0306-4522/91 $3.00 + 0.00 Pergamon Press plc c; 1991 IBRO
Neuroscience Vol. 41, No. 213, pp. 695-701, 1991 Printed in Great Britain
EXPRESSION OF NEURONAL INTERMEDIATE FILAMENT PROTEINS ON, AND ON, DURING GOLDFISH OPTIC NERVE REGENERATION: EFFECT OF TECTAL ABLATION C. M. HALL*$§ and N. SCHECHTERt$ Departments of tBiochemistry and §Psychiatry and Behavioral Science, Health Sciences Center, State University of New York at Stony Brook, Stony Brook, NY 11794, U.S.A. Abstract--Goldfish retinal explants were used to study optic tectum participation in the regulation of intermediate filament protein synthesis in retinal ganglion cells during optic nerve regeneration. Retinas were explanted at various times after removal of the contralateral optic tectum. The synthesis of the intermediate filament proteins ON, and ON, in the cultures was quantitated by labeling with [‘5S]methionine, followed by two-dimensional gel electrophoresis, autoradiography, and densitometry. Neuritic growth from the explants was quantitated based on fiber length and density. In retinal explants placed in culture after 23 days of optic nerve regeneration, the synthesis of ON, and ON, was reduced when the tectum had been ablated. In contrast, synthesis of these proteins in explants placed in culture at an earlier stage of regeneration was not affected by tectal ablation. At all time points tested, neuritic outgrowth from retinal explants was stimulated by tectal ablation. These findings indicate that the synthesis of the ON, and ON, intermediate filament proteins during regeneration is not directly regulated by axonal volume. Further, our findings suggest that interaction between growing axons and tectum is important for sustained expression of these proteins during the later stages of optic nerve regeneration
CNS.‘.” During peripheral nerve regeneration, expression of peripherin is induced*’ while expression of neurofilament-L is down-regulated.” Axonal outgrowth, growth termination, and eventual synapse formation and maturation can be organized into growth stages along a time course of regulated molecular events. These stages can be biochemically characterized in terms of a differential expression of macromolecules that meet the physiological requirements of the developing axon at each stage of growth. In goldfish, the expression of the ON, and ON, proteins responds to optic nerve injury. During regeneration of the optic nerve, the synthesis of these proteins remains normal during the first five days of axonal outgrowth but a gradual increase in expression starts between five and 10 days after optic nerve crush, a time when the first optic fibers begin to reach the rostra1 optic tectum.2’,36 This time-course was observed both at the transcriptional38 and translationa133 levels. The gradual increase in expression of the ON, and ON, proteins during optic nerve regeneration reaches a maximum between 20 and 30 days after nerve crush, so that in the retina expression of mRNA for ON, and ON, is up-regulated to levels l&30-fold over normal after three weeks of regeneration. At this time the growing optic fibers exhibit vigorous collateral sprouting over the tectum, and both total optic fiber number and axonal volume is greatly elevated compared to norma1.9,25 Thus, regulatory mechanisms for expression of the ON, and ON, proteins persist for an extended period of time
goldfish visual pathway is a model system for neuronal development and regeneration. This pathway displays a continued growth and development throughout life, ‘x*~along with a remarkable capacity for functional regeneration after optic nerve injury. Within the retinal ganglion cells which give rise to the regenerating axons, the intermediate filaments expressed are different from the conventional neurofilament triplet proteins expressed in the ganglion cells of non-regenerating higher vertebrates.‘5.34 The predominant low molecular weight intermediate filament proteins in the axons of the goldfish optic nerve are two isoelectric variants of molecular weight 58,000, previously designated as ON, and ON, (optic nerve 58,000 mol. wt intermediate filament proteins).‘* These proteins are reminiscent of type III intermediate filament proteins, such as vimentin and peripherin, in that their expression has been linked to neuronal growth and development. Vimentin is transiently expressed before the neurofilament proteins during neuronal differentiation,‘J3’ while peripherin is co-expressed with the neurofilament triplet proteins in the mature PNS and in certain nuclei of the The
*Present address: Institute of Neurobiology, University of Goteborg, Box 33031, S-400, 33 Goteborg, Sweden. fTo whom correspondence should be addressed. Abbreviations: CHES, (2-IN-Cvclohexvlaminolethane_ _ sulfonic acid; EDTA, ethylenediaminetetra-acetate; HEPES, N-2-hvdroxvethvloincrazine-N’-2-ethanesulfonic acid; ON,,Z ,_two isoel&Z variants of optic nerve 58,000 mol. wt intermediate filament proteins; PBS, phosphate-buffered saline. 695
696
C. M. HALL and N. SCHECHTER
during regeneration. In contrast, increased expression of components such as actin, beta-tubulin, GAP-43. and other fast axonally transported proteins of unknown identity’~‘“~‘2~)“” is induced earlier during axonal regeneration. These observations indicate that separate regulatory mechanisms are in operation at different stages of the growth process. With respect to the growing axon, gene expression can be regulated both intrinsically (intra-axonal factors) and extrinsically (extra-axonal factors). The timing of changes in ON protein expression suggests either may be operating. For example, initially increased ON, and ON2 expression at the time of first tectal contact, and maximally increased expression with the later maximal axonal volume, may indicate tither an interaction between the growing axons and their tectal target, or that the larger axonal arbor volume per se stimulates ON, and ON, expression. To address this issue, we utilized the retinal explant system developed by Landreth and Agranoff’Y,20 in order to examine ON, and ON, expression under conditions where increased axonal arborization occurs in vitro in the absence of the tectum. In addition, we endeavored to determine the synthesis of the ON proteins and neuritic outgrowth in retinal explants prepared during nerve regeneration with and without prior
tectal
ablations.
EXPERIMENTAL PROCEDURES
Animuls
and surgical procedure.s
goldfish (Carassius uuratus, 5-8 cm) were obtained commercially from Mt. Parnell Fisheries, Mercersburg, PA, and maintained in 40-gallon tanks at ambient temperatures. After the fish were anesthetized in 0.05% tricaine methanesulfonate, a flap of the skull was opened over the optic tectum. In the control fish the left optic tract was transected. In experimental fish the lobes of the tectum were separated and the left tectum was removed by aspiration through a blunt-ended needle. The skull flap was replaced and sealed with petrolatum. Two days after surgery the fish were reanesthetized and retrobulbar crush was performed on the right optic nerve.
Common
Retinal explant
tissue culture
The culturing of retinal explants was performed essentially as described by Landreth and Agranoff 9.20with some modifications. Glass coverslips were cleaned with acetone, air-dried and incubated for 335 h in 2 mg/ml poly-L-lysine, approximately 52,000 mol. wt, in sterile 0.1 M sodium borate buffer (pH 8.4). After drying overnight, the coverslips were washed with 10 changes of sterile distilled water for a total of 6 h. The coverslips were attached to 35 mm sterile plastic Petri dishes (Nunc) by melting a spot of the culture dish plastic and dragging the melted plastic over the edge of the coverslip at a few points around its sides. The dishes were then welled with culture medium (Gibco L-15, supplemented with 20 mM HEPES, pH 7.2. 0.1 mg/ml 0.1 mM 5-fluoro-2’-deoxyuridine. gentamycin sulfate, 0.2 mM uridine and 5% fetal calf serum). Retinal explants from goldfish that had undergone tectal ablation/tract transection followed by optic nerve crush were prepared as described earlier.” The explants (24 retina pieces in each dish) were grown in humidity chambers at ambient temperature. New medium was added at 335 day intervals.
Quantitation
of neuritic growth
We employed a scoring procedure developed by Landreth and Agranoff.‘P Neurite density was scored on a scale of 14; the score 1 was given for explants with less than IO single neurites while the score 4 was given for explants with very high fiber densities (see Ref. 20 for details of the scoring procedure). The neurites grew both as single fibers and as thick fascicles, which was accounted for by the scoring procedure. The average fiber length for each explant was determined with an ocular micrometer to the nearest IOOnm. The neuritic growth index for each explant was obtained by multiplying the density score with the average neurite length. Radionctiue lubeling und two-dimensional
gel electrophoresi,r
of retinal explant proteins Before the explants were exposed to labeled methionine. after different times in culture, they were washed with 1ml of methionine-free L-15 medium and then incubated with I ml of methionine-free L-15 medium supplemented with 20 mM HEPES, pH 7.2, 0.1 mg/ml gentamycin sulfate. 0.1 mM 5-fluoro-2’-deoxyuridine, 0.2 mM uridine, 5% fetal calf serum dialysed against sterile phosphate-buffered saline (PBS). and 50 DCi f?Slmethionine (1100 Ciimmol, 9.7 mCi/ml ‘NEN Reskarch’Prbducts, Boston, MA). After 3 h, 7 nl of complete medium was added to give a final methionine concentration of 1 mg/ml. After 24 h the dishes were rinsed in 1ml of PBS and explants were harvested by flushing,and scraping the dish with a total of 1.5 ml cold PBS. After centrifugation for IO min at 14,000 r.p.m. at 4’C in an Eppendorf centrifuge (model 5415), the explant pellet was resuspended in 40~1 0.05 M CHES buffer (pH 9.0) containing 2% sodium dodecylsulfate. I % dithiothreotol and 5% glycerol and homogenized in a glass-glass tissue grinder. The homogenate was boiled for 2 min and rapidly frozen at -80°C. Labeled proteins from the retinal explants were separated by two-dimensional gel electrophoresis essentially as consisting of described by O’Farrell. ** First dimensions 4.5% acrylamide and 5% ampholines (PH 46: pH 557: pH 68, 3:4:3, LKB) were run in 2.5mm x 9.5cm tubes at 350 V overnight. Equivalent amounts (7 x 1O’c.p.m.) of trichloroacetic acid-precipitable radioactivity of each sample were loaded. Second dimension gels were 0.7 mm thick, 10% acrylamide resolving gels with a 5% stacking gel. Following electrophoresis, the slab gels were impregnated in 5% glycerol (in 10% acetic acid and 10% methanol) and dried in a Bio-Rad gel drier. Autoradiography was carried out by exposing the gel to Kodak X-OMAT-R film at room temperature for 5 days. Quantitative analysis of the ON, and ON, proteins was performed by scanning the autoradiographic film in an LKB Ultrascan XL laser densitometer using the LKB 2400 Gelscan SL software (LKB, Bromma, Sweden). In this analysis proteins ON, and ON, were combined because it is not known with certainty if the two proteins are different at the amino acid level or if ON? is a phosphorylated product of ON,. In all experiments the levels of the two proteins were coordinately linked. In addition, the labeling of ON, and ON, was normalized for differences in total amount of radioactivity on different gels by selecting a major, non-variant polypeptide as a reference standard. The amount of ON, and ON, synthesized in the explants during the 24 h labeling period was expressed as the sum of the optic densities of the ON, and ON, spots divided by the optic density of the reference spot. Immunoprecipitation Two culture dishes of retinal explants labelled during their first 24 h in culture were harvested and homogenized in CHES buffer as described under “sample preparation”. This sample (50 ~1) was dialysed extensively against 20 mM Tris-HCI (PH 7.4 containing 0.3% sodium dodecylsulfate). The dialysed sample was added to 150~1 of
Neuronal intermediate filament expression in retinal explants immunoprecipitation buffer (20 mM Tris-HCl, pH 7.4; 100mM NaCl; 5 mM EDTA; 1% deoxycholate; 1% Triton X-100; 0.2% sodium dodecylsulfate) and 10 ~1 of a 50% Protein A-Sepharose CL4B suspension (Pharmacia, Uppsala, Sweden) and 10~1 rabbit preimmune serum were added. The mixture was agitated at room temperature for 2 h and subsequently centrifuged. Ten microliters of anti-ONJON, serum was added to the supematant.i6 After an overnight incubation at 4”C, the immune complexes were precipitated at room temperature with 10 ~1 of 50% Protein A-Sepharose for 2 h with agitation.‘s After centrifugation, the pellet was washed twice with immunoprecipitation buffer. The Sepharose beads were resuspended in 40 ~1 CHES buffer and the immunoprecipitate solubilized by heating at 100°C for 2 min. This suspension was filtered through glass-wool and frozen at -80” until electrophoresis. RESULTS Synthesis of ON, and ON, proteins
To identify the ON, and ON, proteins on twodimensional gels, cultures were labeled with [35S]methionine, solubilized and immunoprecipitated with an antiserum raised against the ON, and ON, proteins. I6The immunoprecipitate was subjected to two-dimensional gel electrophoresis and autoradiography. The two labeled spots on the resulting autoradiograph co-migrated with the intermediate filament proteins from the goldfish optic nerve and retina previously designated as ON, and ON, (Fig. I). To address the question of whether the optic tectum provides regulatory signals that modulate the level of synthesis of the ON, and ON, proteins, the regenerating axons were prevented from interacting with their principal target by ablating the contralateral optic tectum prior to optic nerve crush. In control experiments the contralateral optic tract was transected prior to optic nerve crush. The synthesis of ON, and ON, was monitored in retinal explants cultured at five and 23 days after surgery. These times were chosen because they represent growth within the optic nerve (five days postcrush) vs growth over the optic tectum (23 days postcrush) respectively. To
691
investigate if an increasing quantity of axons would induce increased synthesis of ON, and ON,, the retinal explants were kept in culture for up to 11 days and the synthesis of ON, and ON, in the cultures was determined at different time points by densitometric quantitation of the relative incorporation of [35S]methionine into these proteins on autoradiographs of two-dimensional electrophoretic gels. The explants were exposed to [35S]methionine for 24 h because the ON, and ON, proteins are only expressed in a minority of cells present in retinal explants and therefore labeling of these proteins is negligible if the explants are labeled for short time periods. The turnover of intermediate filament proteins is very slow, therefore the 24 h exposure increased the relative labeling of ON, and ON, compared to other proteins expressed in these cultures. In all experimental conditions the synthesis of the ON, and ON, proteins was maximal during the first day in culture. With increasing time in culture their synthesis gradually decreased to barely detectable levels (Fig. 2A, B). During this time period of declining ON, and ON, synthesis there was a continuous outgrowth of neurites from the explants. In cultures of retina that had been previously regenerating their optic axons with the tectum left intact, the relative expression of ON, and ON, in the cultures varied with the time interval between nerve crush and retinal explanting. During the first day in culture the expression of ON, plus ON, was 65% higher in retinas explanted 23 days, as compared to five days, after optic nerve crush (Figs 2A, B). In cultures of retina that had been previously regenerating their optic axons for 23 days before explantation, the amount of ON, and ON, synthesized in the cultures was strongly affected by tectal ablation. In these experiments, with the tectum left intact, the synthesis of ON, and ON, was prominent during the first (Fig. 3A) and second (Fig. 3C) days in culture. In contrast, when the tectum had been ablated there was normal labeling of ON, in the explants during the first day in culture but no detectable synthesis of ON, was observed (Fig. 3B). During the second day in culture and all subsequent days, neither ON, nor ON, was expressed at detectable levels (Figs 2B and 3D). In cultures explanted five days after optic nerve crush, the absence or presence of the optic tectum during axonal regeneration did not significantly affect the levels of ON, and ON_, synthesized in the explants (Fig. 2A). Neuritic growth
Fig. 1. Identification of ON, and ON, by immunoprecipitation and two-dimensional gel electrophoresis. Autoradiograph shows ON, and ON, immunoprecipitated from a total protein extract of retinal explants labeled with [?3]methionine.
Prior optic nerve crush induces vigorous neuritic outgrowth from explanted goldfish retinas.‘9,2o Axonal backfilling experiments have shown that the growing neurites emanate from retinal ganglion ~~11s.‘~ We measured the extent of neuritic outgrowth from explanted retinas to establish whether the reduced synthesis of the ON, and ON, proteins after tectal ablation was attributable to a decreased growth
C. M. HALL and N. SCHECHTEX Retinas explanted 5 days post-crush
0
2
-
control
-
tectal ablation
4
10
8
6
12
DISCUSSION
days
B
++
Retinas explanted 23 days post-crush 1.2
1.0 I-
-
control tectal ablation
0.8
0.6
0.4
0.2
0.0 f I,
1
2
3
The extent of neuritic growth from explanted retina varies with the interval of time between optic nerve crush and explanting (Table 1; Ref. 20). Maxima1 growth is obtained from cultures explanted IO-15 days after optic nerve crush. Cultures explanted after a longer (32 days, Table 1) or shorter (data not shown) postcrush interval exhibit less neuritic growth. Removal of the contralateral optic tectum increased both the average neuritic density and average length of the neurites so that the neuritic growth index was increased by 30% (IO days postcrush), 58% (15 days postcrush), and 53% (32 days postcrush), as compared to transection of the contralateral optic tract (Table 1). These differences were statistically significant (P < 0.005. Student’s two-tailed t-test).
4
5
6
7
Days in culture Fig. 2. Time-course of synthesis of ON, and ON, in retinal explants after tectal ablation. The optic nerve was crushed two days after removal of the contralateral optic tectum (closed diamonds) or transection of the contralateral optic tract (open squares). Retinas were explanted five days (A) or 23 days (B) after crush and labeled with [‘5S]methionine after different times in culture. The relative synthesis of ON, and ON, was determined by densitometric scanning of autoradiographs prepared from two-dimensional electrophoretic gels. Bars indicate f S.E.M. of duplicate samples.
of neurites from the explants under these conditions. The first neurites could be detected within 24 h following explantation and growth of neurites. both in length and number, continued for several weeks in culture. The neuritic growth index shown in Table I was measured after four days in culture. This represents a time of intense neuritic growth. After longer times in culture, the increasing length and complexity of the neurites make quantitation difficult.
The principal finding in this study is that tectal ablation dramatically reduces the synthesis of ON, and ON, in retinal explants cultured 23 days after induction of regeneration by optic nerve crush. This decreased synthesis occurred despite vigorous neuritic outgrowth. Thus, the presence of their target is important for a sustained expression of these proteins in retinal ganglion cells, and suggests that extrinsic regulation of the ON proteins occurs in this system. Earlier studies in the goldfish have demonstrated both positive and negative regulatory influences of the optic tectum on the metabolism and morphology of retinal ganglion cells. Crushing of the optic nerve induces general as well as specific increases in RNA and protein synthesis, along with cell body enlargement.‘.“‘.26.3’ These events begin to diminish at the time when the ganglion cell axons reach the tectum.22.26 Removal of the optic tectum prolongs the cell body reaction,4 prevents return to normal levels of synthesis of some of the proteins whose synthesis is increased at an early stage of regeneration and prevents increased synthesis of proteins normally induced at a later stage during regeneration.‘,6,39 The present experiments show that neuritic growth is stimulated in retinal explants cultured after tectal ablation. This finding is in agreement with a previous study of the growth characteristics of goldfish retinal explants cultured after deprivation of their targeLz4 The stimulated neuritic growth activity resulting from lack of contact with the correct target tissue was accompanied by a decreased synthesis of ON, and ON,. This finding suggests that the increased synthesis of ON, and ON, during the later stages of optic nerve regeneration in oivo33~3s is not simply due to intra-axonal factors associated with the increasing axonal volume due to collateral sprouting over the optic tectum.‘,*’ This view is further supported by the time-course of expression of the ON, and ON, proteins in culture. Neurites from the explanted retina grew at a rapid
Neuronal
intermediate
filament
expression
in retinal
explants
699
Fig. 3. Tectal ablation decreases the synthesis of ON, and ON, in retinal explants. The optic nerve was crushed 2 days after removal of the contralateral optic tectum (tectal ablation) or after transection of the contralateral optic tract (control). Retinas were explanted 23 days after the crush. Autoradiographs of two-dimensional gels of total proteins labeled with @.]methionine during the first (A, B) and second (C, D) days in culture. The left panel (A, C) shows labeled proteins extracted from control explants. The right panel (B, D) shows labeled proteins extracted from explants cultured after tectal ablation. R indicates a non-variant polypeptide used as a reference for quantitation of ON, and ON, synthesis. The square bracket shows the position of the ON, (left) and ON, (right) proteins.
rate
for
2 weeksss”
while
the
synthesis
of ON,
and
the first week in culture. In regenerating goldfish retina, the ON, and ON, proteins are expressed predominantly by retinal ON,
decreased
by
90%
during
Table 1.Neuritic growth index of retinal explants different times after tectal ablation compared explants Days after crush 10 I5 32
NC1 of control 961 k 57 (42) 920 f 73 (57) 410 & 43 (68)
cultured at to control
NGI after tectal ablation 1257 k 90 (47) 1460 + 94 (55) 629 + 54 (74)
Explants were prepared from control (transection of the contralateral optic tract) or experimental (removal of the contralateral optic tectum) retinas 10, 15, and 32 days after optic nerve crush and the neuritic growth index was calculated after four days in culture. Number of explants in parentheses. Mean & S.E.M. NGI, neuritic growth index.
ganglion cells. Small amounts are also synthesized by cells in the inner nuclear layer and by Muller cells, which seem to co-express these intermediate filament proteins with glial fibrillary acidic protein.“.” The profound decline in the synthesis of the ON, and ON, proteins with increasing time in explant culture therefore reflects changed expression in the retinal ganglion cells and not changes in expression or death of other retinal cell types. Furthermore, the declining synthesis was not due to a decreasing viability of the retinal ganglion cells because the number and morphological appearance of these cells in explant culture remain stable during two weeks of culture.‘4 In cultures explanted five days after optic nerve crush, the absence or presence of the optic tectum during axonal regeneration did not change the level of ON, and ON, synthesized in the explants. A similar observation was also reported by Benowitz et al.,’
C. M. HALL and
700
where goldfish optic nerves regenerating with and without the tecta present showed marked differences in the pattern of axonally transported proteins at 24 days but minimal differences at 10 days after nerve crush. During normal regeneration the expression of the ON, and ON, proteins remains at precrush levels five days after crush.” This observation shows that ON, and ON, expression can be maintained, in the absence of tectal contact, during a window of time after nerve crush. However, if appropriate target contact is not restored after a certain interval of time, expression may decrease. CONCLUSION
We propose that the synthesis of the ON, and ON> proteins is regulated, at least in part. by molecular signals arising from interactions between optic axons
and the tectum. Although the cellular origin or nature of the hypothesized signal that may regulate ON, and ON2 protein expression is unknown, our findings indicate that it may have a slow turnover, allowing for its availability in high enough concentration to maintain normal ON, and ON, synthesis until tectal contact is re-established. Since the volume of optic fibers present in the tectum increases during regeneration up to six-fold over normal at 335 weeks after nerve crush,9,25 one possibility may be that the amount of this proposed regulatory signal reaching the retina is related to the number of optic fibers in contact with the optic tectum. il~knowlrdgements--This work was supported by a postdoctoral fellowship from the Swedish Natural Science Research Council to C.H.. The Lars Hierta Foundation (C.H.), and a grant from the National Institutes of Health (EY 05212) to N.S.
REFERENCES 1. Benowitz
L. I., Shashoua V. E. and Yoon M. G. (1981) Specific changes in rapidly transported proteins during regeneration of the goldfish optic nerve. J. Neurosci. 1, 300-307. 2. Benowitz L., Yoon M. and Lewis E. (1983) Transported proteins in the regenerating optic nerve: regulation by interactions with the optic tectum. Science 222, 1855188. 3. Bignami A., Raju T. and Dahl D. (1982) Localization of vimentin, the non-specific intermediate filament protein in embrvonal glia and in early differentiating neurons. Deal Biol. 91, 286-295. Burmeister b. W. and Grafstein B. (198
Neuroscience Vol. 41, No. 213, pp. 695-701, 1991 Printed in Great Britain
EXPRESSION OF NEURONAL INTERMEDIATE FILAMENT PROTEINS ON, AND ON, DURING GOLDFISH OPTIC NERVE REGENERATION: EFFECT OF TECTAL ABLATION C. M. HALL*$§ and N. SCHECHTERt$ Departments of tBiochemistry and §Psychiatry and Behavioral Science, Health Sciences Center, State University of New York at Stony Brook, Stony Brook, NY 11794, U.S.A. Abstract--Goldfish retinal explants were used to study optic tectum participation in the regulation of intermediate filament protein synthesis in retinal ganglion cells during optic nerve regeneration. Retinas were explanted at various times after removal of the contralateral optic tectum. The synthesis of the intermediate filament proteins ON, and ON, in the cultures was quantitated by labeling with [‘5S]methionine, followed by two-dimensional gel electrophoresis, autoradiography, and densitometry. Neuritic growth from the explants was quantitated based on fiber length and density. In retinal explants placed in culture after 23 days of optic nerve regeneration, the synthesis of ON, and ON, was reduced when the tectum had been ablated. In contrast, synthesis of these proteins in explants placed in culture at an earlier stage of regeneration was not affected by tectal ablation. At all time points tested, neuritic outgrowth from retinal explants was stimulated by tectal ablation. These findings indicate that the synthesis of the ON, and ON, intermediate filament proteins during regeneration is not directly regulated by axonal volume. Further, our findings suggest that interaction between growing axons and tectum is important for sustained expression of these proteins during the later stages of optic nerve regeneration
CNS.‘.” During peripheral nerve regeneration, expression of peripherin is induced*’ while expression of neurofilament-L is down-regulated.” Axonal outgrowth, growth termination, and eventual synapse formation and maturation can be organized into growth stages along a time course of regulated molecular events. These stages can be biochemically characterized in terms of a differential expression of macromolecules that meet the physiological requirements of the developing axon at each stage of growth. In goldfish, the expression of the ON, and ON, proteins responds to optic nerve injury. During regeneration of the optic nerve, the synthesis of these proteins remains normal during the first five days of axonal outgrowth but a gradual increase in expression starts between five and 10 days after optic nerve crush, a time when the first optic fibers begin to reach the rostra1 optic tectum.2’,36 This time-course was observed both at the transcriptional38 and translationa133 levels. The gradual increase in expression of the ON, and ON, proteins during optic nerve regeneration reaches a maximum between 20 and 30 days after nerve crush, so that in the retina expression of mRNA for ON, and ON, is up-regulated to levels l&30-fold over normal after three weeks of regeneration. At this time the growing optic fibers exhibit vigorous collateral sprouting over the tectum, and both total optic fiber number and axonal volume is greatly elevated compared to norma1.9,25 Thus, regulatory mechanisms for expression of the ON, and ON, proteins persist for an extended period of time
goldfish visual pathway is a model system for neuronal development and regeneration. This pathway displays a continued growth and development throughout life, ‘x*~along with a remarkable capacity for functional regeneration after optic nerve injury. Within the retinal ganglion cells which give rise to the regenerating axons, the intermediate filaments expressed are different from the conventional neurofilament triplet proteins expressed in the ganglion cells of non-regenerating higher vertebrates.‘5.34 The predominant low molecular weight intermediate filament proteins in the axons of the goldfish optic nerve are two isoelectric variants of molecular weight 58,000, previously designated as ON, and ON, (optic nerve 58,000 mol. wt intermediate filament proteins).‘* These proteins are reminiscent of type III intermediate filament proteins, such as vimentin and peripherin, in that their expression has been linked to neuronal growth and development. Vimentin is transiently expressed before the neurofilament proteins during neuronal differentiation,‘J3’ while peripherin is co-expressed with the neurofilament triplet proteins in the mature PNS and in certain nuclei of the The
*Present address: Institute of Neurobiology, University of Goteborg, Box 33031, S-400, 33 Goteborg, Sweden. fTo whom correspondence should be addressed. Abbreviations: CHES, (2-IN-Cvclohexvlaminolethane_ _ sulfonic acid; EDTA, ethylenediaminetetra-acetate; HEPES, N-2-hvdroxvethvloincrazine-N’-2-ethanesulfonic acid; ON,,Z ,_two isoel&Z variants of optic nerve 58,000 mol. wt intermediate filament proteins; PBS, phosphate-buffered saline. 695
696
C. M. HALL and N. SCHECHTER
during regeneration. In contrast, increased expression of components such as actin, beta-tubulin, GAP-43. and other fast axonally transported proteins of unknown identity’~‘“~‘2~)“” is induced earlier during axonal regeneration. These observations indicate that separate regulatory mechanisms are in operation at different stages of the growth process. With respect to the growing axon, gene expression can be regulated both intrinsically (intra-axonal factors) and extrinsically (extra-axonal factors). The timing of changes in ON protein expression suggests either may be operating. For example, initially increased ON, and ON2 expression at the time of first tectal contact, and maximally increased expression with the later maximal axonal volume, may indicate tither an interaction between the growing axons and their tectal target, or that the larger axonal arbor volume per se stimulates ON, and ON, expression. To address this issue, we utilized the retinal explant system developed by Landreth and Agranoff’Y,20 in order to examine ON, and ON, expression under conditions where increased axonal arborization occurs in vitro in the absence of the tectum. In addition, we endeavored to determine the synthesis of the ON proteins and neuritic outgrowth in retinal explants prepared during nerve regeneration with and without prior
tectal
ablations.
EXPERIMENTAL PROCEDURES
Animuls
and surgical procedure.s
goldfish (Carassius uuratus, 5-8 cm) were obtained commercially from Mt. Parnell Fisheries, Mercersburg, PA, and maintained in 40-gallon tanks at ambient temperatures. After the fish were anesthetized in 0.05% tricaine methanesulfonate, a flap of the skull was opened over the optic tectum. In the control fish the left optic tract was transected. In experimental fish the lobes of the tectum were separated and the left tectum was removed by aspiration through a blunt-ended needle. The skull flap was replaced and sealed with petrolatum. Two days after surgery the fish were reanesthetized and retrobulbar crush was performed on the right optic nerve.
Common
Retinal explant
tissue culture
The culturing of retinal explants was performed essentially as described by Landreth and Agranoff 9.20with some modifications. Glass coverslips were cleaned with acetone, air-dried and incubated for 335 h in 2 mg/ml poly-L-lysine, approximately 52,000 mol. wt, in sterile 0.1 M sodium borate buffer (pH 8.4). After drying overnight, the coverslips were washed with 10 changes of sterile distilled water for a total of 6 h. The coverslips were attached to 35 mm sterile plastic Petri dishes (Nunc) by melting a spot of the culture dish plastic and dragging the melted plastic over the edge of the coverslip at a few points around its sides. The dishes were then welled with culture medium (Gibco L-15, supplemented with 20 mM HEPES, pH 7.2. 0.1 mg/ml 0.1 mM 5-fluoro-2’-deoxyuridine. gentamycin sulfate, 0.2 mM uridine and 5% fetal calf serum). Retinal explants from goldfish that had undergone tectal ablation/tract transection followed by optic nerve crush were prepared as described earlier.” The explants (24 retina pieces in each dish) were grown in humidity chambers at ambient temperature. New medium was added at 335 day intervals.
Quantitation
of neuritic growth
We employed a scoring procedure developed by Landreth and Agranoff.‘P Neurite density was scored on a scale of 14; the score 1 was given for explants with less than IO single neurites while the score 4 was given for explants with very high fiber densities (see Ref. 20 for details of the scoring procedure). The neurites grew both as single fibers and as thick fascicles, which was accounted for by the scoring procedure. The average fiber length for each explant was determined with an ocular micrometer to the nearest IOOnm. The neuritic growth index for each explant was obtained by multiplying the density score with the average neurite length. Radionctiue lubeling und two-dimensional
gel electrophoresi,r
of retinal explant proteins Before the explants were exposed to labeled methionine. after different times in culture, they were washed with 1ml of methionine-free L-15 medium and then incubated with I ml of methionine-free L-15 medium supplemented with 20 mM HEPES, pH 7.2, 0.1 mg/ml gentamycin sulfate. 0.1 mM 5-fluoro-2’-deoxyuridine, 0.2 mM uridine, 5% fetal calf serum dialysed against sterile phosphate-buffered saline (PBS). and 50 DCi f?Slmethionine (1100 Ciimmol, 9.7 mCi/ml ‘NEN Reskarch’Prbducts, Boston, MA). After 3 h, 7 nl of complete medium was added to give a final methionine concentration of 1 mg/ml. After 24 h the dishes were rinsed in 1ml of PBS and explants were harvested by flushing,and scraping the dish with a total of 1.5 ml cold PBS. After centrifugation for IO min at 14,000 r.p.m. at 4’C in an Eppendorf centrifuge (model 5415), the explant pellet was resuspended in 40~1 0.05 M CHES buffer (pH 9.0) containing 2% sodium dodecylsulfate. I % dithiothreotol and 5% glycerol and homogenized in a glass-glass tissue grinder. The homogenate was boiled for 2 min and rapidly frozen at -80°C. Labeled proteins from the retinal explants were separated by two-dimensional gel electrophoresis essentially as consisting of described by O’Farrell. ** First dimensions 4.5% acrylamide and 5% ampholines (PH 46: pH 557: pH 68, 3:4:3, LKB) were run in 2.5mm x 9.5cm tubes at 350 V overnight. Equivalent amounts (7 x 1O’c.p.m.) of trichloroacetic acid-precipitable radioactivity of each sample were loaded. Second dimension gels were 0.7 mm thick, 10% acrylamide resolving gels with a 5% stacking gel. Following electrophoresis, the slab gels were impregnated in 5% glycerol (in 10% acetic acid and 10% methanol) and dried in a Bio-Rad gel drier. Autoradiography was carried out by exposing the gel to Kodak X-OMAT-R film at room temperature for 5 days. Quantitative analysis of the ON, and ON, proteins was performed by scanning the autoradiographic film in an LKB Ultrascan XL laser densitometer using the LKB 2400 Gelscan SL software (LKB, Bromma, Sweden). In this analysis proteins ON, and ON, were combined because it is not known with certainty if the two proteins are different at the amino acid level or if ON? is a phosphorylated product of ON,. In all experiments the levels of the two proteins were coordinately linked. In addition, the labeling of ON, and ON, was normalized for differences in total amount of radioactivity on different gels by selecting a major, non-variant polypeptide as a reference standard. The amount of ON, and ON, synthesized in the explants during the 24 h labeling period was expressed as the sum of the optic densities of the ON, and ON, spots divided by the optic density of the reference spot. Immunoprecipitation Two culture dishes of retinal explants labelled during their first 24 h in culture were harvested and homogenized in CHES buffer as described under “sample preparation”. This sample (50 ~1) was dialysed extensively against 20 mM Tris-HCI (PH 7.4 containing 0.3% sodium dodecylsulfate). The dialysed sample was added to 150~1 of
Neuronal intermediate filament expression in retinal explants immunoprecipitation buffer (20 mM Tris-HCl, pH 7.4; 100mM NaCl; 5 mM EDTA; 1% deoxycholate; 1% Triton X-100; 0.2% sodium dodecylsulfate) and 10 ~1 of a 50% Protein A-Sepharose CL4B suspension (Pharmacia, Uppsala, Sweden) and 10~1 rabbit preimmune serum were added. The mixture was agitated at room temperature for 2 h and subsequently centrifuged. Ten microliters of anti-ONJON, serum was added to the supematant.i6 After an overnight incubation at 4”C, the immune complexes were precipitated at room temperature with 10 ~1 of 50% Protein A-Sepharose for 2 h with agitation.‘s After centrifugation, the pellet was washed twice with immunoprecipitation buffer. The Sepharose beads were resuspended in 40 ~1 CHES buffer and the immunoprecipitate solubilized by heating at 100°C for 2 min. This suspension was filtered through glass-wool and frozen at -80” until electrophoresis. RESULTS Synthesis of ON, and ON, proteins
To identify the ON, and ON, proteins on twodimensional gels, cultures were labeled with [35S]methionine, solubilized and immunoprecipitated with an antiserum raised against the ON, and ON, proteins. I6The immunoprecipitate was subjected to two-dimensional gel electrophoresis and autoradiography. The two labeled spots on the resulting autoradiograph co-migrated with the intermediate filament proteins from the goldfish optic nerve and retina previously designated as ON, and ON, (Fig. I). To address the question of whether the optic tectum provides regulatory signals that modulate the level of synthesis of the ON, and ON, proteins, the regenerating axons were prevented from interacting with their principal target by ablating the contralateral optic tectum prior to optic nerve crush. In control experiments the contralateral optic tract was transected prior to optic nerve crush. The synthesis of ON, and ON, was monitored in retinal explants cultured at five and 23 days after surgery. These times were chosen because they represent growth within the optic nerve (five days postcrush) vs growth over the optic tectum (23 days postcrush) respectively. To
691
investigate if an increasing quantity of axons would induce increased synthesis of ON, and ON,, the retinal explants were kept in culture for up to 11 days and the synthesis of ON, and ON, in the cultures was determined at different time points by densitometric quantitation of the relative incorporation of [35S]methionine into these proteins on autoradiographs of two-dimensional electrophoretic gels. The explants were exposed to [35S]methionine for 24 h because the ON, and ON, proteins are only expressed in a minority of cells present in retinal explants and therefore labeling of these proteins is negligible if the explants are labeled for short time periods. The turnover of intermediate filament proteins is very slow, therefore the 24 h exposure increased the relative labeling of ON, and ON, compared to other proteins expressed in these cultures. In all experimental conditions the synthesis of the ON, and ON, proteins was maximal during the first day in culture. With increasing time in culture their synthesis gradually decreased to barely detectable levels (Fig. 2A, B). During this time period of declining ON, and ON, synthesis there was a continuous outgrowth of neurites from the explants. In cultures of retina that had been previously regenerating their optic axons with the tectum left intact, the relative expression of ON, and ON, in the cultures varied with the time interval between nerve crush and retinal explanting. During the first day in culture the expression of ON, plus ON, was 65% higher in retinas explanted 23 days, as compared to five days, after optic nerve crush (Figs 2A, B). In cultures of retina that had been previously regenerating their optic axons for 23 days before explantation, the amount of ON, and ON, synthesized in the cultures was strongly affected by tectal ablation. In these experiments, with the tectum left intact, the synthesis of ON, and ON, was prominent during the first (Fig. 3A) and second (Fig. 3C) days in culture. In contrast, when the tectum had been ablated there was normal labeling of ON, in the explants during the first day in culture but no detectable synthesis of ON, was observed (Fig. 3B). During the second day in culture and all subsequent days, neither ON, nor ON, was expressed at detectable levels (Figs 2B and 3D). In cultures explanted five days after optic nerve crush, the absence or presence of the optic tectum during axonal regeneration did not significantly affect the levels of ON, and ON_, synthesized in the explants (Fig. 2A). Neuritic growth
Fig. 1. Identification of ON, and ON, by immunoprecipitation and two-dimensional gel electrophoresis. Autoradiograph shows ON, and ON, immunoprecipitated from a total protein extract of retinal explants labeled with [?3]methionine.
Prior optic nerve crush induces vigorous neuritic outgrowth from explanted goldfish retinas.‘9,2o Axonal backfilling experiments have shown that the growing neurites emanate from retinal ganglion ~~11s.‘~ We measured the extent of neuritic outgrowth from explanted retinas to establish whether the reduced synthesis of the ON, and ON, proteins after tectal ablation was attributable to a decreased growth
C. M. HALL and N. SCHECHTEX Retinas explanted 5 days post-crush
0
2
-
control
-
tectal ablation
4
10
8
6
12
DISCUSSION
days
B
++
Retinas explanted 23 days post-crush 1.2
1.0 I-
-
control tectal ablation
0.8
0.6
0.4
0.2
0.0 f I,
1
2
3
The extent of neuritic growth from explanted retina varies with the interval of time between optic nerve crush and explanting (Table 1; Ref. 20). Maxima1 growth is obtained from cultures explanted IO-15 days after optic nerve crush. Cultures explanted after a longer (32 days, Table 1) or shorter (data not shown) postcrush interval exhibit less neuritic growth. Removal of the contralateral optic tectum increased both the average neuritic density and average length of the neurites so that the neuritic growth index was increased by 30% (IO days postcrush), 58% (15 days postcrush), and 53% (32 days postcrush), as compared to transection of the contralateral optic tract (Table 1). These differences were statistically significant (P < 0.005. Student’s two-tailed t-test).
4
5
6
7
Days in culture Fig. 2. Time-course of synthesis of ON, and ON, in retinal explants after tectal ablation. The optic nerve was crushed two days after removal of the contralateral optic tectum (closed diamonds) or transection of the contralateral optic tract (open squares). Retinas were explanted five days (A) or 23 days (B) after crush and labeled with [‘5S]methionine after different times in culture. The relative synthesis of ON, and ON, was determined by densitometric scanning of autoradiographs prepared from two-dimensional electrophoretic gels. Bars indicate f S.E.M. of duplicate samples.
of neurites from the explants under these conditions. The first neurites could be detected within 24 h following explantation and growth of neurites. both in length and number, continued for several weeks in culture. The neuritic growth index shown in Table I was measured after four days in culture. This represents a time of intense neuritic growth. After longer times in culture, the increasing length and complexity of the neurites make quantitation difficult.
The principal finding in this study is that tectal ablation dramatically reduces the synthesis of ON, and ON, in retinal explants cultured 23 days after induction of regeneration by optic nerve crush. This decreased synthesis occurred despite vigorous neuritic outgrowth. Thus, the presence of their target is important for a sustained expression of these proteins in retinal ganglion cells, and suggests that extrinsic regulation of the ON proteins occurs in this system. Earlier studies in the goldfish have demonstrated both positive and negative regulatory influences of the optic tectum on the metabolism and morphology of retinal ganglion cells. Crushing of the optic nerve induces general as well as specific increases in RNA and protein synthesis, along with cell body enlargement.‘.“‘.26.3’ These events begin to diminish at the time when the ganglion cell axons reach the tectum.22.26 Removal of the optic tectum prolongs the cell body reaction,4 prevents return to normal levels of synthesis of some of the proteins whose synthesis is increased at an early stage of regeneration and prevents increased synthesis of proteins normally induced at a later stage during regeneration.‘,6,39 The present experiments show that neuritic growth is stimulated in retinal explants cultured after tectal ablation. This finding is in agreement with a previous study of the growth characteristics of goldfish retinal explants cultured after deprivation of their targeLz4 The stimulated neuritic growth activity resulting from lack of contact with the correct target tissue was accompanied by a decreased synthesis of ON, and ON,. This finding suggests that the increased synthesis of ON, and ON, during the later stages of optic nerve regeneration in oivo33~3s is not simply due to intra-axonal factors associated with the increasing axonal volume due to collateral sprouting over the optic tectum.‘,*’ This view is further supported by the time-course of expression of the ON, and ON, proteins in culture. Neurites from the explanted retina grew at a rapid
Neuronal
intermediate
filament
expression
in retinal
explants
699
Fig. 3. Tectal ablation decreases the synthesis of ON, and ON, in retinal explants. The optic nerve was crushed 2 days after removal of the contralateral optic tectum (tectal ablation) or after transection of the contralateral optic tract (control). Retinas were explanted 23 days after the crush. Autoradiographs of two-dimensional gels of total proteins labeled with @.]methionine during the first (A, B) and second (C, D) days in culture. The left panel (A, C) shows labeled proteins extracted from control explants. The right panel (B, D) shows labeled proteins extracted from explants cultured after tectal ablation. R indicates a non-variant polypeptide used as a reference for quantitation of ON, and ON, synthesis. The square bracket shows the position of the ON, (left) and ON, (right) proteins.
rate
for
2 weeksss”
while
the
synthesis
of ON,
and
the first week in culture. In regenerating goldfish retina, the ON, and ON, proteins are expressed predominantly by retinal ON,
decreased
by
90%
during
Table 1.Neuritic growth index of retinal explants different times after tectal ablation compared explants Days after crush 10 I5 32
NC1 of control 961 k 57 (42) 920 f 73 (57) 410 & 43 (68)
cultured at to control
NGI after tectal ablation 1257 k 90 (47) 1460 + 94 (55) 629 + 54 (74)
Explants were prepared from control (transection of the contralateral optic tract) or experimental (removal of the contralateral optic tectum) retinas 10, 15, and 32 days after optic nerve crush and the neuritic growth index was calculated after four days in culture. Number of explants in parentheses. Mean & S.E.M. NGI, neuritic growth index.
ganglion cells. Small amounts are also synthesized by cells in the inner nuclear layer and by Muller cells, which seem to co-express these intermediate filament proteins with glial fibrillary acidic protein.“.” The profound decline in the synthesis of the ON, and ON, proteins with increasing time in explant culture therefore reflects changed expression in the retinal ganglion cells and not changes in expression or death of other retinal cell types. Furthermore, the declining synthesis was not due to a decreasing viability of the retinal ganglion cells because the number and morphological appearance of these cells in explant culture remain stable during two weeks of culture.‘4 In cultures explanted five days after optic nerve crush, the absence or presence of the optic tectum during axonal regeneration did not change the level of ON, and ON, synthesized in the explants. A similar observation was also reported by Benowitz et al.,’
C. M. HALL and
700
where goldfish optic nerves regenerating with and without the tecta present showed marked differences in the pattern of axonally transported proteins at 24 days but minimal differences at 10 days after nerve crush. During normal regeneration the expression of the ON, and ON, proteins remains at precrush levels five days after crush.” This observation shows that ON, and ON, expression can be maintained, in the absence of tectal contact, during a window of time after nerve crush. However, if appropriate target contact is not restored after a certain interval of time, expression may decrease. CONCLUSION
We propose that the synthesis of the ON, and ON> proteins is regulated, at least in part. by molecular signals arising from interactions between optic axons
and the tectum. Although the cellular origin or nature of the hypothesized signal that may regulate ON, and ON2 protein expression is unknown, our findings indicate that it may have a slow turnover, allowing for its availability in high enough concentration to maintain normal ON, and ON, synthesis until tectal contact is re-established. Since the volume of optic fibers present in the tectum increases during regeneration up to six-fold over normal at 335 weeks after nerve crush,9,25 one possibility may be that the amount of this proposed regulatory signal reaching the retina is related to the number of optic fibers in contact with the optic tectum. il~knowlrdgements--This work was supported by a postdoctoral fellowship from the Swedish Natural Science Research Council to C.H.. The Lars Hierta Foundation (C.H.), and a grant from the National Institutes of Health (EY 05212) to N.S.
REFERENCES 1. Benowitz
L. I., Shashoua V. E. and Yoon M. G. (1981) Specific changes in rapidly transported proteins during regeneration of the goldfish optic nerve. J. Neurosci. 1, 300-307. 2. Benowitz L., Yoon M. and Lewis E. (1983) Transported proteins in the regenerating optic nerve: regulation by interactions with the optic tectum. Science 222, 1855188. 3. Bignami A., Raju T. and Dahl D. (1982) Localization of vimentin, the non-specific intermediate filament protein in embrvonal glia and in early differentiating neurons. Deal Biol. 91, 286-295. Burmeister b. W. and Grafstein B. (198
