult Neural. 76, 318-330.
N. A. (1982). Intracranial transection mice: preliminary observations. Exp.
Heumann, R., Korsching, S., Bandtlow, C., and Thoenen, H. (1987a). Changes of nerve growth factor synthesis in nonneuronal cells in response to sciatic nerve transection. J. Cell Biol. 704, 1623-1632. Heumann, R., ILindholm, D., Bandtlow, C., Meyer, M., Radeke, M. J., Misko, T. P., Shooter, E., and Thoenen, H. (1987b). Differential regulation of mRNA encoding nerve growth factor and its receptor in rat sciatic nerve during development, degeneration, and regeneration: role of macrophages. Proc. Natl. Acad. Sci. USA 84, 8735-8730. Miller, R. H., Abney, E. R., David, S., ffrench-Constant, C., Lindsay, R., Patel, R., Stone, J., and Raff, M. C. (1986). Is reactivegliosis a property of a distinct subpopulation of astrocytes! J. Neurosci. 6. 22-29. Nathan, C. F. (1987). Invest. 79, 319-326.
Wewish to thank Dr. Peter Richardson for kindly providing NCF and Dr. Nigel Phillips for his advise on the preparation of peritoneal macrophages. This work was supported by grants from the Canadian MRC (MA 8723) and the Spinal Cord Research Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received
References
Secretory
products
of macrophages.
Perry, V. H., and Cordon, 5. (1988). Macrophages and in the nervous system. Trends Neurosci. 77, 273-277. Pesheva, P., Spiess, ? 765-1778.
E., and Schachner,
M. (1989).
J. Clin. microglia
J. Cell Biol.
Phillips, N. C., and Chedid, L. (1988). Muramyi peptides liposomes. In Liposomes as Drug Carriers, C. Gregoriadis, (New York: John Wiley & Sons Ltd.), pp. 243-259. Ramon y Cajal, S. (1928). Degeneration and Nervous System, R. M. May, trans. (London: Press).
709, and ed.
Regeneration of the Oxford University
Changes 469
in the Nonpermissive
CNS Environment
Reier, P. J., Eng, L. F., and Jakeman, L. (1989). Reactive astrocyte and axonal outgrowth tn the injured CNS. In Neural Regeneration and Transplantation, Frontiers of Clinical Neuroscience, Vol. 6, F. J. Seil, ed. (New York: Alan R. Liss, Inc.), pp. 183-209. Richardson, Axons from 264-265.
P. M., McCutnness, U. M., and Aguayo, CNS neurons regenerate into PNS grafts.
A. J. (1980). Nature 284,
Richardson, P. M., Issa, V. M. K., and Shemle, S. (1982). Regeneration and retrograde degeneration of axons in the adult rat optic nerve. J. Neurocytol. 11, 949-966. Savio, T., and Schwab, M. E. (1989). Rat CNS whne matter, gray matter, is nonpermissive for neuronal cell adhesion ber growth. J. Neurosci. 9, 1126-1133.
but not and fi-
Schwab, M. E., and Caroni, P. (1988). Oligodendrocytes and CNS myelin are nonpermissive substrates for neurite growth and ft. broblast spreading in vitro. J. Neurosci. 8, 2381-2393. Schnell, L., and Schwab, M. E. (1990). Axonal regeneration in the rat spinal cord produced by an antlbody against myelin-associated neurite growth inhibitors. Nature 343, 269-272. So, K.-F., and Aguayo, A. J. (1985). Lengthy regrowth of cut axons from ganglion cells after peripheral nerve transplantation into the retina of adult rats. Brain Res. 328, 349-354. Stall, C., Trapp, B. D., and Griffin, J. W. (1989). Macrophage function during Wallerian degeneration of rat optic nerve: clearance of degenerating myelin and la expresslon. J. Neurosci. 9, 23272335. Thanos, S., and Vanselow, J. (1989). Adult retinal gangllon cells retain the ability to regenerate their axons up to several weeks after axotomy. J. Neurosci. Res. 22, 144-149. Villegas-Perez, M., Vidal-Sanz, M., Bray, G. M., and Aguayo, A. J. (1986). Influence of peripheral nerve grafts on the survival and regrowth of axotomized retlnal ganglion cells in adult rats. J. Neurosci. 8, 265-280.
Vol. 5, 463-469,
October,
1990, Copyright
0
1990 by Cell Press
Macrophages Can Modify the Nonpermissive Nature of the Adult Mammalian Central Nervous System S. David, C. Bouchard, 0. Tsatas, and N. Giftochristos Centre for Research in Neuroscience McGill University Montreal General Hospital Research Montreal, Quebec Canada, H3G IA4
Institute
Summary Although astrocytic gliosis has been linked to failure of axonal regeneration in the adult mammalian CNS, its role is not fully established. We used an in vitro assay to investigate the role of reactive astrocytes and macrophages in influencing axonal growth in the lesioned adult rat optic nerve. Soon after optic nerve transection, the nonpermissive nature of the optic nerve is altered to a permissive state near the lesion. This may account for injury-induced axonal sprouting and may contribute to the failure of these sprouts to elongate beyond the site of the lesion in vivo. We provide evidence that this lesion-induced change in the axonal growth-promoting properties of the CNS near the lesion may be produced by mononuclear phagocytes. In addition, several months after optic nerve transection, the degenerated nerves, which consist mainly of astrocytes and lack myelin, i.e., astrocytic “scar” tissue, are a good substrate for neurite growth. Taken together, these results suggest that in this in vitro system, substantial inhibitory effects are not associated with regions of astrocytic gliosis and that the nonpermissive nature of the CNS white matter can be modified by macrophages. Introduction Neurons in the adult mammalian CNS have the intrinsic ability to regenerate their axons for long distances under appropriate experimental conditions (Richardson et al., 1980; David and Aguayo, 1981; Benfey and Aguayo, 1982; Friedman and Aguayo, 1985; So and Aguayo, 1985). The failure of CNS neurons to regenerate axons through adult CNS tissue in vivo appears to be a function of the CNS glial environment (Aguayo, 1985). Many of the earlier studies have suggested that astrocytes in the mammalian CNS might hinder or inhibit axonal regeneration (Reier et al., 1989). However, many of these results are open to other interpretations, and therefore the role of astrocytic gliosis in the failure of regeneration is not yet fully established. Astrocytic gliosis has been shown to be stimulated by macrophages (Giulian et al., 1989). It has been reported recently that a minor protein component of mammalian CNS myelin has a marked inhibitory effect on fibroblast spreading and neurite extension in vitro (Caroni and Schwab, 1988a, 1988b; Schwab and Caroni, 1988) and possibly in vivo (Schnell and Schwab,
1990). Other inhibitory molecules also produced by oligodendrocytes have been identified (Pesheva et al., 1989). In this study, we have used an in vitro assay to determine whether astrocytic gliosis also contributes to the failure of axonal growth and to examine the effects of macrophages on the nonpermissive properties of the adult mammalian CNS white matter. The rat optic nerve is a convenient region of the CNS to study the role of glia or glial components in regeneration, since it lacks neuronal cell bodies and is easily accessible for experimental manipulations. Injuries to the CNS results in axonal sprouting that is not sustained for long (Ramon y Cajal, 1928). We and others have shown that retinal ganglion cell axons undergo sprouting soon after an intracranial optic nerve transection (Richardson et al., 1982; Crafstein and Ingoglia, 1982; Giftochristos and David, 1988). It is not known whether the injury-induced axonal sprouting is accompanied by changes in the glial environment. In the present study, unfixed cryostat sections of transected adult rat optic nerve were used as a tissue culture substrate to study the changes in the ability of the injured CNS tissue to support axonal growth. This approach has been used previously to examine the differences between embryonic and adult CNS tissue (Carbonetto et al., 1987) and the differences between the gray and white matter of the adult mammalian CNS (Savio and Schwab, 1989). Results
and
Discussion
Changes in the Nonpermissive Properties of Adult Optic Nerve 5 Days after Transection Cell Aftachmenf Unfixed, longitudinal cryostat sections of transected adult rat optic nerve were obtained 5 days after transection and picked up on poly+-lysine-coated, round glass coverslips (15 mm). These tissue sections were prepared for tissue culture (Carbonetto et al., 1987) and seeded with PC12 cells. A significantly greater number of PC12 cells attached to the tissue sections near the site of the lesion compared with regions distal to the lesion (Figure 1). The PC12 cells attached to regions that contained astrocytes as determined immunohistochemically (Figures IA and IB). These results suggest that CNS tissue near the site of a lesion has greater adhesive properties than regions distal to a lesion. Neurite Growth To determine the ability of optic nerve tissue near the site of a lesion to support neurite growth (which we were unable to visualize for technical reasons using PC12 cells), we cultured unfixed cryostat sections of adult rat optic nerves obtained 5 days after transection with explants of embryonic days 8-10 (E8-EIO) chick dorsal root ganglia (DRGs). After 5 days in vitro, 30% of the DRG explants seeded near the site of the
Figure
I
0’
E
F
1. PC12 Cells
Plated
on
Unfixed
Cryostat
Sections
of Adult
Rat Optic
Nerve
5 Days after
an Optic
N
Nerve
i TA
TN
Transection
Sections have been double-labeled with anti-CFAP (A and C) and nuclear yellow (B and D). (A) Hypertrophied astrocytes near the site of transection. (B) Same section as in (A), showing the attachment of PC12 cells visualized with nuclear yellow. Note that a large number of pC12 cells are attached predominantly to regions that are CFAP+. (C)An area of a transected optic nerve distal to the site of transection. The anti-GFAP labeling of the astrocytic processes is finer and less intense compared with that of regions nearer the lesion. (D) Fewer pC12 cells are attached to these regions distal to the site of transection. (E) Camera lucida drawing showing the distribution of PC12 cells attached to a slngie cryostat section of a transected optic nerve 5 days after transection. The sectlon has been divided into two parts: the hatched area represents the portion of the nerve distal to the lesion (as in [Cl), in which the anti-CFAP labeling is finer and less Intense compared wtth regions of the nerve near the iesion, which contain hypertrophied astrocytes (as in [A]). (F) Histogram of the numbers of PC12 cells atiached to cryostat sections of the normal optic nerve (N), the transected optic nerve distal to the site of lesion (TA) 5 days after transection (hatched area in [El), and the region of the transected nerve near the lesion (TN). Note that a significantly greater number of PC12 cells attach to regions near the lesion (TN). Mean i SD of three experiments; The asterisk indicates P < 0.01; significance levels between the values of the transected and normal nerves ‘were estimated using Student’s t-test. Magnification: 144x (all panels).
optic nerve transection showed extensive neurite growth onto the tissue sections, whereas none of those plated distal to the lesion showed such growth (Figures 2C and 2D; Table 1). The latter growth was similar to that on sections of normal adult rat optic nerve (Carbonetto et al., 1987) (Figures 2A and 2B; Table 1). These results indicate that within a few days after lesioning, the normally nonpermissive CNS white matter is able to support cell attachment and axonal growth near the site of the lesion. There are several possible explanations for this
change in the growth-promoting properties of the injured CNS. One possibility is that reactive astrocytes near the lesion may be induced to express diffusable growth factors and/or neurite growth-promoting adhesion molecules. Since the glial cells in the frozen sections are not viable in our present assay and the tissue sections are washed extensively prior to culturing, it is unlikely that growth factors which may be produced in vivo by astrocytes are responsible for the observed neurite growth during the 5 days in culture. We have also shown previously that the expression of
Changes 465
in the Nonpermissive
CNS Environment
Figure 2. Neurite Growth from EB-El0 Chick DRGs Plated beside Cryostat Sections of Unfixed Optic Nerve The DRC is located at the top of each micrograph. (A) and (B) show the lack of neurite growth onto sections of normal rat optic nerve. In (A) the DRC is located away from the section; in LB) it lies against the optic nerve section.(C) In contrast to the normal nerve, neurites grow extensively onto sections of the optic nerve near the site of a transection, 5 days after lesioning. (D) However, neurite growth on the transected optic nerve, distal to the site of transection, is similar to that on the normal optic nerve. (E) Neurites grow extensively on sections of the normal adult rat optic nerve treated overnight with a nitrocellulose implant obtained 5 days after implanting into a cerebral cortical lesion. (F) Extensive neurite growth on to a section of optic nerve 4 months after transection. Magnification: 27x (A and B); 42x (C-F).
laminin and heparan sulphate proteoglycan are not correlated with the sprouting response in the injured adult rat optic nerve (Giftochristos and David, 1988). It is not known whether any of the other adhesion molecules are expressed differently by reactive astrocytes in close proximity to the lesion. Another possibility is that inhibitory molecules (Caroni and Schwab, 1988a; Pesheva et al., 1989), such as those associated with CNS myelin, may have been inactivated. It is unlikely that these effects are due to the absence of myelin, since CNS myelin is known to take several weeks or months to be cleared from the le-
Table
1. The
Percentage Normal
Growth Extensive Moderate Minimal
(36) 0 16.5 83.5
of Embryonic
Chick
DRGs
Extending
sioned CNS (Perry and Gordon, 1988; Stall et al., 1989). Furthermore, PC12 cells were found to be attached to regions near the lesion that stained positively with a polyclonal antiserum against myelin basic protein (data not shown). It is possible, however, that the inhibitory proteins associated with myelin are inactivated near the site of the lesion soon after lesioning. This would account not only for the axonal sprouting that occurs soon after the induction of CNS lesions, but also for the failure of these sprouts to elongate for long distances away from the lesion site. These inhibitory molecules may be inactivated by
Neurites
onto
Cryostat
Sections
5 Days Trans. Near Lesion t23Ja
5 Days Trans. Distal to Lesion (23)’
4-6 Months Trans. (29jb
Normal ON, NC Implant (20)’
Normal NC/RS Control
30 48 22
0 13 87
45 27.5 27.5
65 15 20
4.5 19.5 76
of Unfixed
Optic
Nerve
ON, (2V
Normal M&M 46 19 35
ON, (28)’
Normal ON, MDP Control
(34)’
8.5 23.5 68
Growth onto the sections of up to 20 neurites was considered minimal growth, between 20 and 50 was considered moderate and more than 50 was considered extensive growth. Numbers in parentheses indicate the number of DRCs examined. a Five days after optic nerve transection. 6 Four to 6 months after optic nerve transection. c Normal optic nerve sections incubated for 24 hr with nitrocellulose implants obtained 5 days after implanting into cerebral d Normal optic nerve sections incubated with nitrocellulose paper soaked in rat serum. e Normal optic nerve sections incubated for 48 hr with activated peritoneal macrophage-conditioned medium. ‘ Normal optic nerve sections incubated for 48 hr with N-acetylmuramyl-t-alanyl-o-isoglutamine.
growth,
cortex.
Figure 3. Longitudinal Optic Nerve 5 Days
Cryostat Section after Transection
(A) EDl’cells are found only near yellow labeling of the same tissue
of a Transected
Adult
change in the nonpermissive properties of the optic nerve 5 days after transection, we examined whether mononuclear phagocytes might be able to modify the nonpermissive properties of the adult rat optic nerve. This was done by treating cryostat sections of unfixed, normal rat optic nerve with mononuclear phagocytes from the iesioned CNS. These cells were obtained by implanting small pieces of nitrocellulose paper into the cerebral cortex of adult rats. After 5 days the nitrocellulose paper was removed and cultured for 24 hr with unfixed frozen sections of normal adult rat optic nerves. Cells emerged from the nitrocellulose filters within a few minutes after being placed in culture. The majority of these cells appeared to be mononuclear phagocytes, because they labeled positively with the monoclonal antibody ED1 (92.5%) and phagocytosed Latex beads in vitro (80.5%) (Figure 4). These values were obtained from counting a total of 3325 cells in two separate experiments. Qnly 2% of these cells were glial fibrillary acidic protein-positive (CFAP+) astrocytes, and all of these were EDI-. Norma! adult rat optic nerve sections incubated overnight with the nitrocellulose implants were capable of supporting excellent neurite growth from embryonic chick DRGs (Figure 2E; Table 1). Similar results were also obtained with medium conditioned by activated peritoneal macrophages (Table ?), thus suggesting that substances released by macrophages may be responsible for this effect. Some nonneuronal cells were associated with the DRG explants despite the addition of cytosine arabinoside. However, the presence of these nonneuronal cells was not sufficient to ensure the growth of neurites onto sections of untreated, normal optic nerve. For the reasons cited earlier, it is unlikely that the permissive properties of the treated ncrmal nerves are due to the increased expression of diffusable growth factors or adhesion molecules by the glial cells, since there are no viable cells in the tissue sections. This effect may therefore reflect the inactivation of inhibitory molecules in the optic nerve sections, or
Rat
the site of lesion. (B) Nuclear section. Magnification: 80x.
substances released by macrophages (Nathan, 1987). We therefore examined the distribution of mononuclear phagocytes in the transected adult rat optic nerve 5 days after a transection. Immunohistochemical Localization of Mononuclear Phagocytes The pattern of distribution of mononuclear phagocytes in the adult rat optic nerve 5 days after nerve transection was determined in longitudinal cryostat sections. Mononuclear phagocytes, which were identified with the monoclonal antibody ED1 (Dijkstra et al., 1985), were located only near the lesion site (Figure 3). This area corresponds to the region that supports cell attachment and neurite growth in vitro and axonal sprouting in vivo. Macrophages Change Optic Nerve Sections Because of the close tion of mononuclear
Figure (A) The
4. Cells majority
Obtained
the Nonpermissive Adult to a Permissive Substrate correlation between the distribuphagocytes and the localized
from
of the celis
Nitrocellulose are EDI+.
(5) Most
Implants of them
Placed also
in the
Lesioned
phagocytose
Latex
Brain beads
in vitro.
Magnification,
350x
(A); 480x
(B).
Changes 467
in the Nonpermissive
CNS Environment
the expression of growth-promoting molecules by the cells derived from the nitrocellulose implants and peritoneal macrophages. Although we cannot exclude the latter possibility, it appears to be unlikely because neurites were not seen to grow preferentially on top of the cells from the nitrocellulose implants, and the neurites did not show directional growth toward these cells when the majority of them were located away from the tissue section. Macrophages have also been shown to augment the production of nerve growth factor (NGF) by nonneuronal cells in the transected rat sciatic nerve (Heumann et al., 1987a, 1987b). Such an effect is unlikely to account for our results because NGF was added to all our DRG cultures, including the control cultures containing untreated, normal optic nerve. Predegenerated Adult Optic Nerve Sections Are a Permissive Substrate In another set of experiments we investigated whether astrocytic gliosis has an inhibitory effect on axonal growth. For these experiments we used adult rat optic nerves 4-6 months after transection. Such long posttransection survival times were used because of the slow removal of myelin debris in Wallerian degeneration in the CNS, which may be due to the poor recruitment of macrophages to regions distal to the site of the lesion (Perry and Gordon, 1988). These predegenerated nerves lacked intact myelin and were composed of an astrocytic “scar.” When unfixed cryostat sections of these nerves were cultured with E8-El0 chick DRGs, 45% of the ganglia grew neurites extensively onto these sections of degenerated optic nerve (Figure 2F; Table 1). Neurite growth on these sections occurred equally well near and distal to the site of transection. Small segments of these optic nerves that were prepared for electron microscopy contained very little myelin debris and were composed of densely packed astrocytic processes and their perikarya. It has been reported that such nerves contain fewer dark cells that resemble oligodendrocytes compared with normal nerves (Miller et al., 1986). These nerves represent the so called astrocytic scar tissue, which has been proposed in many studies to inhibit axonal regeneration in the mammalian CNS (reviewed by Reier et al., 1989). Many of these studies are open to other interpretations in the light of recent reports that inhibitory molecules are associated with CNS myelin (Caroni and Schwab, 1988a, 1988b). We now provide in vitro evidence that predegenerated optic nerves containing reactive astrocytes and devoid of intact myelin are an excellent substrate for axonal growth. These results suggest that significant inhibitory effects may not be associated with reactive astrocytes in the predegenerated optic nerve. Furthermore, the localized change from a nonpermissive to a permissive state near the lesion at 5 days posttransection becomes more widespread after several months, so as to include the entire length of the degenerated optic nerve.
Conclusions The results of these experiments indicate the following: -The lesioned CNS white matter consisting of astrocytic gliosis and devoid of myelin (4-6 months after lesioning) is capable of supporting neurite growth in vitro, thus suggesting that substantial inhibitory influences may not be present in regions of astrocytic gliosis. -The nonpermissive adult CNS white matter can be rendered permissive near the site of the lesion within 5 days after lesioning and may account for the sprouting seen in vivo soon after lesioning. -This transformation of the adult CNS white matter from a nonpermissive to a permissive state may be the result of the action of cells, such as macrophages, derived from the peripheral circulation. The failure of regeneration in the adult mammalian CNS may therefore be due, in part, to the inability of the sprouting axons to grow beyond the immediate vicinity of the lesion, as a result of the presence of inhibitory molecules such as those associated with CNS myelin. Many of the damaged neurons die during the time required for the removal of CNS,myelin from the degenerating CNS (Ramon y Cajal, 1928; Villegas-P&ez et al., 1988), and those neurons that survive may no longer be in a growth mode (Thanos and Vanselow, 1989). The identification of the substance(s) that is released by macrophages near the lesion and that can alter the nonpermissive nature of CNS tissue may allow one to attempt to modify the CNS glial environment in vivo at an early stage after CNS damage, when injured neurons may still be in a growth mode. Experimental Optic
Procedures
Nerve Transection
Under chloral hydrate anesthesia (42 mg/lOO g BW), the right optic nerve of female Sprague-Dawley rats (250 g) was transected intraorbitally or intracranially as described previously (Giftochristos and David, 1988).
Preparation
of Tissue Sections
Five days or 4-6 months were deeply anesthetized
cryostat
sections
after optic nerve transection, the rats and perfused with PBS. Longitudinal
(10 pm) of the unfixed
optic
nerves
were
picked
up on poly-clysine-coated glass coverslips. The tissue sections were then UV-irradiated for 15 min and washed extensively with Eagle’s minimum essential medium (Carbonetto et al., 1987).
Culturing
of PC12 Cells on Tissue Sections
PC12 cells were cultured in Petri dishes in RPM1 containing 5% fetal bovine serum, 10% horse serum, penicillin/streptomycin, and Fungizone. These cells were triturated, passed through a 23 gauge needle, and plated onto the frozen optic nerve sections at a density of about 2 x 104 cells per coverslip. Cultures were maintained in the above medium with NGF (25 rig/ml) at 37OC in a CO2 incubator. After 5 days, the cultures were fixed with 4% paraformaldehyde and labeled with a rabbit anti-glial fibrillary acidic protein antibody, followed by a goat anti-rabbit secdndary antibody conjugated to fluoresceine. Cultures were also labeled with 0.001% nuclear yellow.
DRC Explant DRGs
from
Cultures E8-10 chicks
were
placed
alongside
the frozen
sec-
tions on glass coverslips and cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum, MEM-vitamins, penicillin/streptomycin, and 25 rig/ml NCF. After overnight culture, the medium was changed to a serum-free, chemically defined medium (Bottenstein and Sato, 1978) containing 50 rig/ml NCF and 20 PM cytosine arabinoside (Carbonetto et al., 1987). After 5 days, the cultures were fixed with 4% paraformaldehyde in 0.1 M phosphate buffer (pH Z4) and stained with Coomassie blue. Preparation of Mononuclear Phagocytes from the Lesioned CNS and Treatment of Normal Rat Optic Nerve Sections Under chloral hydrate anesthesia (42 mg/lOO g BW), a small area of the skull over the parietal lobe was removed in 250 g female Sprague-Dawley rats. The dura was cut and retracted and a 2 mm long incision made in the cerebral cortex. A 2 x 1 mm piece of sterile nitrocellulose paper was inserted into this incision, and the wound was closed. Five days after implantation, the animals were sacrificed by cervical dislocation and the nitrocellulose implants were removed. The nitrocellulose implantswere rinsed in culture media and placed in wells containing coverslips with frozen sections of normal rat optic nerve in DMEM plus 10% fetal bovine serum. These cultures were maintained for 24 hr in a 5% CO2 incubator at 37X The nitrocellulose papers were removed, and the tissue sections were rinsed. Embryonic chick D~Gswere then plated as described above. In thecontrol group, the normal rat optic nerve sections were incubated for 24 hrwith pieces of nitrocellulose paper soaked in normal rat serum. Preparation of Activated Peritoneal Macrophages Sprague-Dawley rats were injected intraperitoneally with 5 ml of 3% thioglycolate. After 4 days, peritoneal cells were obtained by lavage with DMEM. The cells obtained were plated at a density of 2 x 106 cells per ml in Petri dishes. Unattached cells were removed after 18 hr, and the macrophages were activated by treatment with N-acetylmuramyl+alanyl-pisoglutamine (Sigma) (Phillips and Chedid, 1988) at 20 pgiml for 4 days. Immunocytochemistry Five days after an optic nerve transection, rats were deeply anesthetized with chloral hydrate and perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Longitudinal cryostat sections (10 pm) were picked up on gelatin-coated glass slides. These tissue sections were incubated, sequentially, with 3% ovalbumin (Sigma) overnight, with the monoclonal antibody ED1 (Serotec; 1:lOO) overnight, and with a goat anti-mouse secondary antibody conjugated to rhodamine (Cappel, 1: 100) for 1 hr. The tissue sections were then stained with 0.001% nuclear yellow for 30 5. Nitrocellulose implants were cultured overnight on poly+lysine-coated glass coverslips. Cells emerged from the nitrocellulose papers within a few minutes after being placed In culture. These cultures were fixed with ethanol-acetic acid (95:5) for 20 min at -4OC. The cultures were then incubated, sequentially for 30 min each, with the monoclonal antibody EDI, goat antimouse IgC conjugated to rhodamine, rabbit anti-GFAP, and goat anti-rabbit IgG conjugated to fluorecein. Some of these cultures were incubated with Latex beads for 2 hr at 37OC prior to fixation and immunocytochemistry. Acknowledgments
April
Aguayo, A. J. (1985). Axonai regeneration from Injured neurons in the adult mammalian central nervous system. In Synaptic Plasticity, C. W. Cotman, ed. (New York: Cullford Press), pp. 457-484. Benfey, M.. and Aguayo, A. 1. (1982). Extensive elongation axons from rat brain into peripheral nerve grafts. Nature 150-152. Bottenstein, J. E., and Sato, 6. H. (1978). Growth blastoma cell line In serum-free supplemented Natl. Acad. Sci. USA 76, 514-517.
20, 1990; revised
July 27, 1990.
of 296,
of a rat neuromedium. Proc.
Carbonetto, S., Evans, D., and Cochard, P. (1987). growth in culture on tissue substrata from central eral nervous systems. j. Neurosci. Z 610-620.
Nerve fibre and periph-
Caroni, P., and Schwab, M. E. (1988a). Two membrane protein fractions from iat central myelln with inhibitory properties ior neurite growth and fibroblast spreading. J. Cell Biol. 706, 1281-1288. Caroni, P., and Schwab, M. E. (1988b). Antibody against myellnassociated inhibitor of neurite growth neutralizes nonpermissive substrate properties of CNS white matter. Neuron I, 85-96. David, 5, and Aguayo, A. J. (1981). Axonal eiongation into PNS “bridges” after CNS injury In adult rats. Science 274, 931-933. Dijkstra, C. D., Doepp, E. A., Joling, P.. and Kraal, G. (1985). The heterogeneity of mononuclear phagocytes in iymphoid organs: distinct macrophage subpopulations in the rat recognized by monoclonal antibodies EDI, ED2 and ED3. Immunology 54, 589-599. Friedman, B., and Aguayo, A. I. (1985). Injured neurons In the olfactory bulb of adult rat form new axons along peripheral nerve grafts. J. Neurosci. 5, 1616-1625. Giftochristos, RI., and David, 5. (1988). Laminin and heparan su;phate proteoglycan in the lesloned adult mammaljan central nervous system and their possible relationship to axonal sprouting. J. Neurocytol. 17 385-397. Giulian, D., Chen, J., Ingeman, J. E., George, i. K., and Noponen, M. (1989). The role of mononuclear phagocytes in wound healing after traumatic injury to adult mammalian brain. J. Neurosci. 9, 4416-4429. Grafstein, B., and Ingoglia, of the optic nerve in adult Neural. 76, 318-330.
N. A. (1982). Intracranial transection mice: preliminary observations. Exp.
Heumann, R., Korsching, S., Bandtlow, C., and Thoenen, H. (1987a). Changes of nerve growth factor synthesis in nonneuronal cells in response to sciatic nerve transection. J. Cell Biol. 704, 1623-1632. Heumann, R., ILindholm, D., Bandtlow, C., Meyer, M., Radeke, M. J., Misko, T. P., Shooter, E., and Thoenen, H. (1987b). Differential regulation of mRNA encoding nerve growth factor and its receptor in rat sciatic nerve during development, degeneration, and regeneration: role of macrophages. Proc. Natl. Acad. Sci. USA 84, 8735-8730. Miller, R. H., Abney, E. R., David, S., ffrench-Constant, C., Lindsay, R., Patel, R., Stone, J., and Raff, M. C. (1986). Is reactivegliosis a property of a distinct subpopulation of astrocytes! J. Neurosci. 6. 22-29. Nathan, C. F. (1987). Invest. 79, 319-326.
Wewish to thank Dr. Peter Richardson for kindly providing NCF and Dr. Nigel Phillips for his advise on the preparation of peritoneal macrophages. This work was supported by grants from the Canadian MRC (MA 8723) and the Spinal Cord Research Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received
References
Secretory
products
of macrophages.
Perry, V. H., and Cordon, 5. (1988). Macrophages and in the nervous system. Trends Neurosci. 77, 273-277. Pesheva, P., Spiess, ? 765-1778.
E., and Schachner,
M. (1989).
J. Clin. microglia
J. Cell Biol.
Phillips, N. C., and Chedid, L. (1988). Muramyi peptides liposomes. In Liposomes as Drug Carriers, C. Gregoriadis, (New York: John Wiley & Sons Ltd.), pp. 243-259. Ramon y Cajal, S. (1928). Degeneration and Nervous System, R. M. May, trans. (London: Press).
709, and ed.
Regeneration of the Oxford University
Changes 469
in the Nonpermissive
CNS Environment
Reier, P. J., Eng, L. F., and Jakeman, L. (1989). Reactive astrocyte and axonal outgrowth tn the injured CNS. In Neural Regeneration and Transplantation, Frontiers of Clinical Neuroscience, Vol. 6, F. J. Seil, ed. (New York: Alan R. Liss, Inc.), pp. 183-209. Richardson, Axons from 264-265.
P. M., McCutnness, U. M., and Aguayo, CNS neurons regenerate into PNS grafts.
A. J. (1980). Nature 284,
Richardson, P. M., Issa, V. M. K., and Shemle, S. (1982). Regeneration and retrograde degeneration of axons in the adult rat optic nerve. J. Neurocytol. 11, 949-966. Savio, T., and Schwab, M. E. (1989). Rat CNS whne matter, gray matter, is nonpermissive for neuronal cell adhesion ber growth. J. Neurosci. 9, 1126-1133.
but not and fi-
Schwab, M. E., and Caroni, P. (1988). Oligodendrocytes and CNS myelin are nonpermissive substrates for neurite growth and ft. broblast spreading in vitro. J. Neurosci. 8, 2381-2393. Schnell, L., and Schwab, M. E. (1990). Axonal regeneration in the rat spinal cord produced by an antlbody against myelin-associated neurite growth inhibitors. Nature 343, 269-272. So, K.-F., and Aguayo, A. J. (1985). Lengthy regrowth of cut axons from ganglion cells after peripheral nerve transplantation into the retina of adult rats. Brain Res. 328, 349-354. Stall, C., Trapp, B. D., and Griffin, J. W. (1989). Macrophage function during Wallerian degeneration of rat optic nerve: clearance of degenerating myelin and la expresslon. J. Neurosci. 9, 23272335. Thanos, S., and Vanselow, J. (1989). Adult retinal gangllon cells retain the ability to regenerate their axons up to several weeks after axotomy. J. Neurosci. Res. 22, 144-149. Villegas-Perez, M., Vidal-Sanz, M., Bray, G. M., and Aguayo, A. J. (1986). Influence of peripheral nerve grafts on the survival and regrowth of axotomized retlnal ganglion cells in adult rats. J. Neurosci. 8, 265-280.
