GLIA 3:293-300 (1990)
Growth of Adult Rat Retinal Ganglion Cell Neurites on Astrocytes M. BAEHR AND R.P. BUNGE The Miami Project to Cure Paralysis, University o f M i a m i School of Medicine, Miami, Florida 33136
KEY WORDS
Central nervous system, Glia, Regeneration, In vitro
ABSTRACT Astrocytes, as well as Schwann cells (SC),can provide suitable substrata for embryonic neurites during development, but their abilities to support adult regenerating neurites have not been directly compared. The aim of the present study was to determine the ability of astrocytes to promote adult rat retinal ganglion cell (RGC) regeneration in vitro and to compare this to previously determined growth on the surface of Schwann cells. We prepared Type I astrocytes (Raff et al: J . Neurosci. 3:1289-1300, 1983)from perinatal rats. These were subcultured and maintained in either a serum-free medium for at least 2 weeks (stellate astrocytes with little immunoreactivity for laminin) or in serum containing medium for 7 to 10 days (flat and polygonal astrocytes with immunoreactivity for laminin). Stellate astrocytes might therefore represent mature astrocytes in vivo (Ard and Bunge: J . Neurosci. 8:2844-2858,1988), while flat astrocytes might resemble immature brain astrocytes (Liesi et al: J . Cell Biol. 96:920-924, 1983). Adult RGC survival and axonal regrowth on these glia populations was compared to that observed on different SC populations, as previously reported (Baehr and Bunge: Exp. Neurol. 106:27-40, 1989). Both astrocyte populations (either flat or stellate astrocytes) did not enhance RGC survival. Stellate astrocytes were less effective in supporting RGC axon regeneration than flat astrocytes. When these date were compared to RGC survival and axon growth on SC (Baehr and Bunge: Exp.Neurol. 106:2740,1989)only “acitvated mature SC populations were superior to astrocytes in enhancing RGC survival and neurite regrowth. These results suggest 1)that astrocytes and “immature”SC are similar in their ability to support RGC survival; 2) “activated” mature SC populations are significantly better than astrocytes and “immature” SC in enhancing RGC survival and neurite growth; 3) stellate (“mature”) astrocytes, although permissive for regrowing axons, are not a favorable substrate for regenerating adult RGC neurites, nor do they effectively support RGC survival.
INTRODUCTION One of the major problems in studies of mammalian central nervous system (CNS) regeneration is to define the changes that occur in glial populations as the milieu of the CNS changes from being supportive to being inhibitory to neurite growth. In the developing CNS of mammals, outgrowing neurites interact with surrounding glia which provide substrata for growth and guidance for developing axons (Cohen et al., 1987; Fallon, 1985; Liesi and Silver, 1988; Noble et al., 1984; Sidman and Rakic, 1973; Silver and Rutishauser, 1984; Tomaselli et al., 1988). In contrast, mature CNS glia do not 0 1 9 9 0 Wiley-Liss, Inc.
support regeneration and may even be inhibitory for axonal regrowth (Cajal, 1968; Kalderon, 1988; Liuzzi and Lasek, 1987; Reier et al., 1983; Schwab and Caroni, 1988). It has been demonstrated that the adult optic Received November 28,1989; accepted February 13,1990 Address reprint requests to Dr. Mathias Baehr a t his permanent address M. Baehr’s permanent address is Max-Planck Institut fur Entwicklungsbiolopie, Spemannstr. 35/I, 7400 Tubingen, FRG Acknowledgments: The authors thank Dr. Naomi Kleitman for helpful discussions and comments on the paper and Susan Mantia and Charlaine Rowlette for typing the manuscript. Dr. D.D.M. O’Leary generously provided surgical equipment and facilities. This work was supported by the Max-Planck Gesellschaft (M. Baehr) and NIH grant NS09923 (R.P.Bungel.
294
BAEHR AND BUNGE
nerve is not permissive for regenerating neurites in vivo (Bray et al., 1987;Cajal, 1968;Giftochristos and David, 1988) and in vitro (Schwab and Thoenen, 1985). Recently a similar inhibition of neurite regrowth could be observed by confronting regenerating axons with mature oligodendrocytes or components of adult rat CNS myelin (Caroni and Schwab, 1988;Schwab and Caroni, 1988).Therefore, regrowth might be inhibited by mature oligodendrocytesand myelin present at lesion sites in the adult CNS, even if some adult CNS neurons retain their ability to regenerate axons. On the other hand, after lesioning the adult CNS, so-called “reactive” astrocytes can be observed, which may form a fibrous gliosis, thus preventing neurite regeneration (for review see Liuzzi and Lasek, 1987; Reier et al., 1983).Recently a classification of astrocytes into two subpopulations (Type I and 11) has been described (Raff et al., 1983)and it has been suggested that Type I astrocytes which support axon growth during development might be responsible for the formation of the glial scars in adult animals which are generally considered to be inhibitory for axonal regeneration (Miller et al., 1986). We have recently shown that adult rat RGC, defined by prelabeling in vivo from RGC targets and axotomized prior to explanation, survive in tissue culture and regrow axons on defined cellular substrata (Baehr et al., 1988, 1989). Within 48 h after explanation, neurite regrowth could be observed on polylysine plus laminin but not on collagen substrata (Baehr et al., 1988).This system makes it possible to examine interactions of regenerating adult CNS neurites with characterized glia populations in vitro. In the present study we have examined RGC survival and neurite regrowth on two forms of Type I astrocytes cultured on ammoniated collagen which correspond to “immature” astrocytes found during development (flat astrocytes with immunoreactivity for laminin) or to “mature” astrocytes (stellate astrocytes) of mature brain in vivo (Ard and Bunge, 1988;Liesi, 1985;Liesi et al., 1983;Liesi and Silver, 1988).RGC regeneration on these CNS glia populations was compared t o that observed on different SC populations, which are known to support neurite outgrowth and regeneration in vivo (for review see Bray et al., 1981)and in vitro (Baehr and Bunge, 1989;Bunge, 1987;Bunge et al., 1982;Hopkins et al., 1989).In these earlier studies (Baehr and Bunge, 1989),“immature” SC were seeded on purified dorsal root ganglia (DRG)neurites and cultured in serum and ascorbate containing media. This allows SC to repopulate the DRG and t o differentiate in contact with axons (Bunge et al., 1982;Eldridge et al., 1987,1989).These cultures were maintained until SC had started to ensheathe and myelinate DRG axons. The DRG neurons were then removed, which causes the neurites to degenerate, providing a model of Wallerian degeneration in vitro (for review, see Bunge, 1987).Adult retinal explants cultured on these SC monolayers are confronted with SC populations comparable to those found in sciatic nerve grafts. This serves as an in vitro model for the
adult sciatic nerve in vivo (for review, see Bray et al., 1987). MATERIAL AND METHODS Surgery and Prelakreling of RGC Ten adult female Sprague-Dawley rats were used in this study. In order to prelabel RGC the optic nerve was surgically exposed under deep chloralhydrate anesthesia (0.42mgkg body weight) arid completely transected as described recently (Baehr et al., 1989).A combination of the fluorescent tracers Di-I (1,l-dioctedecyl-3,3,3,3tetramethylindocarbocyanine, Molecular Probes, Junction City, 25% in saline containing 1% TritonX-100)and RITC (rhodamine-iso-thiocyariate, Sigma, St. Louis, MO) was applied to the proximal stump of the optic nerve. Both dyes are taken up by RGC axons and retrogradely label RGC in th? retina (Thanos et al., 1987;Vidal-Sanz et al., 1988).For conditioning lesions necessary to induce fast regrowth from retinal explants (Ford-Holevinsky et al., 1986)the optic nerve was exposed, freed from the meninges, and crushed with fine watchmaker’s forceps for 10 s. After labeling or crushing, the retina vasculature was inspected through the lens to make sure that the bload supply was not interrupted. Animals with reduced blood supply were not considered for further experiments. Culture Media The culture media used in the present experiments have been described recently. Briefly, N2 defined medium was prepared as described by Bottenstein and Sat0 (1979)and modified according to Bunge et al. (1982).When adult retina was cultured on glia monolayers (retina-astrocyte coculture:;) N2 chemically defined medium was used with daily addition of ascorbate (20 bg/ml); (Baehr et al., 1988,1989).For astrocyte cultivation Dulbecco’s modified Eagle’s medium (DMEM, Gibco) with 10% fetal bovine serum (FBS)was used. For astrocyte purification fluorodeoxyuridine (FdU, Sigma, lop5)was added as alternating feed every 2 days to DMEM medium. Astrocyte Preparation Astrocytes were prepared from cerebral cortices of 1-day-oldrat pups using the method of McCarthy and de Vellis (1980)as modified by Noble et al. (1984).Briefly, dissociated cells from perinatal cortices were grown in tissue culture flasks (Falcon)until they were confluent. After that, the flask was shaken overnight on a rotary shaker to remove oligodendrocytes and neurons which were sitting on top of the underlying astrocytes. The cultures were then treated with two cycles of FdU to kill dividing fibroblasts or progenitor cells. In our experiments, cultures were used immediately after purification (“immature”astrocytes, 7--10days in vitro with flat
RETINA REGENERATION ON ASTROCYTES AND S C H W A ” CELLS
morphologies) or after subculturing (“mature” astrocytes, more than 3 weeks in vitro, which display stellate morphologies).For cocultures with adult retinae, astrocytes were seeded on ammoniated collagen covered Aclar plastic dishes. The plating densities were 1-lox cells/mm2.All cell cultures used for cocultivation with adult retinae were allowed to grow until they became confluent monolayers before retinal explants were added. Retinal Explants The explantation procedure of adult rat retina has been described recently (Baehr et al., 1988, 1989). Briefly, retinae were dissected in sterile Hanks’ buffered salt solution (HBSS) with penicillin-streptomycin (20 kg/ml Sigma, St. Louis, MO) and whole-mounted on semi-permeable membrane filters (Millipore) with the ganglion cell layer upwards. Eight triangular pieces centered in the optic disk were cut with a tissue chopper and transferred to cellular monolayers on ammoniated collagen. With the ganglion cell layer attached t o the cell layer retinae were cultured in a high oxygen chamber (approx. 70% O2 and 5% COB)at 37°C. Ascorbate (20 kg/ml) was added daily to each culture medium. Cultures were refed with each medium every 2 days. Immunohistochemistry Astrocytes permeabilized with 95% ethanol and 5% acetic acid (at -2O”C, 10 min) were incubated with antisera to glial fibrillary acid protein (GFAF’),which is expressed by astrocytes but not by oligodendrocytes or fibroblasts (Bignami et al., 1972). In purified astrocyte cultures more than 95% of the cells were GFAP positive. For double staining of astrocytes with A2B5 antibodies (American type culture collection, Rockville, MD) and GFAP cultures were incubated with A2B5 (150 in L-15) for 30 min at room temperature (RT), fixed in 4% PFA, and then incubated with a fluorescein conjugated antimouse secondary. After that, cultures were processed for GFAP as described above. Less then 5% of the GFAP positive cells were A2B5 positive (type I1 astrocytes according to Raff et al., 1983). To visualize neurite regrowth, cultures were permeabilized, rinsed, and incubated with a mouse anti-neurofilament mAb (SMI 31, Sternberger-Meyer Immunocytochemicals, Jarretsville) for 45 min a t room temperature. After rinsing, cultures were incubated with a fluorescein conjugated goat anti-rabbit secondary, mounted, and observed on a Zeiss inverted microscope equipped for epifluorescence. The control for specific staining was incubation with secondary antibodies alone without primary antibody.
295
beled RGC per unit area (400 x 300 m) was determined using an inverted microscope with epifluorescence optics. Densities of prelabeled RGC were calibrated from four areas in 200 pm distance of the optic disc in 20 different explants. At the time of explantation RGC densities in this region ranged from 1,200 to 1,700 RGC/mm2, which constitutes 50-70% of the normal RGC population of adult albino rats (Bray et al., 1987; Vidal-Sanz et al., 1988). This cell attrition is probably due to the depletion of the RGC population followingthe axotomy performed 5-7 days before explantation. The number of fluorescent RGC, counted half an hour after explantation, was set as 100%.Two, 4,and 6 days after explantation the numbers of surviving RGC were counted in different regions of the same distance and size to the optic disc (40 areas of 10 explants) for each condition. Densities were normalized and compared with the initial values. Degenerating RGC could be identified by shrinkage and leaking of the dye to the environment. Only viable RGC were included in our measurements. For neurite counts 20 explants were immunostained for neurofilament and scored for each condition 48 h after explantation by means of an inverted fluorescence microscope (Zeiss). Counts were performed at the edges of the explanted retinae, where most of the RGC neurites are defasciculated. Only well-attached retinae which had extended neurites were considered for analysis. To evaluate the growth capacity of RGC neurites on each cellular monolayer, area measurements were performed, because it was impossible to trace individual neurite length from the explant to the front of outgrowth. Therefore, camera lucida drawings of the entire area covered with neurites were performed by connecting the tips of the most outward neurites. Single neurites that had grown for a longer distance than the normal population were not considered in this analysis. At least five different retinal explants from two different experiments were examined (a total of 40 retinal pieces), camera lucida drawings were performed, and the areas were measured on a digitizer tablet. To determine the growth rate the mean diameter of the neuritic halo was determined and the growth rate was calculated. This gves rather an underestimation of the mean growth rate, since axons which have grown for longer distances are not represented and single axon-trajectories which usually do not follow straight lines but might be described as growing in serpentines are shortend to direct distances. For statistical analysis, individual t-tests were performed.
RESULTS RGC Survival
Retinal ganglion cell survival over a period of 6 days in culture with astrocytes is shown in Figure 1 along with a comparison to previous observations on survival RGC and Neurite Counts on SC preparations (see Discussion). In adult rat retinal Viable RGC were scored as described recently (Baehr explants cultured on astrocytes 5040% of the initial et al., 1989). Briefly, the number of Di-I \ RITC prela- RGC population defined by prelabeling in vivo is still
296
BAEHR AND BUNGE 100 80
60 40
20
Days i t i v i t r s
Fig. 1. RGC survival on glia monolayers. In vivo prelabeled RGC in explants from adult retinae were cultured on different glia monolayers. The graph shows the normalized numbers of viable RGC at the time of explanation (DO)and after 2,4, and 6 d i v . (n = 20). At DO the number of fluorescence labeled RGC was counted in standard areas, normalizing, and set 100%.At 2,4, and 6 d.i.v. remaining RGC were scored and compared to the initial cell numbers. On flat ‘immature” astrocytes ( A), “immature” SC (m, +), and stellate “mature” astrocytes (O), a continuous loss of fluorescence prelabeled RGC can be observed. Only activated mature SC (0)are able to stabilize RGC numbers for about 3 d i v . At all time points (2, 4, and 6 d.i.v.1 the retinal explants on activated mature SC show significantly more viable RGC as compared to the other glial populations (t-tests, P c 0.001).
viable after 2 days in vitro (d.i.v.; see Fig. 1).With 38-39% viable RGC at 4 and 15-21% at 6 d.i.v., there are no significant differences among these different glial populations in supporting RGC survival. Therefore, on flat and stellate astrocytes a continuous loss of RGC within the retinal explants can be observed (Fig. 2). Neurite Promotion of Astrocytes Purified astrocytes, cultured in the presence of serum-containing medium for not more than 7 days, show flat morphologies and are known to express the extracellular matrix components laminin and heparin-sulfate proteoglycan (Ard and Bunge, 1988; Liesi et al., 1983). When astrocytes are shifted to serum-free medium and subcultured for at least 10 days, a change from flat to more stellate morphologes occurs and laminin expression ceases. Adult retinal explants cultured on flat astrocytes show defasciculated neurite growth on the astrocyte surfaces and form fine fiber networks within 2 days in vitro (see Fig. 3). Stellate astrocytes are also permissive for axonal regrowth from adult retinal explants and RGC neurites regrow along astrocyte processes (see Figs. 43). After 2 d.i.v., 64 neurites extend on “mature” stellate astrocytes as compared to 69 neurites on the “immature” population (Table 1).We did not observe major differences in neurite branching on either astrocyte monolayer, which could result in different neurite densities within the neuritic halo.
Fig. 2. Time dependent RGC degeneration in vitro. A fluorescent micrograph of living, RITC/Di-I rehbeled RGC is shown. Retinal quadrants in the same distance (afout 200 km) to the optic disc region are shown 2 (A), 4 (B), and 6 (C) days after explantation on an immature astrocyte monolayer. Individual fluorescence labeled viable RGC can be identified. A continuous decrease in the number of viable RGC can be observed after 4 and 6 daj s of observation. Bar, 100 pm.
Retinal explants on stellate astrocytes cover only an area of 0.61 mm2 with neurites but on flat astrocytes 1.06 mm2 of the culture dish are covered with neurites after 2 d.i.v. (see Table 1).Similar results are obtained when the mean growth rates were determined (Table 1). Statistical analysis of these data shows that flat astrocytes are significantly better in supporting RGC axon growth than stellate astrocytes (t-tests, P d 0.001).
RETINA REGENEMTION ON ASTROCYTES AND S C H W A " CELLS
Fig. 3. Coculture of adult retinal explants with flat astrocytes. Retinal neurites that had re-extended on immature, flat astrocyte monolayer are shown. The same area in about 200 pm distance from a retinalexplant is shown in phase contrast (A),after anti-GFAP (B),and anti-neurofilament immunostaining (C). The immature astrocytes show polygonal shapes (A,B) and retinal neurites (C) grow defasciculated on astrocyte surfaces. The arrowhead indicates the nucleus of an astrocyte seen in A-C. Bar, 100 pm.
DISCUSSION Our results suggest that cultivation conditions have a major influence on glial populations in promoting survival and axon regeneration from adult tissues. Astrocytes derived from perinatal rat cortices show flat morphologies when cultured in serum-containing media for
297
Fig. 4. Coculture of adult retina with stellate (mature) astrocytes. Neurite regrowth from retinal explants cultured for 2 days on stellate astrocytes (A,C)is shown. Mature astrocytes attain stellate morphologies (A, phase contrast). Double labeling of the coculture with antiGFAP (B)and anti-neurofilament (C) shows RGC neurite regrowth on small astrocyte processes. The neurites avoid the collagen which serves as a primary substrate for astrocytes. Bar, 100 pm.
up to 10 days. These astrocytes have been shown to resemble immature astrocytes with heparin sulfate proteoglycan and laminin immunoreactivity (Ard and Bunge, 1988; Liesi et al., 1983).Astrocytes, cultured for at least 2 weeks and subcultured for 7 days in chemically defined media, attain stellate morphologies (Ard and Bunge, 1988; Morrison and de Vellis, 1983). This morphological change is correlated with reduced ex-
298
BAEHR AND BUNGE TABLE 1. Neurite growth on glia monolayers
Glia population Flat astrocytes Stellate astrocytes Quiescent SCb Mature SCh
Neurites per explant
Area covered with neurites (mm2)
69 f 23 64 18 82 k 19 361 75
1.06 0.25 0.61 k 0.12 1.25 0.65 2.004 0.34
+
*
+
+ +
Growth rate (Irm/h) 16.7 11.2 20.4 34
“This table shows themean number of neuritesthat h a d extended from each adult retinal explant after 2 d.i.v. ( n = 20 for each condition), t h e mean area on the cellul a r monolayer covered with neurites, and t h ? growth rate of RGC axons under each condition. For measurements, cultures were fixed in 4Y1 PFA, permeabilized, a n d stained with neurofilarnent antibodies. There is a significant increase in RGC axon outgrowth on “activated” mature SC as compared to stellate or flat astrocytes and “immature” SC (for description of the glia preparations see text). Although there are no significantdifferences in t h e numberof neurites extending from each retinal explant on quiescent SC, flat and stellate astrocytes, the mean area covered with astrocytes. On mature “activattd” SC a significant increase of the mean area rovered with neurites and of t h e growth rate is significantly lower un stellate astrocytes. On mature “activated” SC a significant increase of the mean area covered with neuritrs and of the growth rate was observed as cumpared to all other glia preparations (t-tests, P 5 0.001). hParts of t h e w d a t a a r r from a n extensive s t u l y of RGC neurite growth im a variety uf SC surfaces which h a s been puhlishrd wparatrly (Raehr and Bunge, 1989).
Fig. 5. Comparison ofRGC neurite growth on immature quiescent or mature activated SC monolayer. Neurites (visualized by anti-neurofilament staining A,C) extend on activated mature (A,B),quiescent or “immature” SC (C,D).Activated SC (B, phase contrast) retain a linear array and neurites which regrow from the edge of a retinal explant (delineated with arrowheads) grow defasciculated on SC surfaces. Quiescent cultures SC show less alignment (D, phase contrast) and fewer neurites extend from retinal explants (C, anti-neurofilament). The edge of the explant is marked with arrowheads; the dark spot represents the filter used to flat mount the retina. Bar, 100 km.
pression of extracellular matrix (ECM)components like laminin (Ard and Bunge, 1988) and also occurs when astrocytes are in contact with neuronal membranes (Sobue and Pleasure, 1984). &ellate astrocytes therefore might be comparable to mature astrocytes present in the adult optic nerve in situ (for review, see Raff, 1989).When adult rat retina is cultured on these different astrocyte preparations, the flat population (resembling “immature”astrocytes found during developmental growth of RGC axons) supports axon growth significantly better than do stellate astrocytes. Nevertheless, stellate astrocytes (“mature” glia which represents astrocytes of the adult r i t optic nerve in this in vitro paradigm) were somewhat permissive for axonal regrowth. This observation shows some similarity to the in vivo condition, where some sprouting occurs on laminin negative astrocytes after intracranial optic nerve transection in adult rats (Giftochristos and David, 1988).However, none of the aetrocyte populations was able to prevent continuous degeneration of RGC within the explants over time. This wggests that astrocytes which have been shown to provide support for some embryonic tissues (McCaffery ct al., 1984) seem to lack the ability to support survival of adult rat RGC. Interestingly, for SC the situation is quite different (Baehr and Bunge, 1989): Schwann cells, derived from perinatal sciatic nerves and cultured for 7 to 10 days in vitro (as quiescent confluent inonolayers, “immature” SC) support adult neurite regeneration in a manner similar to that seen with immature astrocytes (see Table 1 and Fig. 1).Recently culture conditions have been established in which DECG neurons and SC are cultured as pure populations and then recombined (Bunge et al., 1982; Wood, 1976). When these cultures were maintained in serum- and ascorbate-containing media SC ensheathe and myelinate neurites and produce a basal lamina sheath around the SC-neuron unit (Eldridge et al., 1987,1989). This resembles the organization of the bands of Bungnu in vivo (Bunge, 1987).
RETINA REGENERATION ON ASTROCYTES AND S C H W A ” CELLS
After removal of DRG neurons from these cultures, their axons degenerate and leave behind SC (“activated mature SC) within their basal laminae, a situation comparable to Wallerian degeneration in vivo (for review, see Bray et al., 1981; Bunge, 1987).These SCs and their basal lamina tubes retain a linear array for several days in culture (activated mature SC). Regenerating neurites join these units, which resemble the bands of Bungner in the adult rat sciatic nerve and grow over long distances in vitro (Baehr and Bunge, 1989; Hopkins et al., 1989). This in vitro system therefore allows comparison to CNS regeneration in segments of sciatic nerves transplanted to replace the cut optic nerve (for review, see Bray et al., 1987). It has been shown in vivo that in the presence of a peripheral nerve graft, axotomized adult rat RGC retain their ability to regenerate and RGC extend long neurites into these grafts (Bray et al., 1987). On the other hand, inhibition of neurite regeneration could be observed when a segment of the adult optic nerve was used to bridge the cut ends of a sciatic nerve (Hall and Kent, 1987) which might correspond to our observation that stellate astrocytes comparable to those present in adult rat optic nerves are a bad substrate for regenerating neurites. Our results with adult retinal neurites regenerating on astrocytes of different types in vitro show that flat astrocytes support axonal regrowth in a manner comparable to immature SC (Table 1). On stellate astrocytes, however, neurite extension is significantly reduced, although approximately the same numbers of neurites extend from the explants at 2 d.i.v. This reduced ability of stellate astrocytes to support neurite regeneration (Fawcett et al., 1989) shows a correlation with the reduced laminin expression of astrocytes subcultured for several days in chemically defined medium (Liesi et al., 1983; Ard and Bunge, 1988). In the sciatic nerve of the adult rodent SC start to proliferate after axotomy and provide a particularly favorable milieu for regenerating neurites (for review, see Bray et al., 1981). Under these conditions, SC increase their expression of cell surface molecules like L1 or N-CAM (Seilheimer and Schachner, 1987). It has recently been demonstrated that cell surface molecules like L1 play an important role in promoting neurite growth of embryonic rat retinal neurites on SC (Kleitman et al., 1988a,b). In our in vitro model of regenerating adult retina (Baehr and Bunge, 1989), both RGC survival and the neurite growth rates on SC surfaces are increased on “activated” mature SC as compared to the quiescent (immature) population which contains no proliferating or differentiating SC (see Table 1 and Fig. 1).These results suggest that “immature” central and peripheral glia populations show comparable abilities to support neurite regeneration. Mature astrocytes and SC, however, show substantial differences when confronted with regenerating adult neurites. This might resemble the in vivo situation where the CNS glia is non-permissive for axon regeneration (Fawcett et al., 1989) but peripheral nerves,
299
replacing the normal CNS environment, support regenerating neurites effectively (Bray et al., 1987; VillegezPerez et al., 1988). In the future it will be interesting t o investigate whether these differencesin promoting neurite growth are reflected in changes of cell surface or ECM molecules like L1, N-CAM, or laminin (Bixby et al., 1988; Seilheimer and Schachner, 1987, 1988; Tomaselli et al., 1988)on glial cells and to determine the contribution of single molecules in adult RGC neurite promotion on glial cells. REFERENCES Ard, M.D. and Bunge, R.P. (1988) Heparin sulfate proteoglycan and laminin immunoreactivity on cultured astrocytes: relationship t o differentiation and neurite owth. J . Neurosci., 8:2844-2858. Baehr, M., Vanselow, J., a n d g a n o s , S. (1988)In vitro regeneration of adult rat ganglion cell axons from retinal explants. Exp. Brain Res., 73:393-401. Baehr, M., Vanselow, J., and Thanos, S. (1989) Ability of adult rat ganglion cells to regrow axons in vitro can be influenced by fibroblast growth factor and gangliosides. Neurosci. Lett., 96:197-201. Baehr, M. and Bunge, R.P. (1989) Functional status influences the ability of Schwann cells to promote adult rat retinal ganglion cell survival and axonal regrowth. Exp. Neurol., 106:27-40. Bignami, A,, Eny, L.F.,Dahl, D., and Uyeda, C.T. (1972)Localization of the glial fibri lary acidic protein in astrocytes by immunofluorescence. Brain Res., 43:429-435. Bixby, J.L., Lilien, J., and Reichardt, L. (1988) Identification of the major proteins that promote neuronal process outgrowth on Schwann cells in vitro. J. Cell Biol., 107:353-361. Bottenstein, J.E. and Sato, G.H. (1979) Growth of a rat neuroblastoma cell line in serum-free supplemented medium. Proc. Natl. Acad. Sci. USA, 76:514-517. Bray, G.M., Rasminsky, M., and Aguayo, A.J. (1981) Interactions between axons and their sheath cells. Annu. Reu. Neurosci., 41127-162. Bray, G.M., Villegaz-Perez, M.P., Vidal-Sanz, M., and Aguayo, A. (1987) The use of peripheral nerve grafts to enhance neuronal survival, promote growth and permit terminal reconnections in the central nervous system of adult rats. J . Exp. Biol.,132:5-19. Brockes, J.P., Fields, K.L., and Raff, M.C. (1979)Studies on cultured rat Schwann cells. I. Establishment of purified populations from cultures of peripheral nerve. Brain Res., 165:105-118. Bunge, R.P. (1987)Tissue culture observations relevant to the study of axon-Schwann cell interactions during peripheral nerve development and repair. J. Exp. Biol., 132:21-34. Bunge, R.P., Bunge, M.B., Carey, D.J., Cornbrooks, C.J.,Higgins, D.H., Johnson, M.I., Iacovitti, L., Kleinschmidt, D.C., Moya, F., and Wood, P. (1982) Functional expression in primary nerve tissue cultures maintained in defined medium. Cold Spring Harbor Conference Cell Proliferation, 9:1017-1031. Cajal, S.R.Y. (1968) Traumatic degeneration and regeneration in the optic nerve and retina. In: Degeneration and Regeneration of the Nervous System. Translated by R.M. May. Hafner, New York, pp. 583-596. Caroni, P. and Schwab, M.E. (19881 Two membrane protein fractions from rat central myelin with inhibitory properties for neurit,egrowth and fibroblast spreading. J . Cell Biol., 106:1281-1288. Cohen, J., Burne, J., McKinlay, C., and Winter, J . (1987) The role of laminidfibronectin receptor complex in the outgrowth of retinal ganglion cell axons. Deu. Biol.,122:407418. Eldridge, C.F., Bunge, M.B., Bunge, R.P., and Wood, P.M. (1987) Differentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. J . Cell Biol.,105:1023-1034. Eldridge, C.F., Bunge, M.B., and Bunge, R.P. (19891 Differentiation of axon-related Schwann cells in vitro: 11. Control of myelin formation by basal lamina. J . Neurosci., 9:625-638. Fallon, J.R. (1985) Preferential outgrowth of central nervous system neurites on astrocytes and Schwann cells as compared to nonglial cells in vitro. J . Cell Biol., 100:198-207. Fawcett, J.W., Housden, E., Smith-Thomas, L., and Meyer, R.L. (1989) The growth of axons in three-dimensional astrocyte cultures. Deu. Biol., 135:449485.
300
BAEHR AND BUNGE
Ford-Holevinski, T.S., McCoy, J.P., Hopkins, J.M., and Agranoff, B.W. (1986) Laminin supports neurite outgrowth from explants of axotomized adult rat retinal neurons. Deu. Brain Res., 28:121-126. Giftochristos, N. and David, S.(1988) Laminin and heparan sulfate proteoglycan in the lesioned adult mammalian central nervous system and their possible relationship to axonal sprouting. J . Neurocytol., 17:385-397. Hall, S.M. and Kent, A.P. (1987) The response of regenerating peripheral axons to a grafted optic nerve. J . Neurocytol., 16:317-331. Hopkins, J.M., Schachner, M., and Bunge, R.P. (1989) Antibodies against the cell adhesion molecule L 1do not block growth ofadult rat retinal ganglion cell axons on Schwann cells. SOC.Neurosci. Abstr., 15:331. Kalderon, N. (1988) Differentiating astroglia in nervous tissue histogenesis regeneration: studies in a model system of regenerating peripheral nerve. J . Neurosci. Res., 21:501-512. Kleitman, N., Wood, P., Johnson, M.I., and Bunge, R.P. (1988a) Schwann cell surfaces but not extracellular matrix organized by Schwann cells support neurite outgrowth from embryonic rat retina. J . Neurosci., 8:653-663. Kleitman, N., Simon, D.K., Schachner, M., and Bunge, R.P. (1988b) Growt,h of embryonic retinal neurites elicited by contact with Schwann cell surfaces is blocked by antibodies to L1. Exp. Neurol., 102:298-306. Liesi, P., D. Dahl, and A. Vaheri (1983) Laminin is produced by early rat astrocytes in primary culture. J . Cell Biol., 96:920-924. Liesi, P. (1985)Laminin-immunoreactive glia distinguish regenerative adult CNS systems from non-regenerative ones. EMBO J., 4:2505-2511. Liesi, P. and Silver, J . (1988) Is astrocyte laminin involved in axon guidance in the mammalian CNS? Deu. B i d , 130:774-785. Liuzzi, F.J. and Lasek, R.J. (1987) Astrocytes block axonal regeneration in mammals by activating the physiological stop pathway. Science, 237:642-645. McCaffery, G.A., Raju, T.R., and Bennett, M.R. (1984) Effects of cultured astroglia on the survival of neonatal rat retinal ganglion cells in vitro. Dev. Biol.,104:661-668. McCarthy, K. and deVellis, J . (1980)Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J . Cell Biol., 85:890-902. Miller, R.H., Abney, E.R., David, S., ffrench-Constant, C., Lindsay, R., Patel, R., Stone, J., and Raff, M.C. (1986) Is reactive gliosis a property of a distinct subpopulation of astrocytes? J . Neuroscz., 6122-29. Morrison, R.S. and de Vellis, J . (1983) Differentiation of purified astrocytes in a chemically defined medium. Dev. Bruin Res., 9:337-345. Noble, M., Fok-Seang, J., and Cohen, J. (1984) Glia are a unique substrate for the in vitro growth of central nervous system neurons. J . Neurosci., 4:1892-1903.
Raff, M.C. (1989) Glial cell diversification in the rat optic nerve. Science, 243:1450-1455. Raff, M.C., Abney, E.R., Cohen, J., Lindsay, R., and Noble, M. (1983) Two types of astrocytes in cultures of developing white matter: Differences in morphology, surface ga ngliosides, and growth characteristics. J. Neurosci.. 3:1289-1300. Reier, P.J., Stensaas, L.J., and Guth, L. 1 1983)The astrocytic scar as an impediment to regeneration in the central nervous system. In: Spinal Cord Reconstruction. C.C. Kao, R.P. Bunge, and P.J. Reier, eds. Raven Press, New York, pp. 163--195. Schwab, M.E. and Thoenen, H. (1985) Dissociated neurons regenerate into sciatic but not optic nerve explmts in culture irrespective of neuronotrophic factors. J . Neurosci., 5:2415-2423. Schwab, M.E. and Caroni, P. (1988) 0lii:odendrocytes and CNS myelin are non-permissive substrates for Iieurite growth and fibroblast s reading in vitro. J . Neurosci., 8:2381-2393. SeiPheimer, B. and Schachner, M. (1987) Regulation of neural cell adhesion molecule expression on c u h r e d mouse Schwann cells by nerve growth factor. EMBO J., 6:1611-1616. Seilheimer, B. and Schachner, M. (19881 Studies of adhesion molecules mediating interactions between celle of peripheral nervous system indicate a major role for L1 in mediaXing sensory neuron growth on Schwann cells in culture. J. Cell Bid , 107:341-351. Sidman, R. and Rakic, P. (1973) Neuronal migration, with special reference to developing human brain a review. Bruin Res., 62:l-35. Silver, J. and Rutishauser, U. (1984) Guidance of axons in vivo by a preformed adhesive pathway on neuroepithelial endfeet. Deu. Biol., 106:485-499. Sobue, G. and Pleasure, D. (1984) Astriglial proliferation and phenotype are modulated by neuronal plasma membrane. Brain Res., 324: 175-179. Thanos, S., Vidal-Sanz, M., and Aguayo, A.J. (1987) The use of rhodamine-P-isothiocyanateRITC as a anterograde and retrograde tracer in the adult rat visual system. Brain Res., 406:317-321. Tomaselli, K.J., Neugebaum, U.M., Bixby, J.L., Lilien, J., and Reichardt, L.F. (1988) N-cadherin and integrins: two receptor systems that mediate neuronal process outgrowth on astrocyte surfaces. Neuron, 1:33-43. Vidal-Sanz, M., Villegas-Perez, M.P., Bray, G.M., and Aguayo, A.J. (1988) Persistent retromade labelin:: of adult rat retinal ganglion cells with the carbocya&e dye diI. &p. Neurol., 102:92-161. Villegas-Perez, M., Vidal-Sanz, M., Eiray, G.M., and Aguayo, A.J. (1988) Influences of peripheral neriie grafts on the survival and regrowth of axotomized retinal ganglion cells in adult rats. J . Neurosci., 8:265-280. Wood, P.M. (1976)Separation of functional Schwann cells and neurons from normal peripheral nerve tissue. Brain Res., 115:361-375.
-
Growth of Adult Rat Retinal Ganglion Cell Neurites on Astrocytes M. BAEHR AND R.P. BUNGE The Miami Project to Cure Paralysis, University o f M i a m i School of Medicine, Miami, Florida 33136
KEY WORDS
Central nervous system, Glia, Regeneration, In vitro
ABSTRACT Astrocytes, as well as Schwann cells (SC),can provide suitable substrata for embryonic neurites during development, but their abilities to support adult regenerating neurites have not been directly compared. The aim of the present study was to determine the ability of astrocytes to promote adult rat retinal ganglion cell (RGC) regeneration in vitro and to compare this to previously determined growth on the surface of Schwann cells. We prepared Type I astrocytes (Raff et al: J . Neurosci. 3:1289-1300, 1983)from perinatal rats. These were subcultured and maintained in either a serum-free medium for at least 2 weeks (stellate astrocytes with little immunoreactivity for laminin) or in serum containing medium for 7 to 10 days (flat and polygonal astrocytes with immunoreactivity for laminin). Stellate astrocytes might therefore represent mature astrocytes in vivo (Ard and Bunge: J . Neurosci. 8:2844-2858,1988), while flat astrocytes might resemble immature brain astrocytes (Liesi et al: J . Cell Biol. 96:920-924, 1983). Adult RGC survival and axonal regrowth on these glia populations was compared to that observed on different SC populations, as previously reported (Baehr and Bunge: Exp. Neurol. 106:27-40, 1989). Both astrocyte populations (either flat or stellate astrocytes) did not enhance RGC survival. Stellate astrocytes were less effective in supporting RGC axon regeneration than flat astrocytes. When these date were compared to RGC survival and axon growth on SC (Baehr and Bunge: Exp.Neurol. 106:2740,1989)only “acitvated mature SC populations were superior to astrocytes in enhancing RGC survival and neurite regrowth. These results suggest 1)that astrocytes and “immature”SC are similar in their ability to support RGC survival; 2) “activated” mature SC populations are significantly better than astrocytes and “immature” SC in enhancing RGC survival and neurite growth; 3) stellate (“mature”) astrocytes, although permissive for regrowing axons, are not a favorable substrate for regenerating adult RGC neurites, nor do they effectively support RGC survival.
INTRODUCTION One of the major problems in studies of mammalian central nervous system (CNS) regeneration is to define the changes that occur in glial populations as the milieu of the CNS changes from being supportive to being inhibitory to neurite growth. In the developing CNS of mammals, outgrowing neurites interact with surrounding glia which provide substrata for growth and guidance for developing axons (Cohen et al., 1987; Fallon, 1985; Liesi and Silver, 1988; Noble et al., 1984; Sidman and Rakic, 1973; Silver and Rutishauser, 1984; Tomaselli et al., 1988). In contrast, mature CNS glia do not 0 1 9 9 0 Wiley-Liss, Inc.
support regeneration and may even be inhibitory for axonal regrowth (Cajal, 1968; Kalderon, 1988; Liuzzi and Lasek, 1987; Reier et al., 1983; Schwab and Caroni, 1988). It has been demonstrated that the adult optic Received November 28,1989; accepted February 13,1990 Address reprint requests to Dr. Mathias Baehr a t his permanent address M. Baehr’s permanent address is Max-Planck Institut fur Entwicklungsbiolopie, Spemannstr. 35/I, 7400 Tubingen, FRG Acknowledgments: The authors thank Dr. Naomi Kleitman for helpful discussions and comments on the paper and Susan Mantia and Charlaine Rowlette for typing the manuscript. Dr. D.D.M. O’Leary generously provided surgical equipment and facilities. This work was supported by the Max-Planck Gesellschaft (M. Baehr) and NIH grant NS09923 (R.P.Bungel.
294
BAEHR AND BUNGE
nerve is not permissive for regenerating neurites in vivo (Bray et al., 1987;Cajal, 1968;Giftochristos and David, 1988) and in vitro (Schwab and Thoenen, 1985). Recently a similar inhibition of neurite regrowth could be observed by confronting regenerating axons with mature oligodendrocytes or components of adult rat CNS myelin (Caroni and Schwab, 1988;Schwab and Caroni, 1988).Therefore, regrowth might be inhibited by mature oligodendrocytesand myelin present at lesion sites in the adult CNS, even if some adult CNS neurons retain their ability to regenerate axons. On the other hand, after lesioning the adult CNS, so-called “reactive” astrocytes can be observed, which may form a fibrous gliosis, thus preventing neurite regeneration (for review see Liuzzi and Lasek, 1987; Reier et al., 1983).Recently a classification of astrocytes into two subpopulations (Type I and 11) has been described (Raff et al., 1983)and it has been suggested that Type I astrocytes which support axon growth during development might be responsible for the formation of the glial scars in adult animals which are generally considered to be inhibitory for axonal regeneration (Miller et al., 1986). We have recently shown that adult rat RGC, defined by prelabeling in vivo from RGC targets and axotomized prior to explanation, survive in tissue culture and regrow axons on defined cellular substrata (Baehr et al., 1988, 1989). Within 48 h after explanation, neurite regrowth could be observed on polylysine plus laminin but not on collagen substrata (Baehr et al., 1988).This system makes it possible to examine interactions of regenerating adult CNS neurites with characterized glia populations in vitro. In the present study we have examined RGC survival and neurite regrowth on two forms of Type I astrocytes cultured on ammoniated collagen which correspond to “immature” astrocytes found during development (flat astrocytes with immunoreactivity for laminin) or to “mature” astrocytes (stellate astrocytes) of mature brain in vivo (Ard and Bunge, 1988;Liesi, 1985;Liesi et al., 1983;Liesi and Silver, 1988).RGC regeneration on these CNS glia populations was compared t o that observed on different SC populations, which are known to support neurite outgrowth and regeneration in vivo (for review see Bray et al., 1981)and in vitro (Baehr and Bunge, 1989;Bunge, 1987;Bunge et al., 1982;Hopkins et al., 1989).In these earlier studies (Baehr and Bunge, 1989),“immature” SC were seeded on purified dorsal root ganglia (DRG)neurites and cultured in serum and ascorbate containing media. This allows SC to repopulate the DRG and t o differentiate in contact with axons (Bunge et al., 1982;Eldridge et al., 1987,1989).These cultures were maintained until SC had started to ensheathe and myelinate DRG axons. The DRG neurons were then removed, which causes the neurites to degenerate, providing a model of Wallerian degeneration in vitro (for review, see Bunge, 1987).Adult retinal explants cultured on these SC monolayers are confronted with SC populations comparable to those found in sciatic nerve grafts. This serves as an in vitro model for the
adult sciatic nerve in vivo (for review, see Bray et al., 1987). MATERIAL AND METHODS Surgery and Prelakreling of RGC Ten adult female Sprague-Dawley rats were used in this study. In order to prelabel RGC the optic nerve was surgically exposed under deep chloralhydrate anesthesia (0.42mgkg body weight) arid completely transected as described recently (Baehr et al., 1989).A combination of the fluorescent tracers Di-I (1,l-dioctedecyl-3,3,3,3tetramethylindocarbocyanine, Molecular Probes, Junction City, 25% in saline containing 1% TritonX-100)and RITC (rhodamine-iso-thiocyariate, Sigma, St. Louis, MO) was applied to the proximal stump of the optic nerve. Both dyes are taken up by RGC axons and retrogradely label RGC in th? retina (Thanos et al., 1987;Vidal-Sanz et al., 1988).For conditioning lesions necessary to induce fast regrowth from retinal explants (Ford-Holevinsky et al., 1986)the optic nerve was exposed, freed from the meninges, and crushed with fine watchmaker’s forceps for 10 s. After labeling or crushing, the retina vasculature was inspected through the lens to make sure that the bload supply was not interrupted. Animals with reduced blood supply were not considered for further experiments. Culture Media The culture media used in the present experiments have been described recently. Briefly, N2 defined medium was prepared as described by Bottenstein and Sat0 (1979)and modified according to Bunge et al. (1982).When adult retina was cultured on glia monolayers (retina-astrocyte coculture:;) N2 chemically defined medium was used with daily addition of ascorbate (20 bg/ml); (Baehr et al., 1988,1989).For astrocyte cultivation Dulbecco’s modified Eagle’s medium (DMEM, Gibco) with 10% fetal bovine serum (FBS)was used. For astrocyte purification fluorodeoxyuridine (FdU, Sigma, lop5)was added as alternating feed every 2 days to DMEM medium. Astrocyte Preparation Astrocytes were prepared from cerebral cortices of 1-day-oldrat pups using the method of McCarthy and de Vellis (1980)as modified by Noble et al. (1984).Briefly, dissociated cells from perinatal cortices were grown in tissue culture flasks (Falcon)until they were confluent. After that, the flask was shaken overnight on a rotary shaker to remove oligodendrocytes and neurons which were sitting on top of the underlying astrocytes. The cultures were then treated with two cycles of FdU to kill dividing fibroblasts or progenitor cells. In our experiments, cultures were used immediately after purification (“immature”astrocytes, 7--10days in vitro with flat
RETINA REGENERATION ON ASTROCYTES AND S C H W A ” CELLS
morphologies) or after subculturing (“mature” astrocytes, more than 3 weeks in vitro, which display stellate morphologies).For cocultures with adult retinae, astrocytes were seeded on ammoniated collagen covered Aclar plastic dishes. The plating densities were 1-lox cells/mm2.All cell cultures used for cocultivation with adult retinae were allowed to grow until they became confluent monolayers before retinal explants were added. Retinal Explants The explantation procedure of adult rat retina has been described recently (Baehr et al., 1988, 1989). Briefly, retinae were dissected in sterile Hanks’ buffered salt solution (HBSS) with penicillin-streptomycin (20 kg/ml Sigma, St. Louis, MO) and whole-mounted on semi-permeable membrane filters (Millipore) with the ganglion cell layer upwards. Eight triangular pieces centered in the optic disk were cut with a tissue chopper and transferred to cellular monolayers on ammoniated collagen. With the ganglion cell layer attached t o the cell layer retinae were cultured in a high oxygen chamber (approx. 70% O2 and 5% COB)at 37°C. Ascorbate (20 kg/ml) was added daily to each culture medium. Cultures were refed with each medium every 2 days. Immunohistochemistry Astrocytes permeabilized with 95% ethanol and 5% acetic acid (at -2O”C, 10 min) were incubated with antisera to glial fibrillary acid protein (GFAF’),which is expressed by astrocytes but not by oligodendrocytes or fibroblasts (Bignami et al., 1972). In purified astrocyte cultures more than 95% of the cells were GFAP positive. For double staining of astrocytes with A2B5 antibodies (American type culture collection, Rockville, MD) and GFAP cultures were incubated with A2B5 (150 in L-15) for 30 min at room temperature (RT), fixed in 4% PFA, and then incubated with a fluorescein conjugated antimouse secondary. After that, cultures were processed for GFAP as described above. Less then 5% of the GFAP positive cells were A2B5 positive (type I1 astrocytes according to Raff et al., 1983). To visualize neurite regrowth, cultures were permeabilized, rinsed, and incubated with a mouse anti-neurofilament mAb (SMI 31, Sternberger-Meyer Immunocytochemicals, Jarretsville) for 45 min a t room temperature. After rinsing, cultures were incubated with a fluorescein conjugated goat anti-rabbit secondary, mounted, and observed on a Zeiss inverted microscope equipped for epifluorescence. The control for specific staining was incubation with secondary antibodies alone without primary antibody.
295
beled RGC per unit area (400 x 300 m) was determined using an inverted microscope with epifluorescence optics. Densities of prelabeled RGC were calibrated from four areas in 200 pm distance of the optic disc in 20 different explants. At the time of explantation RGC densities in this region ranged from 1,200 to 1,700 RGC/mm2, which constitutes 50-70% of the normal RGC population of adult albino rats (Bray et al., 1987; Vidal-Sanz et al., 1988). This cell attrition is probably due to the depletion of the RGC population followingthe axotomy performed 5-7 days before explantation. The number of fluorescent RGC, counted half an hour after explantation, was set as 100%.Two, 4,and 6 days after explantation the numbers of surviving RGC were counted in different regions of the same distance and size to the optic disc (40 areas of 10 explants) for each condition. Densities were normalized and compared with the initial values. Degenerating RGC could be identified by shrinkage and leaking of the dye to the environment. Only viable RGC were included in our measurements. For neurite counts 20 explants were immunostained for neurofilament and scored for each condition 48 h after explantation by means of an inverted fluorescence microscope (Zeiss). Counts were performed at the edges of the explanted retinae, where most of the RGC neurites are defasciculated. Only well-attached retinae which had extended neurites were considered for analysis. To evaluate the growth capacity of RGC neurites on each cellular monolayer, area measurements were performed, because it was impossible to trace individual neurite length from the explant to the front of outgrowth. Therefore, camera lucida drawings of the entire area covered with neurites were performed by connecting the tips of the most outward neurites. Single neurites that had grown for a longer distance than the normal population were not considered in this analysis. At least five different retinal explants from two different experiments were examined (a total of 40 retinal pieces), camera lucida drawings were performed, and the areas were measured on a digitizer tablet. To determine the growth rate the mean diameter of the neuritic halo was determined and the growth rate was calculated. This gves rather an underestimation of the mean growth rate, since axons which have grown for longer distances are not represented and single axon-trajectories which usually do not follow straight lines but might be described as growing in serpentines are shortend to direct distances. For statistical analysis, individual t-tests were performed.
RESULTS RGC Survival
Retinal ganglion cell survival over a period of 6 days in culture with astrocytes is shown in Figure 1 along with a comparison to previous observations on survival RGC and Neurite Counts on SC preparations (see Discussion). In adult rat retinal Viable RGC were scored as described recently (Baehr explants cultured on astrocytes 5040% of the initial et al., 1989). Briefly, the number of Di-I \ RITC prela- RGC population defined by prelabeling in vivo is still
296
BAEHR AND BUNGE 100 80
60 40
20
Days i t i v i t r s
Fig. 1. RGC survival on glia monolayers. In vivo prelabeled RGC in explants from adult retinae were cultured on different glia monolayers. The graph shows the normalized numbers of viable RGC at the time of explanation (DO)and after 2,4, and 6 d i v . (n = 20). At DO the number of fluorescence labeled RGC was counted in standard areas, normalizing, and set 100%.At 2,4, and 6 d.i.v. remaining RGC were scored and compared to the initial cell numbers. On flat ‘immature” astrocytes ( A), “immature” SC (m, +), and stellate “mature” astrocytes (O), a continuous loss of fluorescence prelabeled RGC can be observed. Only activated mature SC (0)are able to stabilize RGC numbers for about 3 d i v . At all time points (2, 4, and 6 d.i.v.1 the retinal explants on activated mature SC show significantly more viable RGC as compared to the other glial populations (t-tests, P c 0.001).
viable after 2 days in vitro (d.i.v.; see Fig. 1).With 38-39% viable RGC at 4 and 15-21% at 6 d.i.v., there are no significant differences among these different glial populations in supporting RGC survival. Therefore, on flat and stellate astrocytes a continuous loss of RGC within the retinal explants can be observed (Fig. 2). Neurite Promotion of Astrocytes Purified astrocytes, cultured in the presence of serum-containing medium for not more than 7 days, show flat morphologies and are known to express the extracellular matrix components laminin and heparin-sulfate proteoglycan (Ard and Bunge, 1988; Liesi et al., 1983). When astrocytes are shifted to serum-free medium and subcultured for at least 10 days, a change from flat to more stellate morphologes occurs and laminin expression ceases. Adult retinal explants cultured on flat astrocytes show defasciculated neurite growth on the astrocyte surfaces and form fine fiber networks within 2 days in vitro (see Fig. 3). Stellate astrocytes are also permissive for axonal regrowth from adult retinal explants and RGC neurites regrow along astrocyte processes (see Figs. 43). After 2 d.i.v., 64 neurites extend on “mature” stellate astrocytes as compared to 69 neurites on the “immature” population (Table 1).We did not observe major differences in neurite branching on either astrocyte monolayer, which could result in different neurite densities within the neuritic halo.
Fig. 2. Time dependent RGC degeneration in vitro. A fluorescent micrograph of living, RITC/Di-I rehbeled RGC is shown. Retinal quadrants in the same distance (afout 200 km) to the optic disc region are shown 2 (A), 4 (B), and 6 (C) days after explantation on an immature astrocyte monolayer. Individual fluorescence labeled viable RGC can be identified. A continuous decrease in the number of viable RGC can be observed after 4 and 6 daj s of observation. Bar, 100 pm.
Retinal explants on stellate astrocytes cover only an area of 0.61 mm2 with neurites but on flat astrocytes 1.06 mm2 of the culture dish are covered with neurites after 2 d.i.v. (see Table 1).Similar results are obtained when the mean growth rates were determined (Table 1). Statistical analysis of these data shows that flat astrocytes are significantly better in supporting RGC axon growth than stellate astrocytes (t-tests, P d 0.001).
RETINA REGENEMTION ON ASTROCYTES AND S C H W A " CELLS
Fig. 3. Coculture of adult retinal explants with flat astrocytes. Retinal neurites that had re-extended on immature, flat astrocyte monolayer are shown. The same area in about 200 pm distance from a retinalexplant is shown in phase contrast (A),after anti-GFAP (B),and anti-neurofilament immunostaining (C). The immature astrocytes show polygonal shapes (A,B) and retinal neurites (C) grow defasciculated on astrocyte surfaces. The arrowhead indicates the nucleus of an astrocyte seen in A-C. Bar, 100 pm.
DISCUSSION Our results suggest that cultivation conditions have a major influence on glial populations in promoting survival and axon regeneration from adult tissues. Astrocytes derived from perinatal rat cortices show flat morphologies when cultured in serum-containing media for
297
Fig. 4. Coculture of adult retina with stellate (mature) astrocytes. Neurite regrowth from retinal explants cultured for 2 days on stellate astrocytes (A,C)is shown. Mature astrocytes attain stellate morphologies (A, phase contrast). Double labeling of the coculture with antiGFAP (B)and anti-neurofilament (C) shows RGC neurite regrowth on small astrocyte processes. The neurites avoid the collagen which serves as a primary substrate for astrocytes. Bar, 100 pm.
up to 10 days. These astrocytes have been shown to resemble immature astrocytes with heparin sulfate proteoglycan and laminin immunoreactivity (Ard and Bunge, 1988; Liesi et al., 1983).Astrocytes, cultured for at least 2 weeks and subcultured for 7 days in chemically defined media, attain stellate morphologies (Ard and Bunge, 1988; Morrison and de Vellis, 1983). This morphological change is correlated with reduced ex-
298
BAEHR AND BUNGE TABLE 1. Neurite growth on glia monolayers
Glia population Flat astrocytes Stellate astrocytes Quiescent SCb Mature SCh
Neurites per explant
Area covered with neurites (mm2)
69 f 23 64 18 82 k 19 361 75
1.06 0.25 0.61 k 0.12 1.25 0.65 2.004 0.34
+
*
+
+ +
Growth rate (Irm/h) 16.7 11.2 20.4 34
“This table shows themean number of neuritesthat h a d extended from each adult retinal explant after 2 d.i.v. ( n = 20 for each condition), t h e mean area on the cellul a r monolayer covered with neurites, and t h ? growth rate of RGC axons under each condition. For measurements, cultures were fixed in 4Y1 PFA, permeabilized, a n d stained with neurofilarnent antibodies. There is a significant increase in RGC axon outgrowth on “activated” mature SC as compared to stellate or flat astrocytes and “immature” SC (for description of the glia preparations see text). Although there are no significantdifferences in t h e numberof neurites extending from each retinal explant on quiescent SC, flat and stellate astrocytes, the mean area covered with astrocytes. On mature “activattd” SC a significant increase of the mean area rovered with neurites and of t h e growth rate is significantly lower un stellate astrocytes. On mature “activated” SC a significant increase of the mean area covered with neuritrs and of the growth rate was observed as cumpared to all other glia preparations (t-tests, P 5 0.001). hParts of t h e w d a t a a r r from a n extensive s t u l y of RGC neurite growth im a variety uf SC surfaces which h a s been puhlishrd wparatrly (Raehr and Bunge, 1989).
Fig. 5. Comparison ofRGC neurite growth on immature quiescent or mature activated SC monolayer. Neurites (visualized by anti-neurofilament staining A,C) extend on activated mature (A,B),quiescent or “immature” SC (C,D).Activated SC (B, phase contrast) retain a linear array and neurites which regrow from the edge of a retinal explant (delineated with arrowheads) grow defasciculated on SC surfaces. Quiescent cultures SC show less alignment (D, phase contrast) and fewer neurites extend from retinal explants (C, anti-neurofilament). The edge of the explant is marked with arrowheads; the dark spot represents the filter used to flat mount the retina. Bar, 100 km.
pression of extracellular matrix (ECM)components like laminin (Ard and Bunge, 1988) and also occurs when astrocytes are in contact with neuronal membranes (Sobue and Pleasure, 1984). &ellate astrocytes therefore might be comparable to mature astrocytes present in the adult optic nerve in situ (for review, see Raff, 1989).When adult rat retina is cultured on these different astrocyte preparations, the flat population (resembling “immature”astrocytes found during developmental growth of RGC axons) supports axon growth significantly better than do stellate astrocytes. Nevertheless, stellate astrocytes (“mature” glia which represents astrocytes of the adult r i t optic nerve in this in vitro paradigm) were somewhat permissive for axonal regrowth. This observation shows some similarity to the in vivo condition, where some sprouting occurs on laminin negative astrocytes after intracranial optic nerve transection in adult rats (Giftochristos and David, 1988).However, none of the aetrocyte populations was able to prevent continuous degeneration of RGC within the explants over time. This wggests that astrocytes which have been shown to provide support for some embryonic tissues (McCaffery ct al., 1984) seem to lack the ability to support survival of adult rat RGC. Interestingly, for SC the situation is quite different (Baehr and Bunge, 1989): Schwann cells, derived from perinatal sciatic nerves and cultured for 7 to 10 days in vitro (as quiescent confluent inonolayers, “immature” SC) support adult neurite regeneration in a manner similar to that seen with immature astrocytes (see Table 1 and Fig. 1).Recently culture conditions have been established in which DECG neurons and SC are cultured as pure populations and then recombined (Bunge et al., 1982; Wood, 1976). When these cultures were maintained in serum- and ascorbate-containing media SC ensheathe and myelinate neurites and produce a basal lamina sheath around the SC-neuron unit (Eldridge et al., 1987,1989). This resembles the organization of the bands of Bungnu in vivo (Bunge, 1987).
RETINA REGENERATION ON ASTROCYTES AND S C H W A ” CELLS
After removal of DRG neurons from these cultures, their axons degenerate and leave behind SC (“activated mature SC) within their basal laminae, a situation comparable to Wallerian degeneration in vivo (for review, see Bray et al., 1981; Bunge, 1987).These SCs and their basal lamina tubes retain a linear array for several days in culture (activated mature SC). Regenerating neurites join these units, which resemble the bands of Bungner in the adult rat sciatic nerve and grow over long distances in vitro (Baehr and Bunge, 1989; Hopkins et al., 1989). This in vitro system therefore allows comparison to CNS regeneration in segments of sciatic nerves transplanted to replace the cut optic nerve (for review, see Bray et al., 1987). It has been shown in vivo that in the presence of a peripheral nerve graft, axotomized adult rat RGC retain their ability to regenerate and RGC extend long neurites into these grafts (Bray et al., 1987). On the other hand, inhibition of neurite regeneration could be observed when a segment of the adult optic nerve was used to bridge the cut ends of a sciatic nerve (Hall and Kent, 1987) which might correspond to our observation that stellate astrocytes comparable to those present in adult rat optic nerves are a bad substrate for regenerating neurites. Our results with adult retinal neurites regenerating on astrocytes of different types in vitro show that flat astrocytes support axonal regrowth in a manner comparable to immature SC (Table 1). On stellate astrocytes, however, neurite extension is significantly reduced, although approximately the same numbers of neurites extend from the explants at 2 d.i.v. This reduced ability of stellate astrocytes to support neurite regeneration (Fawcett et al., 1989) shows a correlation with the reduced laminin expression of astrocytes subcultured for several days in chemically defined medium (Liesi et al., 1983; Ard and Bunge, 1988). In the sciatic nerve of the adult rodent SC start to proliferate after axotomy and provide a particularly favorable milieu for regenerating neurites (for review, see Bray et al., 1981). Under these conditions, SC increase their expression of cell surface molecules like L1 or N-CAM (Seilheimer and Schachner, 1987). It has recently been demonstrated that cell surface molecules like L1 play an important role in promoting neurite growth of embryonic rat retinal neurites on SC (Kleitman et al., 1988a,b). In our in vitro model of regenerating adult retina (Baehr and Bunge, 1989), both RGC survival and the neurite growth rates on SC surfaces are increased on “activated” mature SC as compared to the quiescent (immature) population which contains no proliferating or differentiating SC (see Table 1 and Fig. 1).These results suggest that “immature” central and peripheral glia populations show comparable abilities to support neurite regeneration. Mature astrocytes and SC, however, show substantial differences when confronted with regenerating adult neurites. This might resemble the in vivo situation where the CNS glia is non-permissive for axon regeneration (Fawcett et al., 1989) but peripheral nerves,
299
replacing the normal CNS environment, support regenerating neurites effectively (Bray et al., 1987; VillegezPerez et al., 1988). In the future it will be interesting t o investigate whether these differencesin promoting neurite growth are reflected in changes of cell surface or ECM molecules like L1, N-CAM, or laminin (Bixby et al., 1988; Seilheimer and Schachner, 1987, 1988; Tomaselli et al., 1988)on glial cells and to determine the contribution of single molecules in adult RGC neurite promotion on glial cells. REFERENCES Ard, M.D. and Bunge, R.P. (1988) Heparin sulfate proteoglycan and laminin immunoreactivity on cultured astrocytes: relationship t o differentiation and neurite owth. J . Neurosci., 8:2844-2858. Baehr, M., Vanselow, J., a n d g a n o s , S. (1988)In vitro regeneration of adult rat ganglion cell axons from retinal explants. Exp. Brain Res., 73:393-401. Baehr, M., Vanselow, J., and Thanos, S. (1989) Ability of adult rat ganglion cells to regrow axons in vitro can be influenced by fibroblast growth factor and gangliosides. Neurosci. Lett., 96:197-201. Baehr, M. and Bunge, R.P. (1989) Functional status influences the ability of Schwann cells to promote adult rat retinal ganglion cell survival and axonal regrowth. Exp. Neurol., 106:27-40. Bignami, A,, Eny, L.F.,Dahl, D., and Uyeda, C.T. (1972)Localization of the glial fibri lary acidic protein in astrocytes by immunofluorescence. Brain Res., 43:429-435. Bixby, J.L., Lilien, J., and Reichardt, L. (1988) Identification of the major proteins that promote neuronal process outgrowth on Schwann cells in vitro. J. Cell Biol., 107:353-361. Bottenstein, J.E. and Sato, G.H. (1979) Growth of a rat neuroblastoma cell line in serum-free supplemented medium. Proc. Natl. Acad. Sci. USA, 76:514-517. Bray, G.M., Rasminsky, M., and Aguayo, A.J. (1981) Interactions between axons and their sheath cells. Annu. Reu. Neurosci., 41127-162. Bray, G.M., Villegaz-Perez, M.P., Vidal-Sanz, M., and Aguayo, A. (1987) The use of peripheral nerve grafts to enhance neuronal survival, promote growth and permit terminal reconnections in the central nervous system of adult rats. J . Exp. Biol.,132:5-19. Brockes, J.P., Fields, K.L., and Raff, M.C. (1979)Studies on cultured rat Schwann cells. I. Establishment of purified populations from cultures of peripheral nerve. Brain Res., 165:105-118. Bunge, R.P. (1987)Tissue culture observations relevant to the study of axon-Schwann cell interactions during peripheral nerve development and repair. J. Exp. Biol., 132:21-34. Bunge, R.P., Bunge, M.B., Carey, D.J., Cornbrooks, C.J.,Higgins, D.H., Johnson, M.I., Iacovitti, L., Kleinschmidt, D.C., Moya, F., and Wood, P. (1982) Functional expression in primary nerve tissue cultures maintained in defined medium. Cold Spring Harbor Conference Cell Proliferation, 9:1017-1031. Cajal, S.R.Y. (1968) Traumatic degeneration and regeneration in the optic nerve and retina. In: Degeneration and Regeneration of the Nervous System. Translated by R.M. May. Hafner, New York, pp. 583-596. Caroni, P. and Schwab, M.E. (19881 Two membrane protein fractions from rat central myelin with inhibitory properties for neurit,egrowth and fibroblast spreading. J . Cell Biol., 106:1281-1288. Cohen, J., Burne, J., McKinlay, C., and Winter, J . (1987) The role of laminidfibronectin receptor complex in the outgrowth of retinal ganglion cell axons. Deu. Biol.,122:407418. Eldridge, C.F., Bunge, M.B., Bunge, R.P., and Wood, P.M. (1987) Differentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. J . Cell Biol.,105:1023-1034. Eldridge, C.F., Bunge, M.B., and Bunge, R.P. (19891 Differentiation of axon-related Schwann cells in vitro: 11. Control of myelin formation by basal lamina. J . Neurosci., 9:625-638. Fallon, J.R. (1985) Preferential outgrowth of central nervous system neurites on astrocytes and Schwann cells as compared to nonglial cells in vitro. J . Cell Biol., 100:198-207. Fawcett, J.W., Housden, E., Smith-Thomas, L., and Meyer, R.L. (1989) The growth of axons in three-dimensional astrocyte cultures. Deu. Biol., 135:449485.
300
BAEHR AND BUNGE
Ford-Holevinski, T.S., McCoy, J.P., Hopkins, J.M., and Agranoff, B.W. (1986) Laminin supports neurite outgrowth from explants of axotomized adult rat retinal neurons. Deu. Brain Res., 28:121-126. Giftochristos, N. and David, S.(1988) Laminin and heparan sulfate proteoglycan in the lesioned adult mammalian central nervous system and their possible relationship to axonal sprouting. J . Neurocytol., 17:385-397. Hall, S.M. and Kent, A.P. (1987) The response of regenerating peripheral axons to a grafted optic nerve. J . Neurocytol., 16:317-331. Hopkins, J.M., Schachner, M., and Bunge, R.P. (1989) Antibodies against the cell adhesion molecule L 1do not block growth ofadult rat retinal ganglion cell axons on Schwann cells. SOC.Neurosci. Abstr., 15:331. Kalderon, N. (1988) Differentiating astroglia in nervous tissue histogenesis regeneration: studies in a model system of regenerating peripheral nerve. J . Neurosci. Res., 21:501-512. Kleitman, N., Wood, P., Johnson, M.I., and Bunge, R.P. (1988a) Schwann cell surfaces but not extracellular matrix organized by Schwann cells support neurite outgrowth from embryonic rat retina. J . Neurosci., 8:653-663. Kleitman, N., Simon, D.K., Schachner, M., and Bunge, R.P. (1988b) Growt,h of embryonic retinal neurites elicited by contact with Schwann cell surfaces is blocked by antibodies to L1. Exp. Neurol., 102:298-306. Liesi, P., D. Dahl, and A. Vaheri (1983) Laminin is produced by early rat astrocytes in primary culture. J . Cell Biol., 96:920-924. Liesi, P. (1985)Laminin-immunoreactive glia distinguish regenerative adult CNS systems from non-regenerative ones. EMBO J., 4:2505-2511. Liesi, P. and Silver, J . (1988) Is astrocyte laminin involved in axon guidance in the mammalian CNS? Deu. B i d , 130:774-785. Liuzzi, F.J. and Lasek, R.J. (1987) Astrocytes block axonal regeneration in mammals by activating the physiological stop pathway. Science, 237:642-645. McCaffery, G.A., Raju, T.R., and Bennett, M.R. (1984) Effects of cultured astroglia on the survival of neonatal rat retinal ganglion cells in vitro. Dev. Biol.,104:661-668. McCarthy, K. and deVellis, J . (1980)Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J . Cell Biol., 85:890-902. Miller, R.H., Abney, E.R., David, S., ffrench-Constant, C., Lindsay, R., Patel, R., Stone, J., and Raff, M.C. (1986) Is reactive gliosis a property of a distinct subpopulation of astrocytes? J . Neuroscz., 6122-29. Morrison, R.S. and de Vellis, J . (1983) Differentiation of purified astrocytes in a chemically defined medium. Dev. Bruin Res., 9:337-345. Noble, M., Fok-Seang, J., and Cohen, J. (1984) Glia are a unique substrate for the in vitro growth of central nervous system neurons. J . Neurosci., 4:1892-1903.
Raff, M.C. (1989) Glial cell diversification in the rat optic nerve. Science, 243:1450-1455. Raff, M.C., Abney, E.R., Cohen, J., Lindsay, R., and Noble, M. (1983) Two types of astrocytes in cultures of developing white matter: Differences in morphology, surface ga ngliosides, and growth characteristics. J. Neurosci.. 3:1289-1300. Reier, P.J., Stensaas, L.J., and Guth, L. 1 1983)The astrocytic scar as an impediment to regeneration in the central nervous system. In: Spinal Cord Reconstruction. C.C. Kao, R.P. Bunge, and P.J. Reier, eds. Raven Press, New York, pp. 163--195. Schwab, M.E. and Thoenen, H. (1985) Dissociated neurons regenerate into sciatic but not optic nerve explmts in culture irrespective of neuronotrophic factors. J . Neurosci., 5:2415-2423. Schwab, M.E. and Caroni, P. (1988) 0lii:odendrocytes and CNS myelin are non-permissive substrates for Iieurite growth and fibroblast s reading in vitro. J . Neurosci., 8:2381-2393. SeiPheimer, B. and Schachner, M. (1987) Regulation of neural cell adhesion molecule expression on c u h r e d mouse Schwann cells by nerve growth factor. EMBO J., 6:1611-1616. Seilheimer, B. and Schachner, M. (19881 Studies of adhesion molecules mediating interactions between celle of peripheral nervous system indicate a major role for L1 in mediaXing sensory neuron growth on Schwann cells in culture. J. Cell Bid , 107:341-351. Sidman, R. and Rakic, P. (1973) Neuronal migration, with special reference to developing human brain a review. Bruin Res., 62:l-35. Silver, J. and Rutishauser, U. (1984) Guidance of axons in vivo by a preformed adhesive pathway on neuroepithelial endfeet. Deu. Biol., 106:485-499. Sobue, G. and Pleasure, D. (1984) Astriglial proliferation and phenotype are modulated by neuronal plasma membrane. Brain Res., 324: 175-179. Thanos, S., Vidal-Sanz, M., and Aguayo, A.J. (1987) The use of rhodamine-P-isothiocyanateRITC as a anterograde and retrograde tracer in the adult rat visual system. Brain Res., 406:317-321. Tomaselli, K.J., Neugebaum, U.M., Bixby, J.L., Lilien, J., and Reichardt, L.F. (1988) N-cadherin and integrins: two receptor systems that mediate neuronal process outgrowth on astrocyte surfaces. Neuron, 1:33-43. Vidal-Sanz, M., Villegas-Perez, M.P., Bray, G.M., and Aguayo, A.J. (1988) Persistent retromade labelin:: of adult rat retinal ganglion cells with the carbocya&e dye diI. &p. Neurol., 102:92-161. Villegas-Perez, M., Vidal-Sanz, M., Eiray, G.M., and Aguayo, A.J. (1988) Influences of peripheral neriie grafts on the survival and regrowth of axotomized retinal ganglion cells in adult rats. J . Neurosci., 8:265-280. Wood, P.M. (1976)Separation of functional Schwann cells and neurons from normal peripheral nerve tissue. Brain Res., 115:361-375.
-
