DEVELOPMENTAL BRAIN

RESEARCH

ELSEVIER Developmental Brain Research 90 (1995) 122-128

Research report

K252a, a potent inhibitor of protein kinases, inhibits the migration of cerebellar granule cells in vitro

Satoshi Kobayashi a,b,*, Kaoru Isa a, Kensuke Hayashi a, Hiroshi K Inoue b, Keiichi Uyemura c, Tomoaki Shirao a

a Department of Neurobiology and Behavior, Gunma University School of Medicine, Gunma, Japan b Department ofNeurosurgery, Gunma University School of Medicine, Gunma, Japan

c Department of Physiology, Keio University School of Medicine, Tokyo, Japan

Accepted 5 September 1995

Abstract

In order to elucidate the cellular mechanisms of migrating neurons, we developed an assay system in vitro, using an aggregation culture of developing granule cells from the rat cerebellum. This assay system allowed us to eliminate the effects of various factors other than neurons and to examine the direct effects of individual molecules on neuronal migration. In this assay system, we examined the effects of several protein kinase inhibitors on cerebellar granule cell migration, and revealed that K252a, an inhibitor of protein kinases and of the actions of neurotrophins, inhibited the migration. Within 5 min after the addition of K252a to the culture medium, most of the migrating spindle-shaped cells changed into non-migrating large and polygonal cells, which had many microspikes. Staining with rhodamine-phalloidin revealed the appearance of actin bundles that resembled stress fibers within these large cells. On the other hand, extension of neurites was not severely inhibited by the addition of K252a. These results suggest that the migration is regulated by a different mechanism from that of neurite growth.

Keywords: Cell migration; Granule cell; Reaggregation culture, K252a; Kinase inhibitor; Microfilament

1. Introduction

One of the critical issues in neuronal histogenesis is the control of neuronal migration in the developing brain. During the course of development of the central nervous system (CNS), neuroblasts pass through a migratory phase. In the cerebellum, for example, granule cell progenitors migrate along the cerebellar surface, and postmitotic gran- ule cells migrate radially from the germinal external granu- lar layer toward the internal granular layer along Bergmann's glial processes [21-23]. Previous studies re- vealed that several molecules, including extracellular ma- trix and cell adhesion molecules, regulated their migration. More recently Komura and Rakic have reported that intra-

* Corresponding author. Doktorsringen 12B, Plan 6, Histology, Depart- ment of Neuroscience, Berzelius Laboratory, Karolinska Institute, S-171 77 Stockholm, Sweden. Fax: (46) (8) 32 37 42.

0165-3806/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSD1 0165-3806(95)00152-2

cellular Ca 2+ dynamics and activity of NMDA receptor are implicated in the postmitotic granule cell migration [11-13]. However very little is known about the intra- cellular mechanisms through which the signals are trans- duced during neuronal migration. It seems likely that intracellular signal pathways and changes in the phospho- rylation state of cytoskeletal elements are implicated in the neuronal migration.

In this experiment, to elucidate the cellular mechanisms of migrating neurons, we developed an assay system in vitro, using an aggregation culture of developing granule cells from the rat cerebellum. This assay system allowed us to eliminate the effects on cell migration of various factors other than neurons and to evaluate cell migration quantitatively. In this assay system, we examined the effects of several protein kinase inhibitors on cerebellar granule cell migration. In the present communication, we report that K252a, an inhibitor of protein kinases and of the actions of neurotrophins, inhibits the migration of granule cells in vitro.

S. Kobayashi et al. / Developmental Brain Research 90 (1995) 122-128 123

2. Materials and methods

2.1. Materials

Culture media were from Gibco Laboratories (Grand Island, NY). Culture dishes were from Falcon (Becton Dickinson, Tokyo, Japan). Laminin, purified from mouse EHS sarcoma, was from Sigma (St. Louis, MO). Protein kinase inhibitors were purc, hased from Seikagaku Kogyo Co. (Tokyo, Japan), with the exception of K252a (Kyowa Medex, Tokyo, Japan). Poly-L-lysine hydrobromide, pu- trescine and rhodamine-phalloidin were from Sigma. In- sulin, transferrin, sodium selenite, Dispase II and DNase I were from Boehringer-Mannheim (Indianapolis, IN). The monoclonal antibody against neurofilament protein was purchased from Cosmo Bio (Tokyo, Japan). Monoclonal antibodies against /3-tubulin (125HbA2) [19] was prepared from supernatants of hybridoma cultures.

Table 1 Neurite lengths and migration indices

Neurite length ( /zm) Migration index

Control 4 h 112 ___ 8.8 (n = 6) 31 _+ 3.9 (n = 6) 6 h 238_+ 11.7 (n = 4) 2 5 + 4 . 1 ( n = 4) 8 h 156 +_ 4.4 (n = 2) 29 + 0 (n = 2) 10h 277_+ 18.0 (n = 6) 26+2 .6 (n = 6) 12h 341_+ 14.2 (n = 10) 26_+2.7 (n = 10)

K252a 80nM 12h 2 9 1 _ 1 7 . 1 ( n = 1 3 ) 1 .4+0 .66" ( n = 7 ) 200nM 12h 2 0 5 + 1 0 . 5 ( n = 1 0 ) 3.1_+1.2 * ( n = 6 )

Neurite length is defined as the average of the lengths of the longest neurite on each newly formed aggregate. The migration index is defined as the ratio of the number of neurons that migrated beyond the half neurite length to the total number of migrating neurons. Migration indices in control culture were approximately constant for various neurite lengths from 4 h to 12 h after the start of incubation. In the presence of K252a, migration indices were significantly lower than those in control cultures. Each result is given as a mean + S.E.M. An asterisk indicates a signifi- cant difference from the control value ( P < 0.001).

2.2. Substratum for cell culture 2.3. Aggregation cell cultures

Glass coverslips were first coated with poly-L-lysine (PLL). Sterile coverslips (118 mm in diameter) were im- mersed in a solution of 100 mg/ml PLL hydrobromide, washed twice with distilled water, and dried. About 50 /zl of a solution of EHS sarcoma laminin (50 mg/ml) were spotted on each PLL-coated coverslip. After 2 h at 37°C, coverslips were rinsed with serum-free (SF) culture medium. These coverslips 'were used immediately for ag- gregation cultures.

Male Wistar rats on postnatal day 2 were used for all experiments. After deep anesthesia had been induced with ether, cerebella were dissected from the rats, cleaned of meninges and dissociated into suspensions of single cells by gentle digestion with Dispase II and DNase I under sterile conditions as described previously [1,2]. After four washes in Hanks solution, cells were resuspended in Dul- becco's modified Eagle medium (D-MEM) that contained 10% fetal bovin serum at 1 × 106 cells/ml, incubated for

L

! ! .

Fig. 1. Phase-contrast photomicrographs of newly formed aggregates of cerebellar granule cells from infant rats. The aggregates were cultured on laminin in serum-free culture medium in the absence for 12 hr (A) and in the presence for 12 h (B) and for 32 h (C) of 200 nM K252a. Bar = 100 p,m.

124 s. Kobayashi et al. // Developmental Brain Research 90 (1995) 122-128

12 h at 37°C in an atmosphere of 5% CO 2 and 95% air. As a consequence aggregates of neurons reformed. The newly formed aggregates were washed and cultured on the pre- pared coverslips at 37°C in 5% CO 2 in air in SF medium (a mixture of equal volumes of D-MEM and Ham's F-12 medium supplemented with 6 m g / m l glucose, 450 m g / m l NaHCO3, 25 /zg /ml insulin, 25 ktg/ml transferrin, 25 /zg /ml sodium selenite, 6.3 ng /ml progesterone and 9.7 /zg/ml putrescine). A previous study showed that more than 95% of these cells could be considered to be granule neurons [1].

2.4. Addition of test compounds

A variety of compounds was added to SF medium and tested in the migration assay. Aliquots of stock solutions were added directly to the culture medium to give final concentrations of 1-800 nM for K252a, 20-200 /.tM for H - 7 and HA1004, 5 - 3 0 / z M for H-89 and 1-80 /xM for W-7, respectively. Although H-7 inhibits Ca2+/phospho - lipid-dependent protein kinase (C kinase), cAMP-depen- dent protein kinase (A kjnase) and cGMP-dependent pro- tein kinase (G kinase) [9], H-7 has already proven useful as a C kinase inhibitor [4,7,15]. HA1004 is a more potent inhibitor than H-7 of both A kinase and G kinase but inhibits C kinase relatively weakly. H-89 was shown to have a potent and selective inhibitory action against A kinase [6]. As an antagonist of calmodulin, W-7 was used instead of KN-62 for inhibition of calmodulin-dependent protein kinase II (CaM kinase II) since KN-62 was toxic under our culture conditions and could not be used in migration assays. The stock solution of K252a was pre- pared in dimethylsulfoxide (DMSO), at a concentration of 2 mM, and stored at - 2 0 ° C in darkness.

0.1% Triton X-100 in PBS for 10 min. They were incu- bated with 3% bovine serum albumin in PBS for 30 min and then with diluted first antibody for 2 h. Coverslips were washed three times for 10 min each in PBS and then incubated for 30 min with fuorescein isothiocyanate (FITC) -labeled antibodies raised in goat against mouse Ig G (diluted 1:100 in PBS with 3% bovine serum albumin). Finally, coverslips were washed and mounted with Perma Fluor (Immunon, Pittsburgh, PA). For labeling of actin filament, the fixed and detergent-treated samples were incubated with rhodamine-conjugated phalloidin for 1 h. They were observed under a fluorescence microscope and with Nomarski differential interference contrast optics (DIC; Olympus). All staining procedures were performed at room temperature.

2.5. Quantification of the outgrowth of neurites and cell migration

After taking photographs of the reaggregated neurons under a phase-contrast microscope (Nikon, Tokyo, Japan), we measured the distance between the margin of each aggregate and the tip of the longest neurite. We determined the numbers of cells located within given distances from the margin of each aggregate [2]. A migration index was calculated as the ratio of the number of neurons that migrated beyond half of the neurite length to the total number of migrating neurons. The neurite length was defined as the mean of the lengths of longest 10 neurites in each aggregate.

2.6. Immunocytochemistry

Aggregates were fixed in 4% paraformaldehyde in phosphate-buffered saline (pH 7.4; PBS) for 15 rain. Fixed samples were immersed in a permeabilizing solution of

i i ~ ' ~ - 7 = ! ~ - ~ i . . . . . . . . . . . . . ~ i :!~'¸'~< ~ ~ " .....

5 : ~ , ,

Fig. 2. Phase-contrast photomicrographs of newly formed aggregates of ccrebellar granule cells in vitro. The aggregates of granule cells were cultured on laminin in culture medium supplemented with 200 nM K252a for 6 h (A) and then in control medium for 6 h (B). After removal of K252a and replacement of medium with fresh control medium, neurons migrated outwards once again. Bar = 100 pro.

S. Kobayashi et al. / Developmental Brain Research 90 (1995) 122-128 125

3. Results 3.2. Inhibition of migration by K252a

3.1. Migration of cells on laminin

The extensive migration of many small cells occurred on the laminin-coated cover,dips (Fig. 1A). Within ten min after plating, neuron-like cells with neurites started to migrate outwards from the newly formed aggregates. The cells migrated radially from the aggregates. The outgrowth of neurites and the migration of cells persisted for at least 16 h under our culture conditions (Table 1). In order to evaluate the migratory activity independently of neurite outgrowth, we introduced a 'migration index', namely, the ratio of cells that migrated beyond half the neurite length to the total number of migrating cells. As shown in Table 1, the migration indices in control cultures were approxi- mately constant for various neurite lengths from 4 h to 12 h after the start of incubation. Since the cells were spindle-shaped and were not immunostained with glial fibrillary acidic protein (GFAP)-specific antiserum serum, they seemed to be neurons.

K252a, at a concentration of 80 nM, significantly inhib- ited the migration of cells on laminin-coated coverslips during a 12-h incubation (Table 1). The inhibiton of migration was also significant at a concentration of 200 nM although outgrowth of neurites was also slightly inhib- ited at this concentration after a 12-h incubation (Fig. 1B). It was clear that the cells were not damaged by 200 nM K252a since the effects of K252a at that concentration were reversible. Removal of K252a and replacement with fresh SF medium allowed reinitiation of neuronal migra- tion (Fig. 2).

3.3. Cell shape and the cytoskeleton of migrating cells

Most of the cells that migrated out from the newly formed aggregates in control cultures had a bipolar, spin- dle-shaped morphology and a diameter of 7.7 _+ 0.12 /xm (mean _+ S.E.M., n = 50). Their cell bodies were attached to the growing neurites (Fig. 3A). This specialized mor-

i :

ii~ii

Fig. 3. Immunocytochemical staining of migrating granule cells. The newly formed aggregates were cultured on laminin in the control culture medium in the absence of K252a. Photographs show Nomarski differential interference contrast optics (A), staining with rhodamine-phalloidin (B), and immuno- staining with antibodies against /3-tubulin (C) and with antibodies against neurofilament protein (D). An arrow in C indicates immunostaining of/3-tubulin observed in the cytoplasm as well as in the submembranous regions. Bar = 10 p.m.

126 S. Kobayashi et al. / Der, elopmental Brain Research 90 (1995) 122-128

phology was consistent with that observed in previous studies of glial-guided neuronal migration [16,17].

Staining with rhodamine-phalloidin revealed that F-actin was localized in the submembranous regions of the spin- dle-shaped migrating neurons (Fig. 3B). Immunostaining of /3-tubulin was observed in the cytoplasm (arrow in Fig. 3C) as well as in the submembranous regions. A fine granular pattern of immunostaining of scattered neurofila- ment protein was visible in the cells (Fig. 3D).

3.4. Cell shape and the cytoskeleton of arrested cells

Since, in the presence of K252a, almost no granule cells migrated from the newly formed aggregates, it was diffi- cult to analyze their shapes and cytoskeletons. Therefore, K252a was added directly to the culture medium (final concentration: 200 nM) after cells had migrated from the aggregates in SF culture medium without K252a for 6 to 12 h. Within five min of the addition of K252a at 200 nM, most of the migrating spindle-shaped cells appeared to change into large and polygonal, dark, fiat cells under the phase-contrast microscope. The diameter of these large cells was 23 + 0.57 /xm (mean + S.E.M., n = 50), and

this diameter was significantly greater than that of cells in the control medium. The cells also had many microspikes, which were not observed on the cells in control cultures (Fig. 4A).

Staining with rhodamine-phalloidin revealed stress fiber-like actin bundles within these large cells (Fig. 4B). The pattern of immunostaining of /3-tubulin also changed within these large cells. The fine tubulin meshwork was visible in the cytoplasm of them (Fig. 4C). The patterns of immunostaining of neurofilament proteins within these large polygonal cells were similar to those in control spindle-shaped cells (Fig. 4D).

3.5. Possible target of K252a in migrating cells

Since K252a is known to inhibit a variety of protein kinases [8], we examined the effects of inhibitors of A kinase, C kinase and CaM kinase II on cell migration (Table 2). H-89, an inhibitor of A kinase, at 15 /zM and W-7, an inhibitor of CaM kinase II, at 10 /xM did not inhibit cell migration. These inhibitors could not be used at higher concentrations because of their toxicity in our cul- ture system. At 200 /zM, H-7, an inhibitor of C kinase,

i

Fig. 4. Immunocytochemical staining of non-migrating granule cells. The newly formed aggregates were cultured on laminin in the control culture medium in the absence of K252a for 6 h and then in the same medium in the presence of 200 nM K252a for 10 min (A and B), and 6 h (C and D). Photographs show Nomarski differential interference contrast optics (A), staining with rhodamine-phalloidin (B), and immunostaining with antibodies against /3-tubulin (C) and against neurofilament protein (D). Bar = 10 /xm.

S. Kobayashi et al. / Developmental Brain Research 90 (1995) 122-128 127

Table 2 Neurite lengths and migration indices in the presece of various inhibitors of protein kinases

Neurite length (/zm) Migration index

Control 341 _+ 14.2 (n = 10) 27 + 2.7 (n = 10) K252a 2 0 0 n M 205_+10.5 * (n=10) 3.1+1.2 ~ (n=6) H-89 15/xM 225+25.3 * (n=5) 25+4.7(n=5) W-7 10 brM 246_+8.13 * (n = 6) 27_+3.5 (n = 6) H-7 200/xM 198_+7.87 * (n=10) 12_+1.7 * (n=9) HA1004 200 ArM 309_+24.1 (n = 11) 24+1.5 (n = 5)

Neurite lengths and migration in6ices after incubation for 12 h in the presence of various agents are shown. Each result represents a mean +S.E.M.. Migration indices for H-7 and K252a are significantly lower than those for control cultures (see Table 1). Migration indices for the other agents are not significantly different from those for the control cultures irrespective of neurite length. ( * : P < 0.001)

inhibited neurite outgrowth to a similar extent to K252a at 200 nM. Although H-7 also inhibited cell migration, the migration index at 200 p,M H-7 was 12%, while a value of only 3% was obtained for the migration index at 200 nM K252a, at which the outgrowth of neurites was inhibited to the same extent as it was by 200 /xM H-7. HA1004 did not have any significant effects on the outgrowth of neu- rites or cell migration.

4 . D i s c u s s i o n

In the present study, we demonstrated that K252a, which is an inhibitor of protein kinases and of neu- rotrophic activity, inhibited the migration of granule cells in a reaggregation culture system. In our assay system, both the cells in the newly formed aggregates and the migrating cells did not express GFAP, which is a marker of astroglia. Asou reported that more than 95% of cells in a single-cell suspension of 6-day-old mouse cerebellum were immunocytochemically identifiable as granule neu- rons [1]. Nagata and Nakatsuji demonstrated in GABA-up- take experiments that more than 90% of cells that migrated out from mouse cerebellar explants on laminin were gran- ule cells [16]. These results indicated that our assay system allowed us to eliminate the effects of various factors other than neurons, and to examine the direct effects of individ- ual molecules on neuronal migration. In this respect, our assay system is superior to the other published assay systems, including slice cultures and living animals, al- though they have provided valuable information about possible mechanisms of granule cell migration in vivo.

The extent of cell migration could be examined inde- pendently of neurite growth by use of the migration index. Our results showed that the migration was specifically inhibited by K252a, even though neurite growth was also inhibited to a certain extent. It was suggested that the migration was regulated by a different mechanism from that of neurite growth. Since K252a is also known to inhibit various protein kinases [8], we examined the effects

of inhibitors of A kinase, C kinase and CaM kinase II on cell migration. Neither H-89, an inhibitor of A kinase, nor W-7, a calmodulin antagonist, inhibited cell migration. These results demonstrate that the specific inhibition by K252a of cell migration did not involve A kinase or CaM kinase II. However, migration indices in the presence of H-7 were lower than those in control cultures. Although H-7 inhibits A kinase and G kinase as well, this inhibitory effect on cell migration could be linked specifically to inhibition of C kinase because a related drug, HA1004 did not inhibit cell migration in our assay system [4,5,9]. Inhibition by H-7 was less conspicuous than that by K252a, and the characteristic morphological and cytoskeletal changes seen in the presence of K252a were not observed in that of H-7 (data not shown). These observations indi- cate that the mechanism of action of K252a is different from that of H-7.

Recent studies demonstrated that K252a inhibits the neurotrophic activities of NGF, brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3) by the inhibit- ing autophosphorylation of the respective receptors [3,10,18,20,25]. Furthermore, it has been reported that BDNF and NT-3 are involved in the migration of granule cells in the developing cerebellum [14,24,26]. Our results and those of previous studies suggest that K252a inhibits cell migration by inhibiting neurotrophic activity.

In the presence of K252a, characteristic morphological and cytoskeletal changes in granule cell were observed. Small spindle-shaped cells were attached to the neurites in control cultures, and they migrated outwards at a consider- able rate. In the presence of K252a, cell migration was inhibited, and the majority of spindle-shaped granule cells became enlarged and flat. Stress fibers, namely bundles of actin filaments were observed in the cytoplasm of the large, flat cells. It appeared that small spindle-shaped cells had the ability to migrate along the neurites, and that the large, flat granule cells did not. Our immunocytochemical studies revealed that the localization of F-actin in granule cells changed drastically as the shape of cells changed. This fact suggests that microfilament system plays impor- tant roles in neuronal migration.

Very little is known about the mechanism for regulation of the cytoskeletons in migrating neurons. In the present experiments, K252a, an inhibitor of neurotrophic activities and protein kinases, changed the localization of the cyto- skeletal elements through a signal pathway other than those of A kinase, C kinase and CaM kinase II. What signal pathways are involved, and what changes of the other cytoskeletal elements are implicated in neuronal migration remain to be identified.

A c k n o w l e d g e m e n t s

The authors are grateful to Dr. Hiroaki Asou (Keio University) for his advice and Professor Lars Olson (De-

128 S. Kobayashi et aL /Developmental Brain Research 90 (1995) 122-128

partment of Neuroscience, Karolinska Institute) for his valuable comments on this manuscript. This work is sup- ported in part by a Grant-in-Aid for scientific research from the Ministry of Education, Culture and Science, Japan.

References

[1] Asou, H., Miura, M., Kobayashi, M., Uyemura, K. and Itoh, K., Cell adhesion molecule L1 guides cell migration in primary reaggrega- tion cultures of mouse cerebellar cells, Neurosci. Lett., 144 (1992) 221-224.

[2] Asou, H., Miura, M., Kobayashi, M. and Uyemura, K., The cell adhesion molecule L1 has a specific role in neural cell migration, NeuroReport, 3 (1992) 481-484.

[3] Berg, M.M., Sternberg, D.W., Parada, L.F. and Chao, M.V., K-252a inhibits nerve growth factor-induced trk proto-oncogene tyrosine phosphorylation and kinase activity, J. Biol. Chem., 267 (1992) 13-16.

[4] Bixby, J.L., Protein kinase C is involved in laminin stimulation of neurite outgrowth, Neuron, 3 (1989) 287-297.

[5] Bixby, J.L. and Jhabvala, P., Extracellular matrix molecules and cell adhesion molecules induce neurites through different mechanisms, J. Cell Biol., 111 (1990)2725-2732.

[6] Chijiwa, T., Mishima, A., Hagiwara, M., Sano, M., Hayashi, K., Inoue, T., Naito, K., Toshioka, T. and Hidaka, H., Inhibition of forskolin-induced neurite outgrowth and protein phosphorylation by a newly synthesized selective inhibitor of cyclic AMP-dependent protein kinase, N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinoline sulfonamide (H-89), of PC12D pheochromocytoma cells, J. Biol. Chem., 265 (1990) 5267-5272.

[7] Conn, P. J., Strong, J. A., Azhderian, E. M., Narin, A. C., Green- gard, P. and Kaczmarek, L.K., Protein kinase inhibitors selectively block phorbol ester- or forskolin-induced changes in excitability of Aplysia neurons, J. Neurosci., 9 (1989) 473-479.

[8] Hashimoto, Y., Nakayama, T., Teramoto, T., Kato, H., Watanabe, T., Kinoshita, M., Tsukamoto, K., Tokunaga, K., Kurokawa, K., Nakanishi, S., Matsuda, Y. and Nonomura, Y., Potent and prefer- ential inhibition of Ca 2÷/calmodulin-dependent protein kinase II by K252a and its derivative, KT5926, Biochem. Biophys. Res. Com- mun., 181 (1991) 423-429.

[9] Hidaka, H., Inagaki, M., Kawamoto, S. and Sakaki, Y., Isoquinoli- nesulfonamides, novel and potent inhibitors of cyclic nucleotide-de- pendent protein kinases and protein kinase C, Biochemistry, 23 (1984) 5036-5041.

[10] Kniisel, B., Kaplan, D.R., Winslow, J.W., Rosenthal, A., Burton, L.E., Beck, K.D., Rabin, S., Nikolics, K. and Hefti, F., K-252b selectively potentiates cellular actions and trk tyrosine phosphoryla- tion mediated by neurotrophin-3, J. Neurochem., 59 (1992) 715-722.

[11] Komuro, H. and Rakic, P., Selective role of N-type calcium chan- nels in neuronal migration, Science, 257 (1992) 806-809.

[12] Komuro, H. and Rakic, P., Modulation of neuronal migration by NMDA receptors, Science, 260 (1993) 95-97.

[13] Komuro, H. and Rakic, P., Dynamics of granule cell migration: a confocal microscopic study in acute cerebellar slice preparations, J. Neurosci., 15 (1995) 1110-120.

[14] Maisonpierre, P.C., Belluscio, L., Friedman, B., Alderson, R.F., Wiegand, S.J., Furth, M.E., Lindsay, R.M. and Yancopoulos, G.D., NT-3, BDNF, and NGF in the developing rat nervous system: parallel as well as reciprocal patterns of expression, Neuron, 5 (1990) 501-509.

[15] Malinow, R., Madison, D.V. and Tsien, R. W., Persistent protein kinase activity underlying long-term potentiation, Nature, 335 (1988) 820-824.

[16] Nagata, I. and Nakatsuji, N., Granule cell behavior on laminin in cerebellar microexplant cultures, Dev. Brain Res., 52 (1990) 63-73.

[17] Nakatsuji, N. and Nagata, I., Paradoxical perpendicular contact guidance displayed by mouse cerebellar granule cell neurons in vitro, Development, 106 (1989) 441-447.

[18] Nye, S.H., Squinto, S.P., Glass, D.J., Stitt, T.N., Hantzopoulos, P., Macchi, M.J., Lindsay, N.S., Ip, N.Y. and Yancopoulos, G.D., K-252a and staurosporine selectively block autophosphorylation of neurotrophin receptors and neurotrophin-mediated responses, Mol. Biol. Cell, 3 (1992) 677-683.

[19] Obata, K., Nishiye, H., Fujita, S.C., Shirao, T., Inoue, H.K. and Uchizono, K., Identification of a synaptic vesicle-specific 38,000- Dalton protein by monoclonal antibodies, Brain Res., 375 (1986) 37-48.

[20] Ohmichi, M., Decker, S. J., Pang, L. and Saltiel, A. R., Inhibition of the cellular actions of nerve growth factor by staurosporine and K252a results from the attenuation of the activity of the trk tyrosine kinase, Biochemistry, 31 (1992) 4034-4039.

[21] Rakic, P., Neuron-glia relationship during granule cell migration in developing cerebellar cortex. A Golgi and electronmicroscopic study in Macacus rhesus, J. Comp. Neurol., 141 (1971) 283-312.

[22] Rakic, P., Mode of cell migration of the superficial layers of fetal monkey neocortex, J. Comp. Neurol., 145 (1972) 61-84.

[23] Rakic, P., Principles of neural cell migration, Experientia, 46 (1990) 882-891.

[24] Rocamora, N., Garcia-Ladona, F.J., Palacios, J.M. and Mengod, G., Differential expression of brain-derived neurotrophic factor, neu- rotrophin-3, and low-affinity nerve growth factor receptor during the postnatal development of the rat cerebellar system, Mol. Brain Res., 17 (1993) 1-8.

[25] Tapley, P., Lamballe, F. and Barbacid, M., K252a is a selective inhibitor of the tyrosine protein kinase activity of the trk family of oncogenes and neurotrophin receptors, Oncogene, 7 (1992) 371-381.

[26] Wanaka, A. and Johnson Jr., E.M., Developmental study of nerve growth factor receptor mRNA expression in the postnatal rat cere- bellum, Dev. Brain Res., 55 (1990) 288-292.