- Email: [email protected]
Expression of neurofilament proteins in granule cells of the cerebellum
Brain Research, 509 (1990) 47-54 Elsevier
47
BRES 15182
Expression of neurofilament proteins in granule cells of the cerebellum Maurizio VitadeUo 1 and Suzanne Denis-Donini ~ ¢C.N.R. lstituto di Fisiologia dei Centri Nervosi, Milan (Italy) and 2C.N.R. Center of Cytopharmacology and Department of Biology, Milan (Italy)
(Accepted 11 July 1989) Key words: Neurofilament; Cerebellum; Granule cell; Parallel fiber; Monoclonal antibody; Immunocytochemistry
We have used a panel of monoclonal antibodies directed against the low, middle and high molecular weight subunits of neurofilament triplet, to study their expression in mouse cerebellar granule cells. We demonstrate that in situ such cells only express the 2 lower molecular weight subunits either at various developmental stages or in the adult. The same results were obtained in vitro. This pattern of neurofilament protein expression in adult granule cells is therefore similar to that observed in developing neurons but differs from most neurons in the adult brain. The retention of such 'immature' pattern of neurofilament protein expression throughout adulthood could explain the lack of cytologically identifiable intermediate filaments in these neurons when examined with conventional electron microscopic techniques. It furthermore suggests that various neuronal populations might be characterized by the expression of specific subsets of neuronal intermediate filaments. INTRODUCTION M a m m a l i a n neurofilaments, the neuron-specific interm e d i a t e filaments, are c o m p o s e d of 3 subunits with a p p a r e n t m o l e c u l a r weights of about 68,000 (NF-L), 150,000 ( N F - M ) and 200,000 ( N F - H ) D a as m e a s u r e d by electrophoresis on p o l y a c r y l a m i d e gel in the presence of sodium dodecyi sulfate (SDS) 2°'24'35. The 3 subunits are the products of distinct genes 1~, which have b e e n cloned recently 23'27'31 and whose expression is d e v e l o p m e n t a l l y r e g u l a t e d Is. N F - L and N F - M m R N A s are present very early in the e m b r y o n i c brain, and their levels progressively increase during d e v e l o p m e n t . N F - H m R N A appears at late embryological stages and accumulates only in the p o s t n a t a l brain. A l l 3 polypeptides can be p h o s p h o r y l a t e d at multiple sites 17 and it has been shown that p h o s p h o r y l a t e d variants of N F - M and N F - H are p r e d o m i n a n t l y localized in axons while the non-phosp h o r y l a t e d forms are usually found in p e r i k a r y a and d e n d r i t e s 15'22'36. T h e r e is still d e b a t e w h e t h e r all neurons express n e u r o f i l a m e n t proteins. Some neuronal populations are not visualized with anti-neurofilament antibodies: they are generally small CNS neurons, with scant cytoplasm, whose processes are difficult to follow in situ 33"37. C e r e b e l l a r granule cells provide an a d e q u a t e e x p e r i m e n t a l m o d e l since n e u r o n a l p e r i k a r y a are localized in the internal granular layer and their axons extend into the m o l e c u l a r layer. M o r e o v e r , the timing of
d e v e l o p m e n t is well known and one can study these neurons before they leave the mitotic cycle when they are still located in the external germinal layer, during their migration across the m o l e c u l a r layer towards the internal granular layer and t h e r e f o r e during the p e r i o d when they extend axons and progressively form synapses with the dendrites of Purkinje cells 1'2. Previous reports showed that adult cerebeilar granule cells are not reactive with anti-NF antibodies 7'33, nor contain identifiable intermediate filaments when e x a m i n e d with conventional electron microscopic ( E M ) techniques 29. The lack of reactivity with a n t i - N F antibodies in granule cells and parallel fibers in vivo is in contrast with the expression of N F - L m R N A in the internal granule layer of the cerebellum, d e m o n s t r a t e d through in situ hybridization 25. We have therefore re-investigated the expression of neurofilament proteins both in adult mouse cerebellar granule cells and during d e v e l o p m e n t using a panel of m o n o c i o n a i antibodies directed against each of the 3 N F subunits. MATERIALS AND METHODS Generation and characterization of monoclonal antibodies (mAbs) We have produced 3 anti-NF mAbs IG2, 4E5 and 2F8 with the procedures described in previous reports3~'39 and their specificity following Western blotting is described below. Cytoskeletal extracts from rat and mouse brains were prepared according to Pruss et al. 3°. Electrophoretic separation of proteins on 7.5-15% gradient acrylamide SDS gels and electrophoretic transfer of polypeptides to nitrocellulose paper were performed as previously describeds~'39.
Correspondence: M. Vitadello, C.N.R. Istituto di Fisiologia dei Centri Nervosi, Via privata Mario Bianco 9. 20131 Milano, Italy.
0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
48 Nitrocellulose strips were incubated with appropriate dilutions of our monoclonal primary antibodies. Bound antibodies were revealed by incubation with peroxidase-conjugated rabbit anti-mouse IgG (dilution 1:800) (Dakopatts, Glostrup, Denmark) followed by incubation in the substrate solution (3,3"-diaminobenzidine in 100 mM Tris-imidazole, pH 7.6, 0.05% hydrogen peroxide). To test whether these mAbs were directed against phosphorylated epitopes, parallel strips of nitrocellulose paper were incubated either with 500 /~g/ml alkaline phosphatase (type I S; Sigma) in 0.1 M Tris-HCl, 0.01 M phenylmethylsulfonilfluoride pH 8.0, or in buffer without phosphatase for 2 h at 37 °C, before incubation with the primary antibodies. In addition to the 3 new mAbs described above, we also used mAb iC8 that we had generated and characterized previously and which is directed against a phosphate-independent epitope of rat NF-M and NF-H 38'39. MAb NR4, specific for a phosphateindependent epitope of NF-L 13"34, was purchased from Sigma (St. Louis, MO).
medium consisted of MEM and nutrient mixture FI0 (t:1) (Gibco. U.K.) enriched with 33 mM glucose, 2 mM glutamine and 3 mM sodium bicarbonate and complemented with 10% fetal calf serum. The cells were seeded on acid-cleaned glass coverslips coated with polyornithine 1.5 /~g/ml in Plastic Falcon dishes and grown in a humified 5% Co 2, 95% air atmosphere at 36 °C. After 3-7 clays in vitro (DIV), cells were fixed with 4% paraformaldehyde in 120 mM phosphate buffer pH 7.4, washed extensively and incubated with anti-NF antibodies diluted in PBS containing 0.03% Triton X-100 and 5% fetal calf serum. Bound antibodies were visualized using rhodamine-conjugated anti-mouse secondary antibodies, at the dilutions used for cryostat sections.
RESULTS
Characterization of new anti-NF antibodies T h e specificity o f m A b s
lrnrnunofluorescence staining Cerebella from 6-, 10-, 12-day-old and adult Swiss mice were dissected and immediately frozen, without prior fixation, in Freon cooled to -80 °C. Ten-micron-thick cryosections were mounted on slides and incubated with various dilutions of mAbs iC8 (1:250 to 1:1000), 1G2 (1:400 to 1:800), NR4 (1:40 to 1:100), 4E5 (undiluted supernatant) and 2F8 (1:400 to 1:1000) in phosphate-buffered saline (PBS). Controls were performed using mouse preimmune IgGs or hybridoma culture medium in the first step. Incubation was carried out in a humified chamber at room temperature for 1 h. After 30 min wash in PBS, the sections were incubated with TRITC-labeled anti-mouse IgG (Dakopatts) diluted 1:50 in PBS. Sections were washed in PBS, mounted in Elvanol (Garzanti Chimica, Italy) and examined with a Zeiss fluorescence photomicroscope III equipped with epi-illumination. Pictures were taken with Technical Pan 2415 film (Kodak, Rochester, NY).
1G2, 4E5 and 2F8, d e t e r -
m i n e d by i m m u n o b l o t t i n g t e c h n i q u e o n T r i t o n X-100 insoluble
cytoskeletai
proteins
brains, is s h o w n in Fig. 1. M A b
from
rat
and
mouse
1G2 d e c o r a t e s N F - M ,
while m A b s 4 E 5 and 2F8 stain N F - H , b o t h in rat and mouse
(Fig.
la).
T h e r e a c t i v i t y of the 3 m A b s
abolished after alkaline phosphatase cytoskeletal
proteins
(Fig.
lb),
was
t r e a t m e n t o f the
indicating
that
these
a n t i b o d i e s are d i r e c t e d against p h o s p h o r y l a t e d e p i t o p e s of
NF-M
and
NF-H.
MAb
iC8,
as
previously
d e m o n s t r a t e d 38, reacts with p h o s p h a t e - i n d e p e n d e n t epit o p e s of rat N F - H a n d N F - M , w h i l e it stains o n l y m o u s e N F - M (Fig. l a ) .
Cell cultures Primary cultures from 5- and 10-day-old mouse cerebella were prepared. The cells were dissociated enzymatically according to the procedure of Weber and Shachner 4°. They were subsequently grown under standard conditions as already described TM, The culture
Immunohistochemistry In situ. In adult m o u s e c e r e b e l l u m , 4 E 5 , d i r e c t e d against p h o s p h o r y l a t e d N F - H , labels a x o n s in the i n t e r n a l
la
Fig. 1. a: immunoblots on rat (r) and mouse (m) brain cytoskeletal extracts with mAbs iC8, 1G2, 4E5 and 2F8. The first two lanes show the Coomassie blue-stained gel: arrowheads from top to bottom indicate NF-H, NF-M and NF-L. 1C8 reacts with rat NF-H and NF-M, while it detects only mouse NF-M. IG2 stains NF-M, 4E5 and 2F8 stain NF-H in both species, b: pairs of immunoblots on rat cytoskeletal extract with mAbs 4E5, 1G2 and 2F8 on nitrocellulose strips treated with alkaline phosphatase (second lane) or with buffer without phosphatase (first lane) before incubation with the primary antibody. Phosphatase treatment abolishes staining with all 3 antibodies, indicating that they are directed against phosphorylated epitopes on NF-H (4E5 and 2F8) and NF-M (1G2).
49
granule layer as well as basket cell axons in the lower part of the molecular layer but does not stain granule cell perikarya or parallel fibers (Fig. 2A). An identical
•~ ~,
t~~
staining pattern was obtained with anti NF-H mAb 2F8 (result not shown). On the contrary, mAbs anti-NF-M (iC8 and 1G2) and anti-NF-L (NR4) decorate granule
.,~
Fig. 2. Indirect immunofluorescence staining of coronal cryosections of adult mouse cerebellum with anti-NF antibodies (A) 4E5 (NF-H), (B) iC8 (NF-M), (C) NR4 (NF-L) and (D) IG2(NF-M). Parallel fibers in the molecular layer are labeled by iC8 (B), NR4 (C) and IG2 (D) while they are unreactive with 4E5 (A). Basket fibers surrounding Purkinje cell perikarya and axons in the granule cell layer are labeled by all antibodies. MAb iC8 also decorates Purkinje cell perikarya and dendrites when used at higher concentration (inset). Bar = 30 pm.
Fig. 3. Indirect immunofhtorescence staining with mAb iC8 of coronal cryosections in the internal granular layer (G) and in the Purkinje cell layer (P) are decorated. pial surface. B: higher magnification at the level of the Purkinje cell layer: parallel Arrows indicate Purkinje cell bodies. Bars = 30 ym (A) and 21 pm (B).
cells. Fig. 2B-D shows coronal sections, incubated with each antibody, where parallel fibers in the molecular layer are stained and appear as thin streaks coursing perpendicular to descending basket cell axons. which are
of P6 mouse cerebellum. A: axons in the white matrcr. EGL, external germinal layer. Dotted line indicates the fibers are not clearly distinguishable from Purkinje cells.
more intensely labeled. Granule cell perikarya in the internal granular layer are difficult to identify on such sections and appear unstained. However. the resolution limits of light microscopy do not allow any definite
51
~
~
~
,
~
_
~
~"~
~
Fig. 4. Indirect immunofluorescence staining of coronal cryosections of P10 (A,B) and P12 (C,D) mouse cerebellum with mAbs iC8 (A,C) and 4E5 (B,D). IC8 decorates axons in the white matter, Purkinje cell perikarya and dendrites (arrows) and parallel fibers in the molecular layer (m). In contrast parallel fibers are not stained by 4E5, which only decorates axons in the white matter and in the internal granule layer (g). Bars = 20/~m (A,B) and 30/~m (C and D). conclusion on this issue. These 3 m A b s , in addition, d e c o r a t e axons in the white m a t t e r and in the internal granule layer. M A b iC8 decorates also Purkinje cell p e r i k a r y a and dendrites, when used at higher concentra-
tion (Fig. 2B, inset). N o such staining of Purkinje cell p e r i k a r y a was o b s e r v e d with 1G2 and NR4. A study of cryosections from cerebella at various stages of development, postnatal day 6 (P6), P10 and P12, indicates that
52 NF-L and NF-M are expressed very early in parallel fibers whereas NF-H is undetectable. Fig. 3 shows a cryosection from P6 cerebellum stained with iC8 (anti NF-M). The molecular layer is narrow and only a small number of parallel fibers can be observed in the proximity of Purkinje cell perikarya, which are intensely labeled. At P10 and P12 the molecular layer is much thicker and the number of parallel fibers decorated by iC8 has increased (Fig. 4). Parallel fibers were also stained with NR4 (anti NF-L) and IG2 (anti NF-M) (results not shown). In contrast, anti NF-H antibodies (4E5 and 2F8) do not stain any structure corresponding to parallel fibers at any stage that we have examined (Fig. 4 and result not shown). Only axons in the white and in the internal granule layer are decorated. No staining of mitotic and premigratory cells in the external germinal layer was noticed at these stages. In vitro. It has been shown that granule cells represent more than 80% of the neurons in cerebellar primary cultures from 5-10-day-old pups 16. In vitro, granule cell perikarya often appear clustered and are only weakly labeled by mAbs iC8, 1G2 and NR4 (Fig. 5). Neuronal processes are also faintly stained. In some regions, where neurites seem to form fascicles, a brighter staining can be observed. In contrast, our antibodies directed against phosphorylated NF-H (4E5 and 2F8) did not stain granule cells (result not shown). In granule cells from 10-day-old mouse cerebella and maintained in vitro for 7 days an identical staining pattern was observed with all our antibodies (results not shown). DISCUSSION
Fig. 5. Indirect immunofluorescence staining of cerebellar cultures from 5-day-old pups after 3 DIV with mAb 1G2 (A), iC8 (B) and NR4 (C). Granule cell bodies are often clustered and weakly labeled by iC8 and 1G2. Immunoreactivity is not uniformly distributed along cell processes and shows discontinuities (arrow in B). Bar = 17/zm.
Our findings indicate that during cerebellar development (P6-P12), granule cells and their axons, the parallel fibers display the pattern of NF immunoreactivity encountered in immature neurons: they are decorated only by mAbs directed against NF-L and NF~M and present no immunoreactivity for NF-H 1°'28'32. The same pattern is also observed in cerebellar primary cultures from P5-P10 pups where single cells and their processes can be readily analyzed. The very faint labeling observed in vitro indicates that very low amounts of NF proteins are present. This could explain the results of CambrayDeakin and Burgoyne 9 who reported the absence of NF-L and NF-M analyzed with antibodies that probably had a lower affinity for these antigens than ours. The lack of NF-H that we observed in vivo is in agreement with the report by Marc e t al. 26 who found no NF-H immunoreactivity in developing granule cells with two mAbs directed against phosphorylated and non-phosphorylated epitopes. Altogether these observations show that NF-H is not expressed during migration of cerebellar granule
53 cells from the external germinal layer to the internal granular layer. The transient expression of N F - H observed during granule cell development by CambrayDeakin and Burgoyne might result, as they suggested, from cross-reactivity of the m A b used with a phosphorylated epitope on protein(s) other than N F - H 3'42. The most striking observation is that the absence of N F - H expression encountered in early postnatal stages is maintained into adulthood: adult parallel fibers never display N F - H immunoreactivity while they strongly react with m A b s directed against NF-L and against two epitopes (phosphorylated or not) of NF-M. The presence of N F - L reactivity could already be suspected since it had been shown, through in situ hybridization, that m R N A coding for NF-L was present in the internal granular layer. The lack of immunoreactivity that we observed with our anti N F - H antibodies does not necessarily imply the absence of N F - H subunits in parallel fibers, since it could be present in other forms not recognized by our antibodies. However, the fact that other authors 5'21'26 could not detect any immunoreactivity in parallel fibers, using m A b s directed against phosphorylated and nonphosphorylated epitopes on NF-H, supports this view. Furthermore, in situ hybridization experiments with a N F - H c D N A probe have shown that N F - H m R N A is not synthesized in cerebellar folia ~2. Taken together, these observations indicate that adult cerebellar granule cells only express an incomplete set of NF proteins when compared to other neuronal populations. The lack of N F - H is likely to have important effects on the structure and function of neurofilaments since NF-H is a major
component of crosslinks among neurofilaments t9. Other indications about N F - H functions come from experiments on rabbit retinal ganglion cells where the appearance of NF-H during postnatal development is accompanied by an 8-fold reduction in the velocity of NF transport 41. The pattern of NF expression in adult cerebellar granule cells is quite similar to that reported for most developing neurons 1°'Ls'28'32. In this respect, it is interesting to note that immature optic nerve axons and immature Purkinje cells, which only display NF-L and NF-M immunoreactivity, do not show any morphologically identifiable intermediate filament 4'28. It was proposed that the in situ concentration of NF proteins is below the critical level to permit polymerization, or that subunits may reside in smaller protofilaments not visible at the electron microscope. It is conceivable that a similar organization of neurofilament protein subunits results in the absence of cytologically identifiable intermediate filaments in granule cells and parallel fibers, when examined with conventional E M techniques 29. In conclusion, cerebellar granule cells represent a clear example of mature neurons which express an incomplete set of NF protein subunits and could therefore represent a useful model to further investigate on the role of NFs in controlling neuronal shape and function and on the factor(s) that regulate the expression of the 3 subunits.
Acknowledgements. The authors wish to thank Mrs. L. Aquino and Mr. L. Chiumiento for their skilful technical assistance.
REFERENCES 1 Altman, J., Postnatal development of the cerebellar cortex in the rat. I. The external germinal layer and the transitional molecular layer, J. Comp. Neurol., 145 (1972) 353-398. 2 Altman, J., Postnatal development of the cerebellar cortex in the rat. II. Phases in the maturation of Purkinje cells and of the molecular layer, J. Cornp. Neurol., 145 (1972) 399-464. 3 Balkema, G.W. and Drager, U.C., Light-dependent antibody labelling of photoreceptors, Nature (Lond.), 316 (1985) 630633. 4 Bignami, A., Grossi, M. and Dahl, D., Transient expression of neurofilament protein without filament formation in Purkinje cell development. Immunohistological and electron microscopic study of chicken cerebellum, Int. J. Dev. Neurosci., 3 (1985) 365-377. 5 Bignami, A. and Clark, K., Non phosphorylated and phosphorylated neurofilaments in hypothyroid rat cerebellum, Brain Research, 409 (1987) 143-145. 6 Bignami, A. and Dahl, D., Axonal maturation in development. II. Immunofluorescence study of rat spinal cord and cerebellum with axon-specific neurofilament antibodies, Int. J. Dev. Neurosci., 5 (1987) 29-37. 7 Burgoyne, R.D. and Cambray-Deakin, M.A., The cellular neurobiology of neuronal development: the cerebellar granule cells, Brain Res. Rev., 13 (1987) 77-101. 8 Cambray-Deakin, M.A. and Burgoyne, R.D., Transient expres-
9
10
11 12
13
14
15
sion of neurofilament-like (RT97) immunoreactivity in cerebellar granule cells, Dev. Brain Res., 28 (1986) 282-286. Cambray-Deakin, M.A. and Burgoyne, R.D., Sequential appearance of cytoskeletal components during the early stages of neurite outgrowth from cerebellar granule cells in vitro, Dev. Brain Res., 37 (1987) 197-207. Carden, M.J., Trojanowski, J.O., Schlaepfer, W.W. and Lee, V.M.-Y., Two-stage expression of neurofilament polypeptides during rat neurogenesis with early establishment of adult phosphorylation patterns, J. Neurosci., 7 (1987) 3489-3504. Czosnek, H., Soifer, D. and Wisniewski, H.M., Studies on the biosynthesis of neurofilament proteins, J. Cell. Biol., 85 (1980) 726-734. Dautigny, A., Pham-Dihn, D., Roussel, C,, Felix, J.M., Nussbaum, J.L. and Joll6s, P., The large neurofilament subunit (NF-H) of the rat: cDNA cloning and in situ detection, Biochem. Biophys. Res. Commun., 154 (1988) 1099-1106. Debus, E., Weber, K. and Osborn, M., Monoclonal antibodies specific for glial fibrillary acidic (GFA) protein and for each of the neurofilament triplet polypeptides, Differentiation, 25 (1983) 193-203. Denis-Donini, S., Glowinski, J. and Prochiantz, A., Specific influence of striated target neurons on the in vitro outgrowth of mesencephalic dopaminergic neurites: a morphological quantitative study, J. Neurosci., 3 (1983) 2292-2299. Foster, G.A., Dahl, D. and Lee, V.M.-Y., Temporal and topographic relationship between the phosphorylated and non-
54 phosphorylated epitopes of the 200 kDa neurofilament protein during development in vitro, J. Neurosci, 7 (1987) 2651-2663. 16 Gallo, V., Ciotti, M.T., Coletti, A., Aloisi, E and Levi, G., Selective release of glutamate from cerebellar granule cells differentiating in culture, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 7919-7923. 17 Julien, J.-P. and Mushynski, N.E, Multiple phosphorylation sites in mammalian neurofilament polypeptides, J. Biol. Chem., 257 (1982) 10467-10470. 18 Julien, J.P., Meyer, D., Flavell, D., Hurst, J. and Grosveld, E, Cloning and developmental expression of the murine neurofilament gene family, Mol. Brain Res., 1 (1986) 243-250. 19 Hirokawa, N., Glicksman, M.A. and Willard, M.B., Organization of mammalian neurofilament polypeptides within the neuronal cytoskeleton, J. Cell. Biol., 98 (1984) 1523-1536. 20 Hoffman, P.N. and Lasek, R.J., The slow component of axonal transport. Identification of major structural polypeptides of the axon and their generality among mammalian neurons, J. Ce//. Biol., 66 (1976) 351-366. 21 Langley, O.K., Sternberger, N.H. and Sternberger, L.A., Expression of neurofilament proteins by Purkinje cells: ultrastructural immunolocalization with monoclonal antibodies, Brain Research, 457 (1988) 12-20. 22 Lee. V.M.-Y., Garden, M.J., Schlaepfer, W.W. and Trojanowski, J.Q., Monoclonal antibodies distinguish several differentially phosphorylated states of the two largest rat neurofilament subunits (NF-H and NF-M) and demonstrates their existence in the normal nervous system of adult rats, J. Neurosci., 7 (1987) 3474-3488. 23 Lewis, S.A. and Cowan, N.J., Genetics, evolution and expression of the 68,000 molecular neurofilament protein: isolation of a cDNA probe, J. Cell. Biol., 100 (1985) 843-850. 24 Liem, R.H.K., Yen, S.-H., Solomon, G.E. and Shelanski, M.L., Intermediate filaments in nervous tissues, J. Cell. Biol., 79 (1978) 637-645. 25 Liesi, P., Julien, J.-P., Vilja, P., Grosveld, F. and Rechardt, L., Specific detection of neuronal cell bodies: in situ hybridization with a biotin-labelled neurofilament cDNA probe, J. Histochem. Cytochem., 34 (1986) 923-926. 26 Marc, C., Clavel, M.-C. and Rabi6, A., Non-phosphorylated and phosphorylated neurofilaments in the cerebellum of the rat: an immunohistochemical study using monoclonal antibodies. Development in normal and thyroid-deficient animals, Dev, Brain Res., 26 (1986) 249-260. 27 Napolitano, E.W., Chiu, S.S.M., Colman, D.R. and Liem, R.H.K., Complete amino acid sequence and in vitro expression of rat NF-M, the middle molecular weight neurofilament protein, J. Neurosci., 7 (1987) 2590-2598. 28 Pachter, J.S. and Liem, R.K.H., The differential appearance of neurofilament triplet polypeptides in the developing rat optic
nerve, Dev. Biol., 103 (1984) 200-21U. 29 Palay, S.L. and Chan-Palay, V., Cerebellar cortex. Cytology and organization, Springer, Heidelberg, 1974~ 30 Pruss, R.M., Mirsky, R., Raft, M.C., Thorpe, R., Dowding, A.J. and Anderton, B.H., All classes of intermediate filaments show a common antigenic determinant defined by a monoclonal antibody, Cell, 27 (1981) 419-428. 31 Schneidman, P.S., Carden, M.J., Lees, J.F. and Lazzarini, R.A., The structure of the largest murine neurofilament protein (NF-H) as revealed by cDNA and genomic sequences, Mol. Brain Res., 4 (1988) 217-231. 32 Shaw, G. and Weber, K., Differential expression of neurofilament triplet proteins in brain development, Nature (Lond.), 298 (1982) 277-279. 33 Shaw, G., Osborn, M. and Weber, K., An immunofluorescence microscopical study of the neurofilament triplet proteins, vimentin and glial fibrillary acidic protein within rat brain, Eur. J. Cell Biol., 26 (1981) 68-82. 34 Shaw, G., Osborn, M. and Weber, K., Reactivity of a panel of neurofilament antibodies on phosphorylated and dephosphorylated neurofilaments, Eur. J. Cell Biol., 41 (1986) I-9. 35 Schlaepfer, W.W. and Freeman, L.A., Neurofilament proteins of rat peripheral nerve and spinal cord, J. Cell Biol., 78 (1978) 653-662. 36 Sternberger, L.A. and Sternberger, N.H., Monoclonal antibodies distinguish phosphorylated and nonphosphorylated forms of neurofilaments in situ, Proc. Natl. Acad. Sci. U.S.A., 80 (1983) 6126-6130. 37 Trojanowski, J.Q., Walkenstein, N. and Lee, V.M.-Y., Expression of neurofilament subunits in neurons of the central and peripheral nervous system: an immunohistochemical study with monoclonal antibodies, J. Neurosci., 6 (1986) 650-660. 38 Vitadello, M., Triban, C., Fabris, M., Gorio, A. and Schiaffino, S., Heterogeneity of rat neurofilament polypeptides revealed by a monoclonal antibody, J. Neurochem., 46 (1986) 665-670. 39 VitadeUo, M., Triban, C., Fabris, M., Donh, M., Gorio, A. and Schiaffino, S., A developmentally regulated isoform of 150,000 molecular weight neurofilament protein specifically expressed in autonomic and small sensory neurons, Neuroscience, 23 (1987) 931-941. 40 Weber, A. and Schachner, M., Maintenance of immunocytologically identified Purkinje cells from mouse cerebellum in monotayer culture, Brain Research, 311 (1984) 119-130. 41 Willard, M. and Simon, C., Modulations of neurofilament axonal transport during development of rabbit retinal ganglion cells, Cell, 35 (1983) 551-559. 42 Wood, J.N., Lathangue, N.B., McLachlan, D.R., Smith, B.J., Anderton, B.H. and Dowding, A.J., Chromatin proteins share antigenic determinants with neurofilaments, J. Neurochem., 44 (1985) 149-154.
47
BRES 15182
Expression of neurofilament proteins in granule cells of the cerebellum Maurizio VitadeUo 1 and Suzanne Denis-Donini ~ ¢C.N.R. lstituto di Fisiologia dei Centri Nervosi, Milan (Italy) and 2C.N.R. Center of Cytopharmacology and Department of Biology, Milan (Italy)
(Accepted 11 July 1989) Key words: Neurofilament; Cerebellum; Granule cell; Parallel fiber; Monoclonal antibody; Immunocytochemistry
We have used a panel of monoclonal antibodies directed against the low, middle and high molecular weight subunits of neurofilament triplet, to study their expression in mouse cerebellar granule cells. We demonstrate that in situ such cells only express the 2 lower molecular weight subunits either at various developmental stages or in the adult. The same results were obtained in vitro. This pattern of neurofilament protein expression in adult granule cells is therefore similar to that observed in developing neurons but differs from most neurons in the adult brain. The retention of such 'immature' pattern of neurofilament protein expression throughout adulthood could explain the lack of cytologically identifiable intermediate filaments in these neurons when examined with conventional electron microscopic techniques. It furthermore suggests that various neuronal populations might be characterized by the expression of specific subsets of neuronal intermediate filaments. INTRODUCTION M a m m a l i a n neurofilaments, the neuron-specific interm e d i a t e filaments, are c o m p o s e d of 3 subunits with a p p a r e n t m o l e c u l a r weights of about 68,000 (NF-L), 150,000 ( N F - M ) and 200,000 ( N F - H ) D a as m e a s u r e d by electrophoresis on p o l y a c r y l a m i d e gel in the presence of sodium dodecyi sulfate (SDS) 2°'24'35. The 3 subunits are the products of distinct genes 1~, which have b e e n cloned recently 23'27'31 and whose expression is d e v e l o p m e n t a l l y r e g u l a t e d Is. N F - L and N F - M m R N A s are present very early in the e m b r y o n i c brain, and their levels progressively increase during d e v e l o p m e n t . N F - H m R N A appears at late embryological stages and accumulates only in the p o s t n a t a l brain. A l l 3 polypeptides can be p h o s p h o r y l a t e d at multiple sites 17 and it has been shown that p h o s p h o r y l a t e d variants of N F - M and N F - H are p r e d o m i n a n t l y localized in axons while the non-phosp h o r y l a t e d forms are usually found in p e r i k a r y a and d e n d r i t e s 15'22'36. T h e r e is still d e b a t e w h e t h e r all neurons express n e u r o f i l a m e n t proteins. Some neuronal populations are not visualized with anti-neurofilament antibodies: they are generally small CNS neurons, with scant cytoplasm, whose processes are difficult to follow in situ 33"37. C e r e b e l l a r granule cells provide an a d e q u a t e e x p e r i m e n t a l m o d e l since n e u r o n a l p e r i k a r y a are localized in the internal granular layer and their axons extend into the m o l e c u l a r layer. M o r e o v e r , the timing of
d e v e l o p m e n t is well known and one can study these neurons before they leave the mitotic cycle when they are still located in the external germinal layer, during their migration across the m o l e c u l a r layer towards the internal granular layer and t h e r e f o r e during the p e r i o d when they extend axons and progressively form synapses with the dendrites of Purkinje cells 1'2. Previous reports showed that adult cerebeilar granule cells are not reactive with anti-NF antibodies 7'33, nor contain identifiable intermediate filaments when e x a m i n e d with conventional electron microscopic ( E M ) techniques 29. The lack of reactivity with a n t i - N F antibodies in granule cells and parallel fibers in vivo is in contrast with the expression of N F - L m R N A in the internal granule layer of the cerebellum, d e m o n s t r a t e d through in situ hybridization 25. We have therefore re-investigated the expression of neurofilament proteins both in adult mouse cerebellar granule cells and during d e v e l o p m e n t using a panel of m o n o c i o n a i antibodies directed against each of the 3 N F subunits. MATERIALS AND METHODS Generation and characterization of monoclonal antibodies (mAbs) We have produced 3 anti-NF mAbs IG2, 4E5 and 2F8 with the procedures described in previous reports3~'39 and their specificity following Western blotting is described below. Cytoskeletal extracts from rat and mouse brains were prepared according to Pruss et al. 3°. Electrophoretic separation of proteins on 7.5-15% gradient acrylamide SDS gels and electrophoretic transfer of polypeptides to nitrocellulose paper were performed as previously describeds~'39.
Correspondence: M. Vitadello, C.N.R. Istituto di Fisiologia dei Centri Nervosi, Via privata Mario Bianco 9. 20131 Milano, Italy.
0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
48 Nitrocellulose strips were incubated with appropriate dilutions of our monoclonal primary antibodies. Bound antibodies were revealed by incubation with peroxidase-conjugated rabbit anti-mouse IgG (dilution 1:800) (Dakopatts, Glostrup, Denmark) followed by incubation in the substrate solution (3,3"-diaminobenzidine in 100 mM Tris-imidazole, pH 7.6, 0.05% hydrogen peroxide). To test whether these mAbs were directed against phosphorylated epitopes, parallel strips of nitrocellulose paper were incubated either with 500 /~g/ml alkaline phosphatase (type I S; Sigma) in 0.1 M Tris-HCl, 0.01 M phenylmethylsulfonilfluoride pH 8.0, or in buffer without phosphatase for 2 h at 37 °C, before incubation with the primary antibodies. In addition to the 3 new mAbs described above, we also used mAb iC8 that we had generated and characterized previously and which is directed against a phosphate-independent epitope of rat NF-M and NF-H 38'39. MAb NR4, specific for a phosphateindependent epitope of NF-L 13"34, was purchased from Sigma (St. Louis, MO).
medium consisted of MEM and nutrient mixture FI0 (t:1) (Gibco. U.K.) enriched with 33 mM glucose, 2 mM glutamine and 3 mM sodium bicarbonate and complemented with 10% fetal calf serum. The cells were seeded on acid-cleaned glass coverslips coated with polyornithine 1.5 /~g/ml in Plastic Falcon dishes and grown in a humified 5% Co 2, 95% air atmosphere at 36 °C. After 3-7 clays in vitro (DIV), cells were fixed with 4% paraformaldehyde in 120 mM phosphate buffer pH 7.4, washed extensively and incubated with anti-NF antibodies diluted in PBS containing 0.03% Triton X-100 and 5% fetal calf serum. Bound antibodies were visualized using rhodamine-conjugated anti-mouse secondary antibodies, at the dilutions used for cryostat sections.
RESULTS
Characterization of new anti-NF antibodies T h e specificity o f m A b s
lrnrnunofluorescence staining Cerebella from 6-, 10-, 12-day-old and adult Swiss mice were dissected and immediately frozen, without prior fixation, in Freon cooled to -80 °C. Ten-micron-thick cryosections were mounted on slides and incubated with various dilutions of mAbs iC8 (1:250 to 1:1000), 1G2 (1:400 to 1:800), NR4 (1:40 to 1:100), 4E5 (undiluted supernatant) and 2F8 (1:400 to 1:1000) in phosphate-buffered saline (PBS). Controls were performed using mouse preimmune IgGs or hybridoma culture medium in the first step. Incubation was carried out in a humified chamber at room temperature for 1 h. After 30 min wash in PBS, the sections were incubated with TRITC-labeled anti-mouse IgG (Dakopatts) diluted 1:50 in PBS. Sections were washed in PBS, mounted in Elvanol (Garzanti Chimica, Italy) and examined with a Zeiss fluorescence photomicroscope III equipped with epi-illumination. Pictures were taken with Technical Pan 2415 film (Kodak, Rochester, NY).
1G2, 4E5 and 2F8, d e t e r -
m i n e d by i m m u n o b l o t t i n g t e c h n i q u e o n T r i t o n X-100 insoluble
cytoskeletai
proteins
brains, is s h o w n in Fig. 1. M A b
from
rat
and
mouse
1G2 d e c o r a t e s N F - M ,
while m A b s 4 E 5 and 2F8 stain N F - H , b o t h in rat and mouse
(Fig.
la).
T h e r e a c t i v i t y of the 3 m A b s
abolished after alkaline phosphatase cytoskeletal
proteins
(Fig.
lb),
was
t r e a t m e n t o f the
indicating
that
these
a n t i b o d i e s are d i r e c t e d against p h o s p h o r y l a t e d e p i t o p e s of
NF-M
and
NF-H.
MAb
iC8,
as
previously
d e m o n s t r a t e d 38, reacts with p h o s p h a t e - i n d e p e n d e n t epit o p e s of rat N F - H a n d N F - M , w h i l e it stains o n l y m o u s e N F - M (Fig. l a ) .
Cell cultures Primary cultures from 5- and 10-day-old mouse cerebella were prepared. The cells were dissociated enzymatically according to the procedure of Weber and Shachner 4°. They were subsequently grown under standard conditions as already described TM, The culture
Immunohistochemistry In situ. In adult m o u s e c e r e b e l l u m , 4 E 5 , d i r e c t e d against p h o s p h o r y l a t e d N F - H , labels a x o n s in the i n t e r n a l
la
Fig. 1. a: immunoblots on rat (r) and mouse (m) brain cytoskeletal extracts with mAbs iC8, 1G2, 4E5 and 2F8. The first two lanes show the Coomassie blue-stained gel: arrowheads from top to bottom indicate NF-H, NF-M and NF-L. 1C8 reacts with rat NF-H and NF-M, while it detects only mouse NF-M. IG2 stains NF-M, 4E5 and 2F8 stain NF-H in both species, b: pairs of immunoblots on rat cytoskeletal extract with mAbs 4E5, 1G2 and 2F8 on nitrocellulose strips treated with alkaline phosphatase (second lane) or with buffer without phosphatase (first lane) before incubation with the primary antibody. Phosphatase treatment abolishes staining with all 3 antibodies, indicating that they are directed against phosphorylated epitopes on NF-H (4E5 and 2F8) and NF-M (1G2).
49
granule layer as well as basket cell axons in the lower part of the molecular layer but does not stain granule cell perikarya or parallel fibers (Fig. 2A). An identical
•~ ~,
t~~
staining pattern was obtained with anti NF-H mAb 2F8 (result not shown). On the contrary, mAbs anti-NF-M (iC8 and 1G2) and anti-NF-L (NR4) decorate granule
.,~
Fig. 2. Indirect immunofluorescence staining of coronal cryosections of adult mouse cerebellum with anti-NF antibodies (A) 4E5 (NF-H), (B) iC8 (NF-M), (C) NR4 (NF-L) and (D) IG2(NF-M). Parallel fibers in the molecular layer are labeled by iC8 (B), NR4 (C) and IG2 (D) while they are unreactive with 4E5 (A). Basket fibers surrounding Purkinje cell perikarya and axons in the granule cell layer are labeled by all antibodies. MAb iC8 also decorates Purkinje cell perikarya and dendrites when used at higher concentration (inset). Bar = 30 pm.
Fig. 3. Indirect immunofhtorescence staining with mAb iC8 of coronal cryosections in the internal granular layer (G) and in the Purkinje cell layer (P) are decorated. pial surface. B: higher magnification at the level of the Purkinje cell layer: parallel Arrows indicate Purkinje cell bodies. Bars = 30 ym (A) and 21 pm (B).
cells. Fig. 2B-D shows coronal sections, incubated with each antibody, where parallel fibers in the molecular layer are stained and appear as thin streaks coursing perpendicular to descending basket cell axons. which are
of P6 mouse cerebellum. A: axons in the white matrcr. EGL, external germinal layer. Dotted line indicates the fibers are not clearly distinguishable from Purkinje cells.
more intensely labeled. Granule cell perikarya in the internal granular layer are difficult to identify on such sections and appear unstained. However. the resolution limits of light microscopy do not allow any definite
51
~
~
~
,
~
_
~
~"~
~
Fig. 4. Indirect immunofluorescence staining of coronal cryosections of P10 (A,B) and P12 (C,D) mouse cerebellum with mAbs iC8 (A,C) and 4E5 (B,D). IC8 decorates axons in the white matter, Purkinje cell perikarya and dendrites (arrows) and parallel fibers in the molecular layer (m). In contrast parallel fibers are not stained by 4E5, which only decorates axons in the white matter and in the internal granule layer (g). Bars = 20/~m (A,B) and 30/~m (C and D). conclusion on this issue. These 3 m A b s , in addition, d e c o r a t e axons in the white m a t t e r and in the internal granule layer. M A b iC8 decorates also Purkinje cell p e r i k a r y a and dendrites, when used at higher concentra-
tion (Fig. 2B, inset). N o such staining of Purkinje cell p e r i k a r y a was o b s e r v e d with 1G2 and NR4. A study of cryosections from cerebella at various stages of development, postnatal day 6 (P6), P10 and P12, indicates that
52 NF-L and NF-M are expressed very early in parallel fibers whereas NF-H is undetectable. Fig. 3 shows a cryosection from P6 cerebellum stained with iC8 (anti NF-M). The molecular layer is narrow and only a small number of parallel fibers can be observed in the proximity of Purkinje cell perikarya, which are intensely labeled. At P10 and P12 the molecular layer is much thicker and the number of parallel fibers decorated by iC8 has increased (Fig. 4). Parallel fibers were also stained with NR4 (anti NF-L) and IG2 (anti NF-M) (results not shown). In contrast, anti NF-H antibodies (4E5 and 2F8) do not stain any structure corresponding to parallel fibers at any stage that we have examined (Fig. 4 and result not shown). Only axons in the white and in the internal granule layer are decorated. No staining of mitotic and premigratory cells in the external germinal layer was noticed at these stages. In vitro. It has been shown that granule cells represent more than 80% of the neurons in cerebellar primary cultures from 5-10-day-old pups 16. In vitro, granule cell perikarya often appear clustered and are only weakly labeled by mAbs iC8, 1G2 and NR4 (Fig. 5). Neuronal processes are also faintly stained. In some regions, where neurites seem to form fascicles, a brighter staining can be observed. In contrast, our antibodies directed against phosphorylated NF-H (4E5 and 2F8) did not stain granule cells (result not shown). In granule cells from 10-day-old mouse cerebella and maintained in vitro for 7 days an identical staining pattern was observed with all our antibodies (results not shown). DISCUSSION
Fig. 5. Indirect immunofluorescence staining of cerebellar cultures from 5-day-old pups after 3 DIV with mAb 1G2 (A), iC8 (B) and NR4 (C). Granule cell bodies are often clustered and weakly labeled by iC8 and 1G2. Immunoreactivity is not uniformly distributed along cell processes and shows discontinuities (arrow in B). Bar = 17/zm.
Our findings indicate that during cerebellar development (P6-P12), granule cells and their axons, the parallel fibers display the pattern of NF immunoreactivity encountered in immature neurons: they are decorated only by mAbs directed against NF-L and NF~M and present no immunoreactivity for NF-H 1°'28'32. The same pattern is also observed in cerebellar primary cultures from P5-P10 pups where single cells and their processes can be readily analyzed. The very faint labeling observed in vitro indicates that very low amounts of NF proteins are present. This could explain the results of CambrayDeakin and Burgoyne 9 who reported the absence of NF-L and NF-M analyzed with antibodies that probably had a lower affinity for these antigens than ours. The lack of NF-H that we observed in vivo is in agreement with the report by Marc e t al. 26 who found no NF-H immunoreactivity in developing granule cells with two mAbs directed against phosphorylated and non-phosphorylated epitopes. Altogether these observations show that NF-H is not expressed during migration of cerebellar granule
53 cells from the external germinal layer to the internal granular layer. The transient expression of N F - H observed during granule cell development by CambrayDeakin and Burgoyne might result, as they suggested, from cross-reactivity of the m A b used with a phosphorylated epitope on protein(s) other than N F - H 3'42. The most striking observation is that the absence of N F - H expression encountered in early postnatal stages is maintained into adulthood: adult parallel fibers never display N F - H immunoreactivity while they strongly react with m A b s directed against NF-L and against two epitopes (phosphorylated or not) of NF-M. The presence of N F - L reactivity could already be suspected since it had been shown, through in situ hybridization, that m R N A coding for NF-L was present in the internal granular layer. The lack of immunoreactivity that we observed with our anti N F - H antibodies does not necessarily imply the absence of N F - H subunits in parallel fibers, since it could be present in other forms not recognized by our antibodies. However, the fact that other authors 5'21'26 could not detect any immunoreactivity in parallel fibers, using m A b s directed against phosphorylated and nonphosphorylated epitopes on NF-H, supports this view. Furthermore, in situ hybridization experiments with a N F - H c D N A probe have shown that N F - H m R N A is not synthesized in cerebellar folia ~2. Taken together, these observations indicate that adult cerebellar granule cells only express an incomplete set of NF proteins when compared to other neuronal populations. The lack of N F - H is likely to have important effects on the structure and function of neurofilaments since NF-H is a major
component of crosslinks among neurofilaments t9. Other indications about N F - H functions come from experiments on rabbit retinal ganglion cells where the appearance of NF-H during postnatal development is accompanied by an 8-fold reduction in the velocity of NF transport 41. The pattern of NF expression in adult cerebellar granule cells is quite similar to that reported for most developing neurons 1°'Ls'28'32. In this respect, it is interesting to note that immature optic nerve axons and immature Purkinje cells, which only display NF-L and NF-M immunoreactivity, do not show any morphologically identifiable intermediate filament 4'28. It was proposed that the in situ concentration of NF proteins is below the critical level to permit polymerization, or that subunits may reside in smaller protofilaments not visible at the electron microscope. It is conceivable that a similar organization of neurofilament protein subunits results in the absence of cytologically identifiable intermediate filaments in granule cells and parallel fibers, when examined with conventional E M techniques 29. In conclusion, cerebellar granule cells represent a clear example of mature neurons which express an incomplete set of NF protein subunits and could therefore represent a useful model to further investigate on the role of NFs in controlling neuronal shape and function and on the factor(s) that regulate the expression of the 3 subunits.
Acknowledgements. The authors wish to thank Mrs. L. Aquino and Mr. L. Chiumiento for their skilful technical assistance.
REFERENCES 1 Altman, J., Postnatal development of the cerebellar cortex in the rat. I. The external germinal layer and the transitional molecular layer, J. Comp. Neurol., 145 (1972) 353-398. 2 Altman, J., Postnatal development of the cerebellar cortex in the rat. II. Phases in the maturation of Purkinje cells and of the molecular layer, J. Cornp. Neurol., 145 (1972) 399-464. 3 Balkema, G.W. and Drager, U.C., Light-dependent antibody labelling of photoreceptors, Nature (Lond.), 316 (1985) 630633. 4 Bignami, A., Grossi, M. and Dahl, D., Transient expression of neurofilament protein without filament formation in Purkinje cell development. Immunohistological and electron microscopic study of chicken cerebellum, Int. J. Dev. Neurosci., 3 (1985) 365-377. 5 Bignami, A. and Clark, K., Non phosphorylated and phosphorylated neurofilaments in hypothyroid rat cerebellum, Brain Research, 409 (1987) 143-145. 6 Bignami, A. and Dahl, D., Axonal maturation in development. II. Immunofluorescence study of rat spinal cord and cerebellum with axon-specific neurofilament antibodies, Int. J. Dev. Neurosci., 5 (1987) 29-37. 7 Burgoyne, R.D. and Cambray-Deakin, M.A., The cellular neurobiology of neuronal development: the cerebellar granule cells, Brain Res. Rev., 13 (1987) 77-101. 8 Cambray-Deakin, M.A. and Burgoyne, R.D., Transient expres-
9
10
11 12
13
14
15
sion of neurofilament-like (RT97) immunoreactivity in cerebellar granule cells, Dev. Brain Res., 28 (1986) 282-286. Cambray-Deakin, M.A. and Burgoyne, R.D., Sequential appearance of cytoskeletal components during the early stages of neurite outgrowth from cerebellar granule cells in vitro, Dev. Brain Res., 37 (1987) 197-207. Carden, M.J., Trojanowski, J.O., Schlaepfer, W.W. and Lee, V.M.-Y., Two-stage expression of neurofilament polypeptides during rat neurogenesis with early establishment of adult phosphorylation patterns, J. Neurosci., 7 (1987) 3489-3504. Czosnek, H., Soifer, D. and Wisniewski, H.M., Studies on the biosynthesis of neurofilament proteins, J. Cell. Biol., 85 (1980) 726-734. Dautigny, A., Pham-Dihn, D., Roussel, C,, Felix, J.M., Nussbaum, J.L. and Joll6s, P., The large neurofilament subunit (NF-H) of the rat: cDNA cloning and in situ detection, Biochem. Biophys. Res. Commun., 154 (1988) 1099-1106. Debus, E., Weber, K. and Osborn, M., Monoclonal antibodies specific for glial fibrillary acidic (GFA) protein and for each of the neurofilament triplet polypeptides, Differentiation, 25 (1983) 193-203. Denis-Donini, S., Glowinski, J. and Prochiantz, A., Specific influence of striated target neurons on the in vitro outgrowth of mesencephalic dopaminergic neurites: a morphological quantitative study, J. Neurosci., 3 (1983) 2292-2299. Foster, G.A., Dahl, D. and Lee, V.M.-Y., Temporal and topographic relationship between the phosphorylated and non-
54 phosphorylated epitopes of the 200 kDa neurofilament protein during development in vitro, J. Neurosci, 7 (1987) 2651-2663. 16 Gallo, V., Ciotti, M.T., Coletti, A., Aloisi, E and Levi, G., Selective release of glutamate from cerebellar granule cells differentiating in culture, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 7919-7923. 17 Julien, J.-P. and Mushynski, N.E, Multiple phosphorylation sites in mammalian neurofilament polypeptides, J. Biol. Chem., 257 (1982) 10467-10470. 18 Julien, J.P., Meyer, D., Flavell, D., Hurst, J. and Grosveld, E, Cloning and developmental expression of the murine neurofilament gene family, Mol. Brain Res., 1 (1986) 243-250. 19 Hirokawa, N., Glicksman, M.A. and Willard, M.B., Organization of mammalian neurofilament polypeptides within the neuronal cytoskeleton, J. Cell. Biol., 98 (1984) 1523-1536. 20 Hoffman, P.N. and Lasek, R.J., The slow component of axonal transport. Identification of major structural polypeptides of the axon and their generality among mammalian neurons, J. Ce//. Biol., 66 (1976) 351-366. 21 Langley, O.K., Sternberger, N.H. and Sternberger, L.A., Expression of neurofilament proteins by Purkinje cells: ultrastructural immunolocalization with monoclonal antibodies, Brain Research, 457 (1988) 12-20. 22 Lee. V.M.-Y., Garden, M.J., Schlaepfer, W.W. and Trojanowski, J.Q., Monoclonal antibodies distinguish several differentially phosphorylated states of the two largest rat neurofilament subunits (NF-H and NF-M) and demonstrates their existence in the normal nervous system of adult rats, J. Neurosci., 7 (1987) 3474-3488. 23 Lewis, S.A. and Cowan, N.J., Genetics, evolution and expression of the 68,000 molecular neurofilament protein: isolation of a cDNA probe, J. Cell. Biol., 100 (1985) 843-850. 24 Liem, R.H.K., Yen, S.-H., Solomon, G.E. and Shelanski, M.L., Intermediate filaments in nervous tissues, J. Cell. Biol., 79 (1978) 637-645. 25 Liesi, P., Julien, J.-P., Vilja, P., Grosveld, F. and Rechardt, L., Specific detection of neuronal cell bodies: in situ hybridization with a biotin-labelled neurofilament cDNA probe, J. Histochem. Cytochem., 34 (1986) 923-926. 26 Marc, C., Clavel, M.-C. and Rabi6, A., Non-phosphorylated and phosphorylated neurofilaments in the cerebellum of the rat: an immunohistochemical study using monoclonal antibodies. Development in normal and thyroid-deficient animals, Dev, Brain Res., 26 (1986) 249-260. 27 Napolitano, E.W., Chiu, S.S.M., Colman, D.R. and Liem, R.H.K., Complete amino acid sequence and in vitro expression of rat NF-M, the middle molecular weight neurofilament protein, J. Neurosci., 7 (1987) 2590-2598. 28 Pachter, J.S. and Liem, R.K.H., The differential appearance of neurofilament triplet polypeptides in the developing rat optic
nerve, Dev. Biol., 103 (1984) 200-21U. 29 Palay, S.L. and Chan-Palay, V., Cerebellar cortex. Cytology and organization, Springer, Heidelberg, 1974~ 30 Pruss, R.M., Mirsky, R., Raft, M.C., Thorpe, R., Dowding, A.J. and Anderton, B.H., All classes of intermediate filaments show a common antigenic determinant defined by a monoclonal antibody, Cell, 27 (1981) 419-428. 31 Schneidman, P.S., Carden, M.J., Lees, J.F. and Lazzarini, R.A., The structure of the largest murine neurofilament protein (NF-H) as revealed by cDNA and genomic sequences, Mol. Brain Res., 4 (1988) 217-231. 32 Shaw, G. and Weber, K., Differential expression of neurofilament triplet proteins in brain development, Nature (Lond.), 298 (1982) 277-279. 33 Shaw, G., Osborn, M. and Weber, K., An immunofluorescence microscopical study of the neurofilament triplet proteins, vimentin and glial fibrillary acidic protein within rat brain, Eur. J. Cell Biol., 26 (1981) 68-82. 34 Shaw, G., Osborn, M. and Weber, K., Reactivity of a panel of neurofilament antibodies on phosphorylated and dephosphorylated neurofilaments, Eur. J. Cell Biol., 41 (1986) I-9. 35 Schlaepfer, W.W. and Freeman, L.A., Neurofilament proteins of rat peripheral nerve and spinal cord, J. Cell Biol., 78 (1978) 653-662. 36 Sternberger, L.A. and Sternberger, N.H., Monoclonal antibodies distinguish phosphorylated and nonphosphorylated forms of neurofilaments in situ, Proc. Natl. Acad. Sci. U.S.A., 80 (1983) 6126-6130. 37 Trojanowski, J.Q., Walkenstein, N. and Lee, V.M.-Y., Expression of neurofilament subunits in neurons of the central and peripheral nervous system: an immunohistochemical study with monoclonal antibodies, J. Neurosci., 6 (1986) 650-660. 38 Vitadello, M., Triban, C., Fabris, M., Gorio, A. and Schiaffino, S., Heterogeneity of rat neurofilament polypeptides revealed by a monoclonal antibody, J. Neurochem., 46 (1986) 665-670. 39 VitadeUo, M., Triban, C., Fabris, M., Donh, M., Gorio, A. and Schiaffino, S., A developmentally regulated isoform of 150,000 molecular weight neurofilament protein specifically expressed in autonomic and small sensory neurons, Neuroscience, 23 (1987) 931-941. 40 Weber, A. and Schachner, M., Maintenance of immunocytologically identified Purkinje cells from mouse cerebellum in monotayer culture, Brain Research, 311 (1984) 119-130. 41 Willard, M. and Simon, C., Modulations of neurofilament axonal transport during development of rabbit retinal ganglion cells, Cell, 35 (1983) 551-559. 42 Wood, J.N., Lathangue, N.B., McLachlan, D.R., Smith, B.J., Anderton, B.H. and Dowding, A.J., Chromatin proteins share antigenic determinants with neurofilaments, J. Neurochem., 44 (1985) 149-154.