~eurosrienre Vol. 15, No. I, pp. 203-213. 1985 Printed in Great Britain

03064522/85 $3.00 + 0.00

Pergamon Press Ltd c 1985 IBRO

IMPAIRED DEVELOPMENT OF CEREBELLAR CORTEX IN RATS TREATED POSTNATALLY WITH

a-DIFLUOROMETHYLORNITHINE

J. V. BARTOLOME,* L. ScHwErTzERt T. A. SLOTKIN* and J. V. NADLER* Departments of *Pharmacology and tAnatomy, Duke University Medical Center, Durham, NC 27710.

U.S.A.

Abstract-z-Difluoromethylomithine specifically and irreversibly inhibits the enzyme omithine decar- boxylase. Omithine decarboxylase catalyses the initial step in the synthesis of polyamines, which are thought to play an essential role in growth and development of mammalian tissues. The current study examined the effects of a-difluoromethylomithine on the ontogenic development of the rat cerebellar cortex. Animals injected daily with adifluoromethylomithine on postnatal days 1-21 suffered a deficit in the number of granllle cells and many of the remaining granule cells became trapped in the molecular layer during migration. Purkinje cells were also scattered throughout the molecular layer and their mean diameter was 38% smaller than in controls. In general, the cerebellar cortex of adifluoromethylomithine- treated rats failed to progress much beyond the stage of development reached in control rats during the first postnatal week. These effects of adifluoromethylomithine were already clearly visible at 10-l 5 days of age. The final size of the cerebellum as a whole and of individual folia was markedly subnormal.

These data indicate that polyamines play an obligatory role in cerebellar neurogenesis and histogenesis.

Omithine decarboxylase (ODC) catalyses the initial reaction in the biosynthesis of putrescine and the polyamines spermidine and spermine,*‘s” which are postulated to play an essential role in tissue growth and development. 9.21.33 Numerous studies conducted in a wide variety of organisms have shown that ODC activity and the polyamine content are highest in rapidly proliferating tissues and decline as growth processes cease. 9*2i.33~40~41 Thus each tissue has a specific ODC/polyamine developmental profile and any perturbation of this pattern is usually accom- panied by deficits in tissue maturation and/or func- tion.32.37 Nowhere is this relationship more evident than in the developing mammalian brain. In the fetal and neonatal rat the regional distribution of ODC activity correlates with the timing of cellular repli- cation, ODC activity peaks earliest in brain stem regions, later in cerebral cortex and last in the cerebellum.7.8 In the developing cerebellum enzyme levels are highest in proliferating cell populations.‘7*‘8

Until recently there has been no definitive proof that ODC activity is mechanistically linked to the ontogenic development of the brain. However, the availability of DL-a-ditIuoromethylomithine (DFMO), a specific irreversible ODC inhibitor,‘3,24 has now made it possible to evaluate the role of ODC activity in each developmental stage. Daily adminis-

tration of DFMO to neonatal rats rapidly and persis- tently inhibits the ODC activity of brain and

Please address all correspondence to: Dr. J. V. Bartolome, Department of Pharmacology, Duke University Medical Center. P.O. Box 3813, Durham, NC 27710, U.S.A.

AbbretGarions: DFMO, DL-zdifluoromethylomithine; ODC. omithine decarboxylase.

markedly decreases putrescine and spermidine lev- els.“*’ These alterations are accompanied by deficits in cellular replication, as evidenced by reductions in the synthesis and content of brain DNA.39 Indeed, of all tissues in the developing rat that have been studied, the central nervous system appears to be among the most sensitive to the antimitotic effect of DFM0.3uz In addition this agent affects post- proliferative aspects of development; for example, DFMO suppresses the formation of cate- cholaminergic synapses, even when treatment is he- gun after catecholamine neurons have ceased di- viding.‘* Inhibition of synaptogenesis is associated with reductions in the synthesis of RNA, protein and membrane phospholipids, which can exceed the re- duction in DNA synthesis. 38.39 Studies of other tissues also suggest that polyamines are necessary for growth processes which do not involve cell division.“.”

The question remains as to the identity of the cell populations in developing brain which are affected by DFMO treatment. The present study examines the influence of repeated DFMO administration to neo- natal rats on cytogenesis and histogenesis in the cerebellar cortex, a region that develops predom- inantly after birth. “*6*‘6.‘9 Major events in the post- natal differentiation of cerebellar cortex include neuronal proliferation, migration and growth, as well as the development of its characteristically laminated architecture. This region thus provides a convenient model for examining the effects of DFMO on both neuronal replication and differentiation. For com- parison we also studied the effects of DFMO on another brain region that develops predominantly after birth, the fascia dentata of the hippocampal formation.‘2*35

203

304 J. V. Bartolome et al.

Fig. 1. Drawing of a section through the cerebellum and brainstem of the adult rat, which shows the rostrocaudal level at which the subsequent photomicrographs were taken. Crus I (CR I) of the cerebellar cortex is enclosed by the box. CN, cochlear nucleus; DN, dentate nucleus; LL, lateral lemniscus; LS, lobuhrs simplex; NI, nucleus interpositus; PF, paraflocculus; SO, superior olive; TST, spinal tract of the trigeminal nerve; V, fourth ventricle; VL, vermian lobe; VN, vestibular nucleus; (*), corresponds to same position as asterisk in Fig. 7. The dashed lines represent the deep border of the internal granule cell layer.

EXPERIMENTAL PROCEDURES

a-Dipuo*omethylornithine treatment Timed-pregnant Sprague-Dawley rats (Zivic-Miller

Labs., Allison Park, PA) arrived 7-10 days before term. They were individually housed with a 12-h light-dark cycle and allowed food and water ad libitum. Pups from all litters were randomized at birth and redistributed to the nursing mothers. Litter sizes were kept at 10 or 11 pups to maintain a uniform nutritive status. Randomization within each treatment group was again carried out at intervals of several days and, for each experiment, pups of both sexes were selected from several different litters. Neonates were given DFMO (5OOmg/kg, s.c.) daily for as long as 21 days beginning the day after birth. Control animals received equal volumes of saline. This regimen has been shown previously to inhibit the ODC activity of neonatal brain by more than 900/,, leading to depletion of putrescine and spermidine.“’ For histological analysis, brains from six control and six DFMO-treated rats were examined at 5, 10, 15, 22 and 120 days of age. At the first four ages, studies were performed 24 h after the last injection.

Nissl preparation

Rats were anesthetized with pentobarbital and perfused intracardially with loo/!, fonnalin in 0.1 M sodium phos- phate buffer, pH 7.3-7.4. The brains were removed and postfixed for 2 days in buffered formalin which contained 300/, (w/v) sucrose. Frozen sections were cut through the entire cerebellum and hippocampus in the coronal plane at a thickness of either 20 or 35 pm stained with cresyl- echtvioiet or cresyl violet acetate. Control and DFMO- treated rats were processed simultaneously.

Qrumtitarive histological methodr

All quantitative analyses of cerebellar histology were performed on CNS I of the cerebellar hemispheres at the rostocaudal level shown in Fig. 1.

For each layer of the cerebellar cortex, cell counts were obtained at a magnification of 100-fold in a 2OO~rn~ area. Only neurons with nucleoli visible in the plane of section were counted. In addition, the circumference of each Pur- kinje cell soma was drawn at a magnification of 400-fold with use of a microscope equipped with a drawing tube. Only those cells with visible nucleoli were considered and an equal number of Purkinje cells were drawn from each case at a given age. The cross-sectional areas of these somata were then determined with a computerized planimeter.

Widths of the suprapyramidal and infrapyramidal leaves of the hippocampal granule cell layer were measured at a magnification of 415fold. For this purpose, two sections cut from between levels 1 and 2 of Nadler et al.‘5 were selected and measurements were made on each side of the brain normal to the cell layer halfway between the end and apex of the granule cell arch. The four measurements made on each leaf were averaged to obtain a single value for each leaf in each animal.

Two-way analyses of variance (treatment x age) were employed to assess differences in Purkinje cell body size and in cell density. Groups were then compared with either Newman-Keuls or simple effects post hoc analyses. The criterion for a statistically significant effect was P c 0.05.

RESULTS

The anatomy’5.26.3’ and normal develop- ment’s*‘6.20*22 of the cerebellum have heen reviewed in detail. For the purpose of this report only those aspects which were altered by DFMO treatment will be described. The effects of DFMO on the devel- opment of the cerehellar cortex were uniform

Fig. 2. Coronal sections through crus I of the cerebeilar cortex of S-day-old rats. (A) Control. Four layers can be identified. The extremely dense external granule cell layer (egcl) is located beneath the piat surface. The molecular layer (mi) is not so free of cell bodies as in the adult, due to the inward migration of granule ceils to the internal granule ceil layer (igcl). The deep white matter (wm) of the cerebellar cortex is also identifiable. (B) DFMO-treated. Rata were injected with DFMO on postnatal days 14 and were perfused on day 5. No differences from control arc evident. Crcsyl violet stain; scale, 200 /tm; 3Sqm thick sections.

205

Fig. 4. Coronal sections through crus I of the cerebellar cortex of lo-day-old rats. (A) Control. The four layers identified in the 5-day-old can also be distinguished at this age. The major developmental changes are the widening and clearing of the molecular layer, the increased density of the internal granule cell layer and the appearance of a Purkinje cell layer (pcl; borders are marked by the arrowheads). Other abbreviations as in Fig. 2. (B) DFMO-treated. Rats were injected with DFMO on postnatal days l-9 and were perfused on day IO. Many granule cells occupy the molecular layer. These granule cells are intermixed with the abnormally distributed Purkinje cells (between the pans of arrowheads). Cresyl violet stain; scale,

2OO~m; 35pm thick sections.

Fig. 5. Coronal sections through crus I of the cerebellar cortex of IS-day-old rats. (A) Control. The thickness of the external granule cell layer decreases between IO and I5 days of age, whereas the molecular layer broadens. The arrowheads mark the extent of the Purkinje cell layer. Abbreviations as in Figs 2 and 4. (B) DFMO-treated. Rats were injected with DFMO on postnatal days I-14 and were perfused on day 15. The appearance of the cerebellar cortex hardly differs from that in Fig. 4(B), but here the cell density of the internal granule cell layer is clearly less than control. Cresyl vrolet stam: scale. 2(X)/~m;

15-pm thick sections.

Fig. 6. Coronal sections through crus I of the cerebellar cortex of 22day-old rats. (A) C&trol. At this age the external granule cell layer is no longer present. Otherwise the cytoarchitecture closely resembles that of H-day-old rats. Arrows mark the extent of the Purkinje cell layer. Abbreviations as in Figs 2 and 4. (B) DFMO-treated. Rats were injected with DFMO on postnatal days l-21 and were perfused on postnatal day 22. In contrast to controls, the cerebcllar cortex still possesses an external granule cc4 layer. In fact, the cerebellar cortex has progressed little beyond the stage of development reached at IO (Fig. 4B) or 15 (Fig. SB) days of age. Unlike controls, a large proportion of the granule cells are located in

the molecular layer (star). Cresyl violet stain; scale, 2OOpm; 35-gm thick sections.

Fig. 7. Coronal sections through crus I of the cerebellar cortex of adult rats. (A) Control (same section used for Fig. I). The asterisk marks the same position as that in Fig. 1. Abbreviations as in Figs 2 and 4. (B) DFMO-treated. Tbis rat was injected with DFMO on postnatal days l-21 and was allowed to survive to adulthood (I 20 days of age). The layers of cerebellar cortex arc quite disorganized compared to those of controls. Many granule cells failed to migrate into the internal granule cell layer, remaining trapped in the molecular layer (star). These granule cells are intermixed with Purkinje cells that also failed to form a discrete layer. Arrows in (A) and (B) demarcate the broadest extent of the Purkinje cell layer.

Cresyl violet stain; scale, 2OOpm; 35-pm thick sections.

207

‘SU

O!)33S

pq1

turf-()z :rur1(joz ‘apm

:u!els

la[o!A Ika

J3 ‘3~

~3

Bale JO

sip

~ep+ueAd

iedumodd!q

aql sasolx~a .ral]el aql ‘sya3 qdJou+

od JO JaL

e[ ‘Id :~a&[ Jelnxlotu

‘fm :~

aA

e[

1la3 a

jnu

w8

‘1~~

8 ‘uo$ia

i s!q

l II! sa

yyxmo

uq

r! p

z]ua

wd

ola

,tap

sn

o!,tq

o

ou

pa

mp

old

0K

yja

‘9 %LJ

u! p

xgm

ap

w

p

a)sa

J) aJa

m s~

mu

!uv

‘pa

lmJl-o

Wd

a

(a)

‘10~1~

03 (v)

3~3

plo

-hp

-zz Jo

sn

A%

a

mu

ap

a

ql

q8n

oJq

l su

o!lsa

s p

zuo

~o

3 .11

%.J

suo

yas

23!ql rud-~c !S’L

‘uoge3yy&xu

tu!els mio!A

@aI

‘palEa

-oMda

(a) .loquo3

(v) x!sr(pxm

JOJ uasoq3

IaA

al ~

up

mm

tso3

aq

l w

slw

p

lo-d

ep

-0~ 1 JO luals u!srq pue m

nfjaqam

aqt qSnonf3 suogms

pmoJo3

‘6 $5

Effects of DFMO on development of cerebellar cortex 209

Y d !m iJ 4oR

& 300

g 200 m J 100

= 0

EXTERNAL GRANULE CELL LAYER MOLECULAR LAYER

5 IO 15 22 AM1

500

400

300

200

100

0

INTERNAL GRANULE CELL LAYER

5 10 15 22 ADULT

AGE KIAYS> Fig. 3. Cell density in crus I of cerebellar cortex from rats of various ages. Each neuron with a visible nucleolus was counted in a 200 pm2 cross-sectional area of each layer. Solid hoe, control rats; dotted line, DFMO-treated rats. Values are expressed as means *SEM. See Experimental Procedures for other details. *DFMO-treated rats differed from controls at P < 0.05 [simple effects test after two-way analysis

of variance (P < O.Ol)].

throughout this region and thus, for simplicity, only the development of crus I is presented.

Effects of a-dtfiuoromethylorttithine on the devel- opment of cerebellar granule cells

On postnatal day 5, the cerebellar cortex of rats treated with DFMO since the day after birth ap- peared quite normal (Fig. 2). All the layers could be identified and the cell density of each layer did not differ significantly from that of littermate controls of the same age (Fig. 3). Differences between control and DFMO-treated rats became apparent at 10-15 days of age (Figs 4 and 5). Most strikingly, the normal increase in cell density of the internal granule cell layer and decrease in cell density of the molecular layer failed to occur (Fig. 3). However, the external granule cell layer maintained its normal thickness and cell density. The contrast between cerebellar cortices of control and DFMO-treated rats was further accen- tuated at day 22 by the continued differentiation of this brain region in controls and the lack of further development in treated animals. An external granule

-l < 100

E 75

&

!z 50

’ 25

E

DFMO

5 10 15 22 ADULT

AGE (DAYS> Fig. 8. Mean percentage of total cerebellar cortical neurons (excluding Purkinje cells) found in the internal granule cell

layer at various ages.

cell layer was still apparent in DFMO-treated rats (Figs 3 and 6B), whereas in control animals it had disappeared by this time (Figs 3 and 6A).

The deleterious effects of DFMO on the density and distribution of cerebellar granule cells were irre- versible. The only obvious change between days 22 and 120 was the disappearance of the external gran- ule cell layer (Fig. 7). The internal granule cell layer remained relatively thin and never attained the cell density of controls (Fig. 3). Less than 60% of adult cerebellar cortical neurons (excluding Purlcinje cells) were found in this layer compared to about 90% in adult controls (Fig. 8). Approximately one-fourth of the granule cells remained ectopically localized in the molecular layer.

a-Difluoromethylomithine treatment markedly re- duced the overall size of the cerebellum and of individual folia (Fig. 9).

Eflects of a-d@uoromethylornithine on the devel- opment of Purkinje cells

The density of Purkinje cells (cells per 200pm2 cross-sectional area) in crus I progressively declined with age in both control and DFMO-treated rats (Table 1). This observation presumably reflects the increasing dispersion of a constant number of Pur- kinje cells due to growth of the neuropil. The Pur-

Table 1. Density of Purkinje cells in CNS I of cerebellar cortex

Age 5 10 15 22 120

Control 16il 11*1 1051 8kl 6fO DFMO-treated 18&2 14*1 13+1 ISkZ* 8*l

Rats were injected daily with DFMO or saline from the day after birth until 21 days of age or until the day before sacrifice. Units are Purkinje cells per 200 pm2 of Purkinje cell layer. Values are means f SEM. See Experimental Procedures for other details.

‘DFMO-treated rats differed from controls at P < 0.05 [simple effects test after two-way analysis of variance (P < O.Ol)].

210 J. V. Bartolome er al.

AGE UlAYS>

Fig. 10. ES&s of DFMO on the postnatal increase in cross-sectional area of Purkinje cell somata. See Experimental Procedures for details. Solid line, control rats; dotted tine, DFMO-treated rats. Values are

means fSEM for 41-59 celh.

kinje cell density tended to be greater in DFMO- treated rats from 10 days of age onward, although only at 22 days was this difference statistically significant. In contrast, the laminar distribution of Purkinje cells was markedly altered by postnatal treatment with DFMO. In both 5-day-old control and DFMO-beam rats, the Purkinje cells were scattered throughout the molecular layer (Fig. 2). By postnatal day 10 the Purkinje cells of control rats were seen to be forming a discrete layer l-3 ceils thick (Figs 4A-7A). In DFMO-treated rats, however, the Purkinje cells remained scattered in a broad band five or more cells thick which encompassed much of the molecular layer (Figs 4B-7B).

The cross-sectional area of Purkinje cell somata in controls increased approximately 4-fold between 5 and 22 days of age (Fig. 10). a-Difluoro- methylonithine treatment inhibited the growth of Purkinje cell somata during this period (F, 229.99; d.f., 1,501; P < 0.01).

Effects of a -dt~~oromethy!or~ithine on the develop - ment of the fascia dentata

Like the cerebellar granule cells granule cells of the fascia dentata also develop mainly after birth. About 80-U% of these neurons enter their final division between birth and postnatal day 21.‘2s35 Nevertheless, in contrast to its effects on develop~nt of the cerebellar cortex, daily administration of DFMO during the first three weeks after birth failed to alter the lamination of the fascia dentata (Fig. 11). Be- tween postnatal days 5 and 22 the width of the suprapyramidal granule cell layer remained virtually constant at 60-75pm, whereas the width of the

infrapyramidal granule cell layer increased from 35-40 to 80-lOO/lm. This difference was expected since most infrapyramidal granule cells enter their final division later than suprapyramidal granule cell~.‘~ The mean widths of both granule cell layers were qua1 in control and DFMO-treated rats at all

ages.

DlSCUSSION

Daily administration of DFMO during the first three postnatal weeks dramatically altered both the ultimate size and laminar architecture of the rat cerebellar cortex. The cerebellar cortex was much smaller than normal, the number of granule cells was irreversibly reduced and neither granule cells nor Purkinje cells formed a well-defined layer. In general the cerebellar cortex failed to mature much beyond the stage of development reached in control rats during the first postnatal week. These findings can be explained by effects of DFMO on both neuronal replication and postmitotic events.

In the rat all cerebellar cortical neurons, except for Purkinje and Golgi cells, enter their final mitosis after birth.1-4~6*‘*~i9 The great majority of these postnatally generated neurons are granule cells. Although the density of granule cells (per unit cross-sectional area) was unaffected by DFMO trea~ent, the cerebellar cortex as a whole grew to only a fraction of its normal size. Thus the total number of these neurons was correspondingly reduced. This finding implies that DFMO treatment either inhibited the postnatal gen- eration of cerebellar granule cells or increased the incidence of neuronal cell death. Since DFMO in-

Effects of DFMO on development of cerebellar cortex 211

hibits DNA synthesis in the CNS without afftitiiig the survival of postmitotic cells,” the deficit in the total number of granule cells most likely reflects reduced cell division, rather than increased cell death. Thus replication of cerebellar granule cells appears to require ongoing synthesis of polyamines.

In addition to reducing the number of cerebellar granule cells, DFMO inhibited their migration: many of the surviving granule cells failed to reach the internal granule cell layer and were found ectopically located in the molecular layer. This finding is consis- tent with substantial evidence that some post- proliferative events require ODC and the polyamines. Although it is tempting to conclude that polyamines participate in cell migration, granule cell ectopia can also be produced by mutations and antimitotic agents that damage the immature cells of the external gran- ule cell layer.s,23.30*36 Thus the block of granule cell migration by DFMO may not necessarily imply a direct role for polyamines per se in this process. Alternatively, DFMO administration may have pro- duced granule cell ectopia by reducing the number or altering the orientation of Bergmann glia. Rakic has suggested that processes of Bergmann glia guide the migration of granule cells2*.29 and glial pathology has been postulated to account for the presence of gran- ule cells in the molecular layer of weuuer mouse cerebellar cortex.M We could not determine whether DFMO affected the Bergmann glia from the present material.

Adverse effects of DFMO were not restricted to the granule cells, as the lamination and enlargement of Purkinje cells were also disrupted. These results could have arisen from postmitotic actions of DFMO, since polyamine depletion by the drug depresses RNA and protein synthesis in brain.39 The resumption of nor- mal Purkinje cell. growth after the treatment was stopped (Fig. 10) tends to support this hypothesis, but it must be noted that the development of Purkinje

cells also depends on their local milieu. The massive reduction in the number of parallel fibers could thus contribute to the lack of a well-defined Purkinje cell layer as seen in the reeler mouse23 or in rats exposed postnatally to antimitotic treatments which reduce the number of granule cell~.~~~ Unlike DFMO- treated rats, however, the other conditions do not display a reduction in the size of Purkinje cell somau.

He&co at least some direct role of polyamines in Purkinje cell development seems likely.

In contrast to its marked effects on the location and size of Purkinje cells, DFMO administration had little effect on the total number of these neurons in the cerebellar cortex. Indeed, Purkinje cell density actually tended to be greater in DFMO-treated rats as a consequence of the smaller volume of cerebellar cortex. This result again supports the view that DFMO does not destroy postmitotic neurons.

The dendritic branching and synaptic re- lationships of Purkinje (and granule) cells could not be determined from the present material and will be the subject of subsequent investigations. In view of the results with other antimitotic agents, it is likely that these characteristics also were altered by DFMO.

To determine whether postnatal administration of DFMO affected other mitotic neurons, we also stud- ied the development of the fascia dentata of the hippocampal formation. These neurons, like cere- bellar granule cells, are generated predominantly after birth and migrate to their final position.‘2.35 Despite the ontogenetic similarity of dentate and cerebellar granule cells, we observed no effect of DFMO treatment on the architecture of the fascia dentata. If DFMO had affected these neurons in the same manner as cerebellar granule cells, one would expect it to have inhibited the formation of at least the late-developing infrapyramidal granular layer. No such inhibition could be measured, suggesting that either dentate and cerebellar granule cells differ in the degree of dependence upon polyamines for cell replication and differentiation, or alternatively that DFMO less effectively depletes polyamines in fascia dentata than in cerebellar cortex. In either case, our data clearly indicate that not all immature neurons are as sensitive to DFMO as are cerebellar granule cells. This agent may therefore be employed as a relatively selective toxin in studies of brain devel- opment.

Acknowledgemenfs-We thank Ms. D. Evenson, Ms. D. Pistorino, Ms. A. Robbins and Mr. A. Hakim for technical help and Ms. M. Triplett and Mrs. C. Dickens for secretarial assistance. This work was supported by U.S. Public Health Service grants HD-09713 and NS-06233. a-Difluom-

methylomithine (MDL 71, 782 A) was kindly provided by Dr. P. McCann, Merrell Dow Research Institute (Cincinnati. OH).

REFERENCES

Altman J. (1966) Autoradiographic and histological studies of postnatal neurogenesis. II. A longitudinal investigation of the kinetics, migration and transformation of cells incorporating tritiated thymidine in infant rats, with special reference to postnatal neurogenesis in some brain regions. J. camp. Neural. 128, 431474. Altman J. (1972) Postnatal development of the cerebellar cortex in the rat. 1. The external germinal layer and the transitional molecular layer. J. camp. Neural. 145. 353-398. Altman J. (1972) Postnatal develop&ent of the ceiebellar cortex in the rat. II. Phases in the maturation of Purkinje -cells and the molecular layer. J. camp. Neural. 145, 399-464.

4. Altman J. (1972) Postnatal development of the cerebellar cortex in the rat. III. Maturation of the components of the granular layer. J. camp. Neural. 145, 465-514.

5. Altman J. and Anderson W. J. (1969) Early effects of X-irradiation of the cerebellum in infants rats: decimation and reconstitution of the external granular layer. Expl Neural. 24, 196216.

212 J. V. Bartolome er nl.

6.

7.

8.

9. 10.

II.

12.

13.

14.

15. 16.

17.

18.

19.

20. 21. 22. 23.

24.

25.

26. 27.

28.

29. 30

31

32.

33.

34.

35.

36

37 38

39.

40.

41.

Altman J. and Das G. D. (1966) Autoradiographic and histological studres of postnatal neurogenesis. I. A longitudinal investigation of the kinetics, migration and transformation of cells incorporating tritiated thymidine in neonatal rats. with special reference to postnatal neurogenesis in some brain regions. J. cotnp. Neural. 126, 337-390. Anderson T. R. and Schanberg S. M. (1972) Omithine decarboxylase activity in developing rat brain. J. Neurochem. 19, 1471-1481. Anderson T. R. and Schanberg S. M. (1975) Effect of thyroxine and cortisol on brain omithine decarboxylase activity and swimming behavior in developing rat. Eiochem. Pharmac. 24, 495-501. Bachrach U. (1973) Function of Naturally Occurring Pol_vamines. Academic Press, New York. Bartolome J. V.. Huguenard J. and Slotkin T. A. (1980) Role of omithine decarboxylase in cardiac growth and hypertrophy. Science. N. Y. 210, 793-794. Bartolome J. V., Trepanier P. A., Chait E. A. and Slotkin T. A. (1982) Role of polyamines in isoproterenol-induced cardiac hypertrophy: effects of z-difluoromethylomithine, an irreversible inhibitor of ornithine decarboxylase. J. molec. cell. Cardiol. 14, 461-466. Bayer S. A. and Altman J. (1974) Hippocampal development in the rat: cytogenesis and morphogenesis examined with autoradiography and low-level X-irradiation. J. camp. Neurol. 158, 55-80. Bey P. ( 1978) Substrate-induced irreversible inhibition of alpha-aminoacid decarboxylases. Application to glutamate, aromatic-L-alpha-aminoacid and omithine decarboxylases. In Enzyme Activated Irreversible Inhibitors (eds Seiler N.. Jung M. J. and Koch-Weser J.), pp. 27-41. Elsevier/North Holland, New York. Caviness V. S. Jr. and Sidman R. L. (1973) Time of origin of corresponding cell classes in the cerebral cortex of normal and reeler mutant mice: An autoradiographic analysis. J. camp. Neurol. 148, 141-152. Eccles J. C., Ito M. and Szentagothai J. (1967) The Cerebellum as a Neuronal Machine, Springer, New York. Fujita S. (1969) Autoradiographic studies on histogenesis of the cerebellar cortex. In Neurobiology of Cerebellur Evolution and Development (ed. Llinas R.), pp. 743-801. A.M.A., Chicago. Gilad G. M. and Gilad V. H. (1981) Cytochemical localization of omithine decarboxylase with rhodamine or biotin-labeled a-difluoromethylomithine. J. Histochem. Cyrochem. 29, 687-692. Gilad G. M. and Kopin I. J. (1979) Neurochemical aspects of neuronal ontogenesis in the developing rat cerebellum: Changes in neurotransmitter and polyamine synthesizing enzymes. J. Neurochem. 33, 1195-1204. Heinsen H. (1977) Quantitative anatomical studies on the postnatal development of the cerebellum of the albino rat. Anat. Embryo/. 151, 201-218. Jacobson M. (1978) Developmental Neurobiology. Plenum Press, New York. Janne J., Poso H. and Raina A. (1978) Polyamines in rapid growth and cancer. Biochem. biophys. Acta 473,241-293. Lund R. D. (1978) Development and Plasticity of the Bruin. Oxford University Press, New York. Meier G. and Hoag W. G. (1962) The neuropathology of “reeler” a neuromuscular mutation in mice. J. Neuropurh. exp. Neurol. 21, 649-654. Metcalf B. W., Bey P., Danzin C., Jung M. J., Casara P. and Vevert J. P. (1978) Catalytic irreversible inhibition of mammalian omithine decarboxylase (E.C.4.1.1.17) by substrate and products analogues. J. Am. them. Sot. 100, 2551-2553. Nadler J. V., Cotman C. W. and Lynch G. S. (1977) Histochemical evidence of altered development of cholinergic fibers in the rat dentate gyrus following lesions. I. Time course after complete unilateral entorhinal lesion at various ages. J. camp. Neural. 171, 561-588. Palay S. L. and Chan-Palay V. (1974) Cerebellur Cortex. Springer, Berlin. Pegg A. E. and Williams-Ashman H. G. (1968) Biosynthesis of putrescine in the prostate gland of the rat. Biochem. J. 100, 533-539. Rakic P. (1971) Neuron-glia relationship during granule cell migration in developing cerebellar cortex. A Golgi and electron microscopic study in Mucocus rhesus. J. camp. Neural. 141, 283-312. Rakic P. (1975) Cell migration and neuronal ectopias in the brain. Birth Defects 11, 95-129. Rakic P. and Sidman R. L. (1973) Organization of cerebellar cortex secondary to deficit of granule cells in weaver mutant mice. J. camp. Neural. 152, 133-162. Ramon y Cajal S. (191 I) Histologie du Systeme Neweux de I’Homme et des Vertebres, Vol. II (translated by L. Azoulay, 1955). Consejo Superior de lnvestigaciones Cientificas, Madrid. Russell D. H. (1980) Omithine decarboxylase as a biological and pharmacological tool. Pharmacology 2% 117-129. Russell D. H. and Durie B. G. M. (1978) Polyamines as Biochemical Markers @Normal and Malignant Growth: Progress in Cancer Research and Therapy, Vol. 8. Raven Press, New York. Russell D. H. and Snyder S. H. (1968) Amine synthesis in rapidly growing tissues: omithine decarboxylase activity in regenerating rat liver, chick embryo and various tissues. Proc. narn. Acad. Sci. U.S.A. 60, 1420-1427. Schlesinger A. R., Cowan W. M. and Gottlieb D. I. (1975) An autoradiographic study of the time of origin and the pattern of granule cell migration in the dentate gurus of the rat. J. camp. Neurol. 15% 149-176. Shim& M, and Lanaman J. (1970) Repair of the external granular layer after postnatal treatment with

5-fluorodeoxyuridine. Am. J. Anat. 129, 247-260. Slotkin T. A. (1979) Omithine decarboxylase as a tool in developmental neurobiology. Lrye Sci. 24, 1623-1630. Slotkin T. A.. Grignolo A., Whitmore W. L., Lerea L., Trepanier P. A., Barnes G. A., Weigel S. J., Seidler F. J. and Bartolome J. V. (1982) Impaired development of central and peripheral catecholamine neurotransmttter systems m preweanling rats treated with a-difluoromethylomithine, a specific irreversible inhibitor of omithine decarboxylase. J. Pharmac. exp. Ther. 222, 746-751. Slotkin T. A.. Persons D., Slepetis R. J., Taylor D. and Bartolome J. V. (1985) Control of nucleic acid and protein synthesis in developing tissues of the neonatal rat: effects of a-difluoromethylomithine a specific irreversible inhibitor of ornithine decarboxylase. Teratology, 30, 21 l-224. Slotkin T. A., Seidler F. J., Trepanier P. A., Whitmore W. L., Lerea L., Barnes G. A., Weigel S. J. and Bartolome J. V. (1982) Omithine decarboxylase and polyamines in tissues of the neonatal rat: Effects of ordifluoromethylomithine, a specific, irreversible inhibitor of omithine decarboxylase. J. Pharmac. exp. Ther. 222, 741-745. Slotkin T. A., Seidler F. J., Whitmore W. L., Weigel S. J., Slepetis R. J., Lerea L., Trepanier P. A. and Bartolome J. V. (1983) Critical periods for the role of omithine decarboxylase and the polyamines in growth and development

Effects of DFMO on development of cerebellar cortex 213

of the rat: effects of exposure to a-difluoromethylomithine during discrete prenatal or postnatal intervals. hr. J. devl Neurosci. 2, 113-127.

42. Slotkin T. A., Whitmore W. L., Lerea L., Slepetis R. J., Weigel S. J.. Trepanier P. A. and Seidler F. J. (1983) Role of omithine decarboxylase and the polyamines in nervous system development: short-term postnatal administration of adifluoromethylomithine, an irreversible inhibitor of omithine decarboxylase. Int. J. devl Neurosci. 1, 7-16.

(Accepted 15 Nooember 1984)