Ž .Developmental Brain Research 102 1997 305–308

Short communication

Cerebellar granule cells elaborate neurites before mitosis

Evan Wolf a, Joseph P. Wagner a, Ira B. Black a,b, Emanuel DiCicco-Bloom a,b,c,)

a Department of Neuroscience and Cell Biology, UniÕersity of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, 675 HoesLane, Piscataway, NJ 08854, USA

b Cancer Institute of New Jersey, UniÕersity of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, 675 Hoes Lane, Piscataway,NJ 08854, USA

c Department of Pediatrics, UniÕersity of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, 675 Hoes Lane, Piscataway, NJ08854, USA

Accepted 17 June 1997

Abstract

Neuronal birth and neurite outgrowth have been regarded as discrete, sequential stages of development. However, we recently foundwthat sympathetic neuroblasts often elaborate axons before mitosis, in culture E. Wolf, I.B. Black, E. DiCicco-Bloom, Mitotic neuroblasts

Ž . x wdetermine neuritic patterning of progeny, J. Comp. Neurol. 367 1996 623–635 and in vivo E. Wolf, I.B. Black, E. DiCicco-Bloom,Ž .Central and peripheral neuroblasts elaborate neurites prior to division in vivo and in vitro, Soc. Neurosci. Abstr. 21 1995 785; E. Wolf,

Ž .I.B. Black, E. DiCicco-Bloom, Mitotic neuroblasts engage in axonal outgrowth and pathfinding in vivo, Soc. Neurosci. Abstr. 22 1996x525 . Here, we report that cerebellar granule cells often divide with heritable neurites in vitro. Therefore, mitotic CNS precursors, in

addition to peripheral neuroblasts, simultaneously undergo proliferation and process formation. Potentially, neurites on dividingprecursors may allow target fields to influence directly the course of neurogenesis. q 1997 Elsevier Science B.V.

Keywords: Neurogenesis; Proliferation; Neurite outgrowth; Microtubule

Neuroblast proliferation and axodendritic outgrowth un-derlie neural regionalization and connectivity. Traditionalconcepts suggest that the elaboration of neuronal processesnecessarily follows cessation of neuroblast proliferation.This view emerged from classical studies of mammaliantelencephalon development describing migration and dif-ferentiation of postmitotic neurons during the formation of

Žw x w x.cortical cytoarchitecture 17 ; for review, see 11 . Fur-ther, studies of cell division in vitro describe tubulinaggregation in the mitotic spindle and actin localization in

w xthe cytokinetic cleavage furrow 12,1 ; these cytoskeletalelements are presumably not readily available for mitosiswhen assembled into neuritic cytoarchitecture.

However, we recently demonstrated that embryonicw x w xsympathetic neurons in culture 22 and in vivo 21,23

often elaborate long, complex processes before cell divi-Ž .sion paramitotic neurites . Thus, proliferation and neurite

outgrowth need not be sequential in all neuronal popula-tions. Moreover, paramitotic neurites elaborated by parental

) Ž .Corresponding author. Fax: q1 732 235-4990; E-mail:diciccem@umdnj.edu

Ž w x.neuroblasts were maintained during mitosis see also 14and were faithfully preserved by daughter doublets after

w xcell division 22 . This morphologic ‘inheritance’ raises thepossibility that dividing neuronal precursors engage inpathfinding, interacting with target fields, potentially coor-dinating neurogenesis with the formation of circuit archi-

w xtecture 22 .Are sympathetic neuroblasts unique in their capacity to

coordinate mitosis and process formation, or does thisŽ .phenomenon occur in central nervous system CNS neu-

roblasts as well? We have addressed this issue usingvirtually pure and highly characterized cerebellar granule

w xneuroblast cultures as a model CNS population 5,7,19 .While ontogenetic stages such as proliferation and neuriteoutgrowth have been explored for granule neuroblasts on a

Ž w x.population basis for review, see 8 , the sequence ofdevelopmental events in individual cells remains incom-pletely described.

Serum-free cultures of dissociated cerebellar granuleneuron precursors were obtained from postnatal day 7 ratsusing Percoll gradient centrifugation, as previously re-

w xported 5,7,19 . Neurites present during S-phase of the cellŽ .cycle were identified by combining microtubule MT

0165-3806r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0165-3806 97 00111-9

( )E. Wolf et al.rDeÕelopmental Brain Research 102 1997 305–308306

w3 ximmunostaining with H thymidine incorporation auto-w x Ž .radiography 22 . In brief, cells were plated onto poly D -Ž .lysine-coated 0.1 mgrml plastic culture dishes contain-

w x Ž . Žing defined medium 3,4 , insulin 10 mgrml , laminin 10. Ž .mgrml and fibronectin 5 mgrml . Cultures were treated

w3 x Ž .for 2 h with H thymidine 1 mCirml; Dupont Co., MAafter 22 h of incubation. At 24 h, after stabilizing polymer-ized b-tubulin, tubulin monomers were extracted and cellswere fixed. Indirect immunostaining was performed with a

Žmonoclonal IgG specific for b-tubulin 1 : 1000 dilution;. w xAmersham Corp., IL 2,22 and dishes were processed for

w xautoradiography 3,4 . Combined MT staining and thymi-dine labeling were assessed at 400= magnification, usinga Zeiss Axiophot microscope equipped with brightfield,phase-contrast, and FITC optics. Materials were fromSigma Chemical Co. unless otherwise specified. For com-bined neurite and S-phase labeling, experimental groupsconsisted of 3 dishes and 3 rows of ;100 S-phase labeledcells were counted per dish. Data were combined from twoexperiments. Serial time-lapse photography was performedusing a Zeiss Axiovert 10 microscope with phase optics at320= magnification. Forty-eight microscope fields were

Ž .photographed 12 fields in each of 4 dishes at 4, 8, 19, 22,26, and 31 h of incubation.

Cerebellar granule neuroblasts often engaged in processoutgrowth during S-phase of the cell cycle, as indicated bycolocalization of MT-positive neurites and nuclearw3 x Ž .H thymidine incorporation Fig. 1 . In fact, under certainconditions, over 50% of S-phase neuroblasts had neuritesŽ .see below . Overall, with various culture conditions, neu-

Ž .rites were present on 962 of 3628 26.5% S-phase neurob-w xlasts. We have termed these neurites ‘paramitotic’ 22 to

denote processes greater than one cell body diameter inŽ .length i.e. )12 mm that are present at any time during

the cell cycle. Paramitotic neurites were diverse in lengthand shape, and exhibited MT staining that was generallysimilar to that of neurites on non-mitotic cells.

A striking feature of sympathetic paramitotic neuriteswas their consistent morphologic inheritance; neurites pre-sent before cell division were always present after cell

w xdivision with unaltered morphology 22 . To determinewhether cerebellar granule cell paramitotic neurite mor-phology was inherited after cell division, serial pho-tographs of 48 microscope fields were obtained over aperiod of 31 h. Ninety-four cells divided in this time-lapseseries, and 46 of these cells had neurites before mitosis.More than half of the cells with neurites before mitosisŽ .25r46 exhibited inheritance of neuritic morphology:paramitotic neurites had qualitatively similar morphology

Žbefore and after mitosis i.e. shape, length, width, and.orientation, e.g., Fig. 2 . As in the case of sympathetic

neuroblasts, this preservation of neuritic patterning oc-curred through asymmetric cell division, whereby neuriteswere differentially allocated to daughter cells. However,unlike sympathetic neuroblasts, not all granule cell parami-totic neurites were inherited: of the 56 neurites elaborated

w3 xFig. 1. Neuroblasts with MT-containing neurites incorporate H thymi-dine. Cerebellar granule cells incubated for 22 h received a 2 h pulse ofw3 xH thymidine and were then fixed and processed for combined b-tubulin

w3 ximmunostaining and autoradiography. Every neurite on 3628 H thymi-Ž .dine-labeled cells S-phase contained polymerized tubulin along the

entire length. One microscope field with three cells is illustrated. Epifluo-Ž . Ž .rescence a shows that each cell has one long, microtubule MT -contain-

Ž . Žing neurite. Brightfield b reveals that the cell on the far left arrows in a,.b has prominent silver grains over its nucleus, while maintaining a

Ž .neurite arrowhead, a . In this assay, the intensity of neuritic MT stainingw3 xvaried, but did not correlate with H thymidine incorporation. Neurites

in control cultures processed without primary antibody were completelyŽ w x.unlabeled not shown, see 22 . Bar, 30 mm.

Ž .by parental cells, only 33 59% were present on daughterprogeny. The remaining neurites were either remodeled, or

Ž .regressed completely following cell division not shown .Interestingly, 4 cells exhibited both inherited and non-in-herited neurites, suggesting that mechanisms underlyingmorphologic preservation may be neurite-specific, ratherthan cell-specific.

Regulation of paramitotic neuritogenesis may involveintracellular andror extracellular cues. To begin definingregulatory signals, we assessed the incidence of parami-

Ž .totic neurites under different culture conditions Fig. 3 . InŽ .the presence of insulin and poly D -lysine, required for

Žsurvival and adhesion in serum-free medium unpublished. Ž .observations , 1.7"0.4% mean"S.E.M., ns1802 cells

of S-phase cells exhibited neurites. However, the additionof laminin and fibronectin increased neurite elaboration by

ŽS-phase cells to 50.9"0.9% mean"S.E.M., ns1826.cells . Thus, neurite elaboration by mitotic neuroblasts was

dramatically increased in the presence of extracellularŽ w x.neuritogenic cues for reviews, see 15,16 .

Our observations indicate that neuronal precursors fromthe mammalian CNS have the capacity to simultaneously

( )E. Wolf et al.rDeÕelopmental Brain Research 102 1997 305–308 307

divide and elaborate neurites: the presence of neuriticcytoskeleton does not prevent cell mitosis. In conjunctionwith previous characterization of this phenomenon in sym-

w xpathetic neuroblasts 14,21–23 , these observations suggestthat mitosis and neuritic process outgrowth may occur

Fig. 2. Neuritic morphology can be preserved during granule cell prolifer-ation. Cultures were photographed sequentially at the time points indi-cated in hours. In this example, the granule cell indicated by an arrow at

Ž .8 h has elaborated a neurite arrowhead by 19 h in culture. This cellŽ .divides 22 h and the neurite, allocated to the lower daughter cell, is

Žrestored to its original morphology over the next several hours compare.images at 19 and 31 h . In this experiment, 46 cells had neurites before

Ž .they divided, and more than half of these cells 25r46 exhibited‘inheritance’ of neuritic morphology: qualitatively similar morphology

Ž .before and after mitosis. Overall, 33 of 56 59% paramitotic neuritesappeared to be inherited. Bar, 30 mm.

Fig. 3. Extracellular factors increase the incidence of paramitotic neurites.Granule cell cultures were processed as described for Fig. 1. The percent-

w3 x Ž .age of H thymidine-labeled cells S-phase bearing MT-positive neu-rites were scored in each group. The data are the sum of two experimentsand are expressed as mean percent"S.E.M. ) Differs from control atp-0.0001.

simultaneously in several neural regions. Further, sincedividing neuronal precursors are susceptible to cell deathw x w x10,24 and may require specific trophic support 4 , wespeculate that cell division, and survival of dividing cells,may be regulated by distant environmental signals received

w xthrough neurites or mature axons and dendrites 22,23 , inw xaddition to local and intrinsic cellular cues 11,16 .

Our observations on cultured granule neuroblasts arerelevant to recent infrared video microscopic studies of

Ž .precursors in the external germinal layer EGL , recordedfrom intact, living brain sections of newborn cerebellumw x6 . Precursors in the proliferative EGL are frequentlybipolar, exhibiting one process attached to the externalbasement membrane overlying the surface, and a secondradial process extending vertically towards the molecularlayer. By postnatal day 7, the age we performed ourstudies, these processes are well developed in vivo, extend-

Ž w x.ing to the Purkinje cell layer see Figs. 1, 2C in Ref. 6 .While the authors focus on nuclear translocation in theseneuronal processes as a means of granule cell migrationfrom the EGL, it is quite likely that a proportion ofproliferative EGL precursors also possess radial neuronalprocesses extending toward targets, compatible withparamitotic neurites we observed in vitro.

Although studies remain to be performed in vivo, theseobservations in culture and in intact brain sections raise the

( )E. Wolf et al.rDeÕelopmental Brain Research 102 1997 305–308308

possibility that interactions of granule cell processes withPurkinje cell dendrites or secreted factors underlie recipro-cal relationships between these cells during development.For example, the temporal and spatial dependence of gran-

w xule neuron number on Purkinje cell number 9 may bew xactively coordinated through Purkinje-derived trophic 20

w xandror mitogenic 18 signals, retrogradely transported todividing granule neuroblasts. Specifically, Purkinje cells

Ž w x.produce basic fibroblast growth factor bFGF, 13 , whichw xmay stimulate granule cell mitosis 5,19 . Thus, numerical

matching between populations may not be solely a lateevent during development, but may be ongoing even ascells are being generated. Therefore, in many neural re-gions, the component ontogenetic processes, including pro-liferation, cell death, migration, and axodendritic out-growth, may proceed concurrently to generate complexbrain architecture.

Acknowledgements

This work was supported by NIH Grant HD 23315Ž .E.D.-B., I.B.B. .

References

w x1 N. Akkas, Biomechanics of Cell Division, Plenum Press, New York,1987, 374 pp.

w x2 M.M. Black, T. Slaughter, I. Fischer, Microtubule-associated proteinŽ .1b MAP 1b is concentrated in the distal region of growing axons,

Ž .J. Neurosci. 14 1994 857–870.w x3 E. DiCicco-Bloom, I.B. Black, Insulin growth factors regulate the

mitotic cycle in cultured rat sympathetic neuroblasts, Proc. Natl.Ž .Acad. Sci. USA 85 1988 4066–4070.

w x4 E. DiCicco-Bloom, W.J. Friedman, I.B. Black, NT-3 stimulatessympathetic neuroblast proliferation by promoting precursor sur-

Ž .vival, Neuron 11 1993 1101–1111.w x5 W.-Q. Gao, N. Heintz, M.E. Hatten, Cerebellar granule cell neuro-

genesis is regulated by cell–cell interactions in vitro, Neuron 6Ž .1991 705–715.

w x6 G. Hager, H.-U. Dodt, W. Zieglgansberger, P. Liesi, Novel forms ofŽ .neuronal migration in the rat cerebellum, J. Neurosci. Res. 40 1995

207–219.w x7 M.E. Hatten, Neuronal regulation of astroglial morphology and

Ž .proliferation in vitro, J. Cell Biol. 100 1985 384–396.

w x8 M.E. Hatten, N. Heintz, Mechanisms of neural patterning andspecification in the developing cerebellum, Annu. Rev. Neurosci. 18Ž .1995 385–408.

w x9 K. Herrup, K. Sunter, Numerical matching during cerebellar devel-opment: quantitative analysis of granule cell death in staggerer

Ž .mouse chimeras, J. Neurosci. 7 1987 829–836.w x10 S. Homma, H. Yaginuma, R.W. Oppenheim, Cell death during the

earliest stages of spinal cord development in the chick embryo: apossible means of early phenotypic selection, J. Comp. Neurol. 345Ž .1994 377–395.

w x11 M. Jacobson, Developmental Neurobiology, 3rd ed., Plenum Pub-lishing Corp., New York, 1991, 776 pp.

w x12 M. Kirschner, T. Mitchison, Beyond self-assembly: from micro-Ž .tubules to morphogenesis, Cell 45 1986 329–342.

w x13 S. Matsuda, N. Okumura, H. Yoshimura, Y. Koyama, M. Sakanaka,Basic fibroblast growth factor-like immunoreactivity in Purkinje

Ž .cells of the rat cerebellum, Neuroscience 50 1992 99–106.w x14 S.P. Memberg, A.K. Hall, Dividing neuron precursors express neu-

Ž .ron-specific tubulin, J. Neurobiol. 27 1995 26–43.w x15 V. Nurcombe, Laminin in neural development, Pharmacol. Ther. 56

Ž .1992 247–264.w x16 L.F. Reichardt, K.J. Tomaselli, Extracellular matrix molecules and

their receptors: functions in neural development, Annu. Rev. Neu-Ž .rosci. 14 1991 531–570.

w x Ž .17 F.C. Sauer, Mitosis in the neural tube, J. Comp. Neurol. 62 1935377–405.

w x18 R.J. Smeyne, T. Chu, A. Lewin, F. Bian, S. S.-Crisman, C. Kunsch,S.A. Lira, J. Oberdick, Local control of granule cell generation by

Ž .cerebellar Purkinje cells, Mol. Cell. Neurosci. 6 1995 230–251.w x19 Y. Tao, I.B. Black, E. DiCicco-Bloom, Neurogenesis in neonatal rat

brain is regulated by peripheral injection of basic fibroblast growthŽ . Ž .factor bFGF , J. Comp. Neurol. 376 1996 653–663.

w x20 M.W. Vogel, K. Sunter, K. Herrup, Numerical matching betweengranule and Purkinje cells in lurcher chimeric mice: a hypothesis forthe trophic rescue of granule cells from target-related cell death, J.

Ž .Neurosci. 9 1989 3454–3462.w x21 E. Wolf, I.B. Black, E. DiCicco-Bloom, Central and peripheral

neuroblasts elaborate neurites prior to division in vivo and in vitro,Ž .Soc. Neurosci. Abstr. 21 1995 785.

w x22 E. Wolf, I.B. Black, E. DiCicco-Bloom, Mitotic neuroblasts deter-Ž .mine neuritic patterning of progeny, J. Comp. Neurol. 367 1996

623–635.w x23 E. Wolf, I.B. Black, E. DiCicco-Bloom, Mitotic neuroblasts engage

in axonal outgrowth and pathfinding in vivo, Soc. Neurosci. Abstr.Ž .22 1996 525.

w x24 H. Yaginuma, M. Tomita, N. Takashita, S.E. McKay, C. Cardwell,Q.-W. Yin, R.W. Oppenheim, A novel type of programmed neuronaldeath in the cervical spinal cord of the chick embryo, J. Neurosci. 16Ž .1996 3685–3703.