The development of the corticotectal pathway in the albino rat

DevelopmentalBrain Research, 25 (1986) 227-238

227

Elsevier BRD 50354

The Development of the Corticotectal Pathway in the Albino Rat IAN G. THONG and BOGDAN DREHER

Departmentof Anatomy, Universityof Sydney, N.S. W. 2006 (Australia) (Accepted August 28th, 1985)

Key words: corticotectal projection - - development - - rat

To study the development of the corticotectal pathway, the enzyme horseradish peroxidase (HRP) was injected electrophoretically into the superior colliculus (SC) of rats ranging in age from newborn to adult. In animals younger than postnatal day 3 (P3). collicular injections did not label any cells in the cortex while in animals injected at P3-P4, only a few cortical cells were retrogradely labeled. In contrast, injections made at P5 or later resulted in the labeling of a substantial proportion of lamina V cells in a number of cortical areas ipsilateral to the injected colliculus. Although at P5-P7 the bulk of labeled cells was located in the visual cortices (both striate and extrastriate), a substantial proportion of the labeled cells was located in the somatosensory, motor and association cortices. On the other hand, in animals injected at P12 (or later), the labeled cells were largely restricted to the visual cortices with relatively few corticotectal cells located in somatosensory area I. At all ages studied, labeled cortical cells were confined to lamina V and had clear-cut apical dendrites (pyramidal cells). The dendritic morphologies and somal sizes of the corticotectal cells indicate that in animals younger than P12 these cells are immature. These observations suggest that the axons of cortical cells do not reach the SC before P3 and that these early corticotectal projections (P3-P12) are established by immature cells. Furthermore, although the corticotectal projection exhibits, from its onset, a high degree of specificity in terms of the laminar distribution of its cells of origin, its areal distribution is 'exuberant'. The 'exuberant' projections originating from non-visual cortical areas disappear by P12-P14, that is at the time when young rats open their eyes for the fist time.

INTRODUCTION

ciated with a naturally occurring wave of ganglion cell deathS.43.

The visual system of nocturnal rodents is very immature at the time of birth. F u r t h e r m o r e , since nocturnal rodents have a short gestation period and are easily available, they offer an excellent opportunity to study the spatial and temporal patterns of devel-

In the mature rat, a single retinotopic locus in the major retinorecipient nucleus, the superior colliculus (for recent review see ref. 17), receives separate projections, precisely aligned in a point-to-point relationship, from at least seven retinotopically orga-

opment of interconnections within the m a m m a l i a n visual system. It has been clearly established that at least some projections present within the visual system of the newborn nocturnal rodents are ' e x u b e r a n t ' and not very precisely organized. For example, in the newborn rat, the retinofugal projections are clearly more excessively distributed and more n u m e r o u s than those of the adult rat ~t,5,28-30,43,44,49. The precise topo-

nized areas located within the visual cortex 41. Thus, the study of the timecourse of development and the precision of organization of the early corticotectal pathway might provide important insights concerning developmental interrelationships between principal afferent inputs to the retinorecipient parts of the mammalian tectum. To study the time-course of development of the corticotectal system of the rat, we injected the enzyme horseradish peroxidase ( H R P ) into superior colliculi (SC) of rats of different ages and Studied the changes in soma size, cell morphology as well as laminar and areal pattern of distribution of the retro-

graphic pattern of organization characterizing the retinotectal or retinogeniculate pathway of the adult rat and hamster emerges only gradually during the first few postnatal days 19,28.30,37,53,54 and is asso-

Correspondence: I.G. Thong, Department of Anatomy, University of Sydney, N.S.W. 2006, Australia. 0165-3806/86/$03.50© 1986 Elsevier Science Publishers B.V. (Biomedical Division)

228 gradely labeled corticotectal cells. A preliminary communication of some of our results has been already published 57. MATERIALS AND METHODS Thirty Sprague-Dawley albino rats ranging in age from postnatal day i (P1) to adulthood were used (25 of the rats were in the age range 1-21 days, the remaining 5 were adults, 60 days or older).

Surgical procedures The young animals (P1-P21) were anesthetized with 1% halothane in 65%/35% N20/O 2 mixture. They were then placed in a plaster mould tailored to the size of the animal 33 and immobilized with dental impression material (Palginex). To prevent hypothermia during surgery, the animals were placed on a heated table maintained at approximately 37 °C. In the younger animals (1-10 postnatal days) the cortex had not as yet grown over the SC; however, in animals older than P10, the continued caudal growth of the cortex prevented direct access to the SC. Thus, in the animals older than P10, the position of the colliculus was determined stereotaxicallyS2. The adult rats (weighing 250-350 g) were lightly anesthetized with 5% halothane in a mixture of 65%/35% N20/02, then injected i.p. with paraldehyde (0.1 ml/100 g b.wt.). The animals were then mounted in a stereotaxic frame and the position of the colliculus was determined with the help of a stereotaxic atlas 42. In both the younger and older animals, prior to injecting the HRP, the dura overlying the exposed SC or cortex was removed.

HRP injections Micropipettes containing either a 20% solution of HRP (Grade 1, lyophilized, Boehringer Mannheim) in 0.2 M Tris-HC1 buffer (pH 8.6) and 5% dimethyl sulphoxide (DMSO) 25 or a 2.5% solution of H R P conjugated with lectin wheatgerm agglutinin (WGAHRP) in 0.2 M Tris-HC1 buffer (pH 8.6) were prepared. The internal diameters of the tips of the microelectrodes were 20-40 #m and their impedances were in the range of 2 - 6 MfrS. The injections were made electrophoretically (electrode positive, current 1-2 #A, square wave pulses 1 s ON/1 s OFF) over a period of 15-30 min 11.

In animals younger than P10, the injections were made at a depth of 500urn from the surface of the SC. This placed the dense core of the injection site within the superficial (retinorecipient) layers of the colliculus. Since, as mentioned earlier, in animals older than P10 the cortex at least partially covers the colliculi, to minimize the spread of H R P into the non-collicular regions of the brain, the electrode was passed between the cortex and the transverse sinus. In adult animals, in order to minimize the spread of HRP to the cortex overlying the colliculus, a negative current was passed through the electrode before and during the advancement or withdrawal of the electrode through the cortex. Twenty-four hours after an H R P injection animals were reanesthetized by an intraperitoneal injection of sodium pentobarbitone (Sagatal 0.1 ml/100 g b.wt.) and perfused transcardially with a mixture of aldehydes. Animals younger than P18 were perfused with warm (37 °C) Hartmann's solution containing 5% sodium nitrite and heparin followed by fixative (0.5% paraformaldehyde, 1.25% glutaraldehyde) delivered over a period of 20 min and then finally by physiological saline delivered over a period of 20 rain. Adult rats were perfused with Hartmann's solution followed by fixative (1% paraformaldehyde, 2.5% glutaraldehyde) over 30 rain and then physiological saline. Following the perfusions, the brains were stored for 2-3 days at 4 °C in 30% sucrose 0.2 M phosphate buffer (pH 7.4) containing 1% DMSO and then sectioned parasagittally at either 30ktm (animals 10 days old or younger) or 50 ,um on a freezing microtome. Sections were collected at room temperature in 0.1 M phosphate buffer (pH 7.4) and reacted with the chromogen tetramethyi benzidine (TMB) according to the method of Mesulam 3s. Sections were counterstained with neutral red and mounted with mounting medium.

Methods of analysis Cells were considered to have been retrogradely labeled with H R P if: (a) granules of H R P reaction products were within the confines of clearly identifiable outlines of cell bodies; (b) the labeled cells were not associated with blood vessels. Criterion (b) eliminated the possibility of including endothelial and red blood cells which, due to their endogenous perox-

229 idase activity, might also contain precipitate similar to that of the H R P reaction products. For each age the distribution of corticotectal cells was determined by plotting the position of labeled cells in each parasagittal section and then redrawing their position on to a reconstructed dorsal view of the brain. The areal distribution of corticotectal cells for a given animal was derived by comparing the distribution of labeled cells on the reconstructed dorsal view of the brain with the available areal maps of the adult cortex. The mean somal diameter (average of largest and smallest diameter) of labeled cells located within the primary visual area (area 17) was determined by measurements under a microscope with a 100x oil immersion objective. Cells which were densely labeled with H R P reaction product were excluded from the measurements of somal sizes since crystals of reaction product frequently obscured the outline of the cell. Furthermore, only the perikaryal diameters of cells with a clearly distinguishable nucleus (and thus presumably cut through their maximal diameter) were measured. The density of the projection from the striate cortex to the SC was also determined. This was achieved by: (a) identifying the region of the striate cortex with the highest density of labeled corticotectal cells (that is, the region which presumably corresponds retinotopically to the injected collicular region); and (b) expressing the proportion of labeled cells in this high-density cluster as a percentage of the total number of presumed neurons present in the sampled region. Although, in the rat, like in other mammals studied so far, there is no increase in the number of cortical neurons during the postnatal period 3, the cortical laminae continue to increase in thickness for quite some time after birth. The cortical growth observed in rodents during the postnatal period seems to be due to the continuing growth of the neurons and neuropil as well as the proliferation of glia (cf. refs. 2, 13, 14, 31). To compensate for the continuing increase in thickness of lamina V, at all ages studied the sampled regions involved the entire thickness of lamina V. Thus at different ages, sampled regions were of a constant width but of different thickness. In order to exclude glial cells, we restricted the counts of unlabeled cells to those within the size

range of labeled cells. Since the mean somal sizes of glial cells in the adult rat32 are smaller than those of the smallest corticotectal cells labeled in the present study (except in the animals younger than P7) it appears that virtually all glial cells were excluded from our counts. On the other hand, this restriction might have excluded from our counts some of the smaller unlabeled neurons.

Controls Three types of control experiments were performed. (1) H R P injections were made into the subdural space around the colliculus. As expected, these injections did not label any cells in the cortex or visual subcortical nuclei. (2) The electrode containing H R P was placed into the SC but no current was applied. No labeled cells were observed in any part of the cortex. (3) H R P was injected into 'non-visual structures' (e.g. inferior colliculus); no labeled cells were found in the visual cortex although a substantial number of labeled cells were present in a number of other structures. RESULTS

Injection sites Fig. 1 illustrates injection sites of W G A - H R P , typical for younger and older subgroups of animals used in the present study. In animals younger than P7 (Fig. 1A), although the dense 'core' of precipitation of HRP reaction products is restricted to the superficial (retinorecipient) layers of the SC a more diffuse region of lighter precipitate - - the 'halo region' - spreads into the deeper collicular laminae. Similarly, in P16 (Fig. 1B) or older animals, while the 'core' of the injection site is confined to the superficial layers, the 'halo' spreads to the deeper layers.

Morphology of corticotectal cells In animals younger than P10 the retrograde labeling of corticotectal cells is substantially sparser (i.e. fewer granules of H R P reaction products in the labeled cells) than that seen in the older pups or adults (cf. inserts in Fig. 2 B - D ) . This 'light' labeling does not appear to be related to the amount of H R P injected as larger injections into the colliculi of young pups did not improve the quality of the labeling. Thus, our data indicate that 'light' labeling is a feature charac-

230

A

Fig. 1. Photomicrographs of parasagittal sections through the brains of (A) 7- and (B) 16-day-old rat pups in which WGA-HRP was injected electrophoretically into the SC. Note that in both cases the dense 'core' region of precipitate of HRP reaction products is restricted to the superficial collicular laminae. Note also that in the younger animal, the cortex has not yet grown over the SC. By contrast, in the 16-day animal (like in the adult rat) the cortex completely covers the SC. Scale bars in A and B, 1 ram.

teristic of immature neurons (see similar findings on rat's retinal ganglion cells34,43,44; cat's corticotectal cellsSS). In the P 4 - P 5 animals corticotectal cells are clearly immature. A t those ages the labeled cortical cells have small ovoid s o m a t a with long fine-gauge apical dendrites (Fig. 2B, insert) and a m e a n somal diameter of 8.7/~m (range 4.5-11.5/~m; Fig. 3). Only occasionally were axons observed extending from the base of the soma and no basal dendrites were observed at these ages. F r o m P5 onwards, the size and a p p e a r a n c e of the corticotectal cells gradually changes and by P12 they have more or less attained the somal d i a m e t e r characteristic of mature cells (mean soma d i a m e t e r 12.6 ~m, range 8.5-16.5/~m; Fig. 3). Over the same time period, the corticotectal cells gradually become m o r e pyramidal in shape, with the apical dendrites becoming thicker and thus m o r e p r o m i n e n t (Fig. 2C). Basal dendrites become a p p a r e n t for the first time in animals injected at P7. In still older animals basal dendrites gradually increase in number, caliber and extent. Although the mean somal size continues to increase till P21 (13.1 /~m, range 8.5-19.5 /~m) the mean size of labeled corticotectal cells in the adult (11.9/~m, range 8-17.5/~m) is slightly lower than that at P15. This small reduction in mean somal size could be related to the continuing d e v e l o p m e n t of the dendrites and apparent shift of cytoplasm from the somata to the dendrites.

Laminar distribution of corticotectal cells Consistent with earlier reports 41.50, injections into the SC of adult rats resulted in the labeling of many lamina V cells in the ipsilateral striate and extrastriate visual cortices. Although some labeled cells were located outside the visual cortex (mainly in parietal cortex area 7, somatosensory area I, see further), all labeled cells were confined to lamina V (Figs. 2D and 4) or lamina Vb in the case of somatosensory area I. Significantly, after H R P injections into the SC of early postnatal rats, labeled cortical cells were also restricted to lamina V of the ipsilateral cortex. This confinement of corticotectal cells to lamina V is already clearly apparent in P 3 - P 5 animals although at these ages laminae I I - I V have not yet differentiated from the cortical plate and lamina V contutes the upper part of the subplate layer (Fig. 2 A - C ; cf. refs. 21, 39, 45). Thus, it appears that at the time when the first corticotectal fibers reach the SC, the laminar distribution of corticotectal cells is already identical to that of the corticotectal cells in adults.

Time-course of development of tectal projection from striate cortex The youngest animals in which an injection of H R P into the superior colliculus labeled cortical cells were 3 days old (P3). H o w e v e r , in animals injected at P3, the labeled cortical cells were very few in n u m b e r and widely scattered throughout the ipsilateral cortex. H R P injections into the SC of P4 rats labeled

231

C

b

Fig. 2. Photomicrographs of parasagittal sections through the cortices of rats whose ipsilateral SC has been injected with HRP. Note in A that labeled cells are not present in the cortex of the 3-day-old rats. Note also that in the 3-day-old rat (P3), only laminae V and VI are distinguishable while laminae I I - I V comprise the cortical plate (CP). In 5-day-old (B), 15-day-old (C) and adult (D) animals, labeled cells in lamina V are clearly discernable. Note that in older animals the cortical laminae including lamina V are thicker. The inserts in B, C and D are high-magnification photomicrographs of the individual labeled cortical cells at the corresponding ages. Note the changes in the morphology of the corticotectal cells as the animal matures - - in particular the increase in somal size and appearance of basal dendrites. Scale bars in A - D , 100~m; bars in the accompanying inserts, 10~m.

232 SO,

50, 5

Days

15

40,

N : 189 x 868

30,

3O

2oJ

2O

10

10

O

6

50 7

.J .J

40

U LIL

3O

8

10

12

14

Days

40.

16

18

20

6

5O

Days

8

10

209 - .

12 ~_ 14

16

18

20

R

Jl

21 D a y s

N = 179 = 10,5

N : 5 =6

ii

40 3O

o 20 m

10

z 0 50

10 . . . . . . . . . . 6 8 10 12

~. . . . . . 14 16 18

20

......

O 50

11 Days

6 ADULT

40

~

1

N

" 217

40

:

203010

....... 8

10

12

- , , .~14

16

18

20

o.,,, ~

N ; 307

30

20

'

- = '1"I

10 0

6

8

10

12

14

16

18 SOMA

20

0

DIAMETER

( ~m )

Fig. 3. Frequency histograms showing the range of somal sizes and mean somal diameters (arrows) of labeled corticotectal cells in animals of different ages. Note that with increasing age both the range of somal sizes and mean somal diameters increase.

only slightly greater numbers of cortical cells. Within the striate cortex, as indicated in Fig. 5, the ratio of labeled cells to the total number of neurons in lamina V within any specific cluster increased rapidly between P4 and P9 and more slowly between P9 and P13. Thus while in animals injected at P5 the labeled cortical cells in the clusters with the highest density constituted about 20% of lamina V neurons, in animals injected at P14 (at the time of eye opening) a peak value of about 42.5% was reached• By P15-P16 the proportion of labeled cells within the densest clusters started to drop reaching a mature value of about 3 4 - 3 5 % by P21.

Areal distribution of labeled cortical cells Following localized injections of H R P into the SC of the P12 rats or older, labeled cells are distributed in a number of discrete patches in the ipsilateral striate and several extrastriate visual cortices. In ad-

dition, there is a smaller proportion of labeled neurons located in the ipsilateral area 7 and somatosensory area I (Fig. 6D, E; see also Figs. 2C, D and 4). No labeled cells were seen in the contralateral cortex. Within somatosensory area I, the corticotectal cells are located within medial and lateral agranular regions outside the granular or 'barrel field' portion of the somatosensory cortex (cf. ref. 26). Within the visual cortex, the location of the discrete patches of labeled corticotectal cells coincides with the location of discrete retinotopically organized visual areas (Fig. 6 C - E ) . These findings clearly confirm the earlier observations of Olavarria and Van Sluyters 41 and Killackey and Erzurumlu 26 in adult rats. By contrast, after H R P injections in the younger animals (P5-P10), labeled cells are widely distributed over the ipsilateral cortex including somatosensory, motor and association areas (Fig. 6A and B). Furthermore, within the ipsilateral visual cortex,

233

Fig. 4. Photomontage of parasagittal section through the cortex of an adult rat ipsilateral to the HRP-injected SC. Note that labeled cells are confined to lamina V and are arranged in discrete clusters (inserts A, B and C). Scale bar for inserts A, B and C, 100/~m. Horizontal arrow in D points in the rostral direction. Scale bar in D, 1 mm.

w h e r e e v e n at this early age the bulk of the l a b e l e d

U ~

40.

cells is l o c a t e d (Fig. 2B), t h e y are not g r o u p e d into

30,

specific clusters, but r a t h e r f o r m o n e large l a b e l e d a r e a i n v o l v i n g virtually all the r e t i n o t o p i c a l l y orga-

g'_

nized cortical areas (Fig. 6 A , B). A t all the ages studied, h o w e v e r , l a b e l e d cells w e r e c o n f i n e d to the cortex ipsilateral to the i n j e c t e d SC.

i- .a

~

-=1o,

DISCUSSION O.

. . . . . . . . . . . .

0

2

6

10 14 18 P O S T N A T A L AGE ( D A Y S )

l~ 22

, ADULT

Fig. 5. Graph illustrating the percentage of labeled corticotectal cells in lamina V of the striate cortex in animals of different ages. The plotted values are derived from the regions within the striate cortex in which the labeled cells clustered. In every :animal studied, the sampled region had a constant width and extended the entire thickness of lamina V. Vertical bars indicate + S.D. Note the very rapid increase in the percentage of labeled cells between P3 and P5 followed by a slower increase until P13. Note also the small decrease in the percentage of labeled cells occurring between P13 and P21.

Time-course of development of the corticotectal pathway T h e p r e s e n t results suggest that the c o r t i c o t e c t a l fibers in the rat do not r e a c h the superficial l a m i n a e of the ipsilateral SC until several days after birth. W e base this c o n c l u s i o n on the fact that the earliest age at which i n j e c t i o n s of H R P into the superficial layers of the ipsilateral SC l a b e l e d cells in the c o r t e x was P3 and t h e n u m b e r of such n e u r o n s was v e r y low in animals y o u n g e r t h a n P5. A n u m b e r of studies on the de-

234 A

Fll3stral ~

B

I;:;:.:,\ i ~,.;.~:~ \ Ik ~ - -

Lateral

POSTNATAL

c(f

ILate'a'

I T"-rq. ~b c,~

~<."-'U-'.'>Cauda,

a ,,,.

C ~ ~-

al

.,;~

~ _ _

DORSAL view

i~ " j, Medial I.i ~ J / ~ & Lateral

FEFd;

s, IVs"

2.21

O

.os,ral

--

" ~

._~2~'27U L

Caudal

LM

E ADULT ~z-~

I

Lateral

'

FI0 st r aik.~-

. . . .

_

"~ Caudal

Fig. 6. Areal distribution of corticotectal cells in P7 (A, B) and adult (D, E) rats. B and E are drawings of parasagittal sections at the levels indicated by the appropriate letters on the reconstructed dorsal aspects of P7 and adult brain, respectively. The hatched areas on the reconstructed dorsal views of the cortices outline regions containing massive numbers of labeled neurons while the dots represent individual labeled cells. Scale bars, 1 mm. C: dorsal view of neocortex of adult rat 9,12,27,4s. Visual cortex is outlined by solid lines while sensory and motor cortices are delineated by broken lines. Scale bar, 1 mm. M, motor cortex; FEF, frontal eye field; SI, somatosensory area I; SII, somatosensory area II; 7, area 7; 29d, retrosplenial cortex; 39, area 39; 41, area 41; visual cortex: AM, anteromedial; PM, posteromedial; V1, primary visual area (area 17); AL, anterolateral; LM, lateromedial; LL, laterolateral; LI, laterointermediate areas.

velopment of the primary optic pathway of the rat suggest that once axons are capable of the anterograde transport of large molecules (as indicated by very sensitive autoradiographic methods), they are also capable of transporting H R P in a r e t r o g r a d e direction to the p e r i k a r y a 5,34. Thus, it appears very unlikely that the axons of the corticotectal cells reach the SC before P3 but their p a r e n t cell bodies are not being labeled because of an inability of i m m a t u r e axons to transport H R P . The relatively p o o r labeling o b s e r v e d in the younger animals might be related to the spread of

terminal arborization of the corticotectal fibers. Thus, if in those younger animals the axons of corticotectal cells arborize over a wide part of the tectum. relatively few axonal terminals of a given corticotectal cell would be present in the vicinity of the injections. A l t h o u g h it can be argued that the a p p a r e n t immaturity of corticotectal cells in the young rat pups observed by us is due to their i n a d e q u a t e labeling with H R P , previous experiments from other laboratories using the Golgi technique have shown that cortical cells do not develop basal dendrites until postnatal day 721,39. F u r t h e r m o r e , from our results it is clear that the mean somal d i a m e t e r of the younger animals is significantly smaller than those of the mature animals. Thus our data suggest that at the time when the first corticotectal fibers reach the superficial layers of the superior colliculus the corticotectal cells are indeed immature. During the several days after the first corticotectal fibers reach the superficial laminae of the SC ( P 3 - P 4 ) , the n u m b e r of cells in the striate cortex sending their axons to the tectum appears to increase very rapidly. A f t e r P9 the increase appears to proceed at a slower pace and by P 1 3 - P 1 4 , the proportion of lamina V cells in the striate cortex projecting to the injected region of the SC reaches a p e a k value of slightly over 40%. It is interesting to note that in the adult golden hamster (a species closely related to the rat), a similar p r o p o r t i o n (about 40%) of cells in lamina Vb of the s e n s o r i m o t o r cortex projects to the spinal cord 23. Studies of M o n t e r o and others 40 as well as the recent study of Espinoza and T h o m a s 9 indicate that the visual cortex of the rat, like that of virtually all mammalian species studied so far (cf. ref. 22), can be subdivided into a n u m b e r of discrete retinotopically organized areas. The results of the discrete injections of H R P into the SC of adult rats conducted in the present study confirm the results of an earlier report by Olavarria and Van Sluyters 41. I n d e e d , both studies indicate that in adult rats each of the retinotopically organized cortical areas projects in a t o p o g r a p h ically organized pattern to the colliculus. Since from the time when the first corticotectal fibers reach the SC labeled cortical cells are present in both striate and extrastriate areas, it appears that the corticotectal projections from extrastriate areas develop simultaneously with those from the striate cortex.

235 It is important to emphasize that at any age, the somata of labeled corticotectal cells (if present) are found only in layer V, that is, in the layer to which they are restricted in the adult rat 50. Thus it appears that throughout development the laminar location of the corticotectal neurons, like that of other subcortically and callosally projecting cortical neurons (cf. refs. 20, 47), is precisely specified. The precise laminar specification of the location of the neuronal somata might in turn predetermine the direction of the initial axonal outgrowth (cf. ref. 16). Furthermore, it is worthwhile noting that by the time the first corticotectal fibers reach the superficial laminae of the SC (P3-P4) the gross errors in the retinotectap0, 37 and the retinogeniculate 19 projection have been already corrected. Secondly, the axons of geniculate relay cells enter the cortical plate at about the time of birth, reach their principal sites of termination (layers IV and I) on P1-P4 and form synapses in layer IV on P6-P9 35. Furthermore, recent work by Martin et al. 36 indicates that the number of geniculate cells projecting to the cortex increases rapidly during those first few postnatal days and a stable, approximately mature level is reached by P5. A similar sequence of events takes place in the fetal rhesus monkey46, 51 - the corticotectal fibers reach the stratum opticum of the SC at embryonic day 86 (E86), that is, some time after the geniculate input has reached the cortical plate (E78). Thus it appears that by the time the first corticotectal fibers reach the SC the colliculus is already precisely retinotopically organized by its retinal input and the striate cortex is already 'primed' by its geniculate input.

The 'exuberancy' of the early corticotectal projections It appears that during the period of rapid increase in number of cells in the striate cortex innervating the ipsilateral tectum (P5-P10), a substantial proportion of the corticotectal cells is located outside the visual cortex in somatosensory area I, motor and association cortices. Although at this stage of development the corticotectal cells located in the visual cortex form a fairly continuous band in lamina V extending throughout all the visual areas, they are rather patchily distributed outside the visual cortex. One can argue that since the SCs of younger animals are substantially smaller than those of older animals, HRP injections of comparable size in these

younger animals would involve a substantially greater proportion of collicular tissue. Thus, the change in areal distribution in the older animals might simply be the result of labeling a more discrete region in the SC. Indeed, in newborn rabbits the corticotectal projection from a given part of the striate cortex unlike that in the adult rabbits appears to occupy a big part of the collicular circumference in both the rostrocaudal and mediolateral directions and again unlike that in the adult rabbits, terminates mainly in the stratum opticum and the adjacent part of the stratum griseum superficiale 15. During the next two weeks or so (rabbits open their eyes for the first time at P l l - P 1 2 ) , the area of termination of the corticotectal fibers gradually becomes much more focused and shifts toward the middle part of the stratum griseum superficiale and stratum griseum zonale, that is, to the laminae at which corticotectal fibers terminate in the colliculi of the adult rabbits 15. However, the present study indicates that in the older rat pups, the areal distribution of corticotectal cells becomes more adult-like and by P12-P14 (that is, by the time the rat pups open their eyes for the first time) labeled cells within the visual cortex are distributed in a number of distinct clusters located in the regions presumably corresponding retinotopically to the injected collicular region. Furthermore, by that time, like in adult rats, only a fairly small proportion of labeled cells is located outside the visual cortex (almost exclusively in the ipsilateral somatosensory and association cortices; cf. ref. 26). The small drop in the proportion of labeled cells in lamina V of the striate cortex observed between P13-P14 and P21 might be at least partially due to the fact that in older animals we increased the sampling area to accommodate for the continuing increase in thickness of lamina V. That is to say, it might be an artifact of our measurement procedure rather than the result of a real loss of corticotectal cells. However, the almost complete disappearance of corticotectal cells in P12-P14 (or older) animals from the regions outside the visual cortex, suggests that the early corticotectal projection in the postnatal rat is 'exuberant' and a substantial degree of 'pruning' of these projections takes place during the second postnatal week. Other 'corticocaudal '1 projections in early postnatal rats seem to be even more extensive than the corticotectal projection. For example, in P2 rats, cells

236 which send their axons into the corticospinal tract (at least up to the level of the p y r a m i d a l decussation or cervical spinal cord) are distributed virtually throughout the tangential extent of layer V of the entire neocortex including visual cortex 1,56. It appears that many of these corticospinal cells send axon collaterals to the superior colliculus is. During the first two postnatal weeks, the cortical cells projecting to the spinal cord gradually become restricted to the cortical areas in which they are found in the adult. On the other hand, the distribution of corticospinal cells projecting into the spinal gray matter of P 5 - P 2 1 hamsters seems to be very similar to that seen in the adult hamsters 24. H o w e v e r , since in P 3 - P 1 0 hamsters there are about twice as many axons in the pyramidal tract as there are in the adult 23, it appears that the corticospinal p r o j e c t i o n in the neonatal hamsters like that in the neonatal rat is exuberant. The apparent discrepancy is p r e s u m a b l y due to the fact that axons of the majority of 'ectopic' cells o b s e r v e d during the d e v e l o p m e n t of a given p r o j e c t i o n do not invade the p r o p e r target tissue. A p p a r e n t l y many 'misprojecting' fibers wait in the vicinity of the target and withdraw before invading the target tissue (cf. ref. 20). A p p l i e d to the d e v e l o p m e n t of the corticotectal pathway in the postnatal rat, this would suggest that after H R P injections made into the corticotectal tract (rather than into the SC) an even higher p r o p o r tion of labeled cells would be located outside the visual cortex. Interestingly, although the visual system of the newborn cat is much more mature than that of the newborn rat (cf. for recent reviews refs. 6, 55), it has been claimed that the 'pruning' of the 'excessive' corticotectal projections (as indicated by a much m o r e restricted areal distribution of corticotectal cells and rapid decrease in the packing density of corticotectal cells located in the striate cortex) also occurs mostly during the second postnatal week 5s, that is, around or REFERENCES 1 Bates, C.A. and Killackey, H.P., The emergence of a discretely distributed pattern of corticospinal projection neurons, Dev. Brain Res., 13 (1984) 265-273. 2 Brizzee, K.R., Vogt, J. and Kharetchko, X., Postnatal changes in glia/neuron index with a comparison of methods of cell enumeration in the white rat. In D.P. Pnrpura and J.P. Schad6 (Eds.), Growth and Maturation o f the Brain, Progress in Brain Reserach, Vol. 4, Elsevier, Amsterdam,

shortly after eye opening which in the cat occurs at P7-P10. This in turn would suggest that unlike in the rat, in the cat the afferent visual input to the visual cortex might play a significant role in the changes of the pattern of distribution of corticotectal neurons. The more focused distribution of corticotectal cells and the reduction in the proportion of labeled cells in lamina V in the older animals are consistent with the loss of some corticotectal cells and/or the retraction of some of their axon collaterals (for review see ref. 7). The role of neuronal death in shaping corticotectal projections is, however, far from established. Thus, while it has been claimed that in the closely related noctural rodent, the mouse, about 30% of cortical neurons die between P5 and P3014, and in areas 17 and 18b (as well as in the cingulate cortices) of PS-P10 hamsters a n u m b e r of pyknotic (degenerating) cells has been observed, the n u m b e r of pyknotic cells in the infragranular laminae V and VI of the hamster's areas 17 and 18b appears to be small to negligible 10. On the other hand, the gradual restriction of the corticospinal projection in the postnatal rat appears to be achieved by eliminating corticospinal collaterals of cells located in lamina V of the posterior third of the neocortex which contains retinotopically organized visual areas TM. It is thus very likely that a similar mechanism is involved in the areal restriction of the corticotectal projection observed in the present study.

ACKNOWLEDGEMENTS We would like to thank M u r r a y J. McCall for his excellent technical help. Thanks are also due to Gordon Williams and Clive Jeffery for their assistance with photography. This project has been supported by a grant for the National Health and Medical Research Council of A u s t r a l i a to B.D. 1964, pp. 136-149. 3 Bruckner, G., Mares, V. and Biesold, D., Neurogenesis in the visual system of the rat. An autoradiographic investigation, J. Comp. Neurol., 166 (1976) 245-256. 4 Bunt, S.M. and Lund, R.D., Development of a transient retinoretinal pathway in hooded and albino rats, Brain Research, 211 (1981) 399-404. 5 Bunt, S.M., Lund, R.D. and Land, P.W., Prenatal development of the optic projection in albino and hooded rats. Dev. Brain Res., 6 (1983) 149-168.

237 6 Chalupa, L.M. and Williams, R.W., Prenatal development and reorganization in the visual system of the cat. In J. Stone, B. Dreher and D.H. Rapaport (Eds.), Developmental of Visual Pathways in Mammals, Liss, New York, 1984, pp. 89-102. 7 Cowan, W.M., Fawcett, J.W., O'Leary, D.D.M. and Stanfield, B.B., Regressive events in neurogenesis, Science, 225 (1984) 1258-1265. 8 Dreher, B., Potts, R.A. and Bennett, M.R., Evidence that the early postnatal reduction in the number of rat retinal ganglion cells is due to a wave of ganglion cell death, Neurosci. Lett., 36 (1983) 255-260. 9 Espinoza, S.G. and Thomas, H.C., Retinotopic organization of striate and extrastriate visual cortex in the hooded fat, Brain Res., 272 (1983) 137-144. 10 Finlay, B.L. and Slattery, M., Local differences in the amount of early cell death in neocortex predict adult local specializations, Science, 219 (1983) 1349-1351. 11 Graybiel, A.M. and Devor, M., A microelectrophoretic delivery technique for use with horseradish peroxidase, Brain Res., 68 (1974) 167-173. 12 Hall, R.D. and Lindholm, E.P., Organization of motor and somatosensory neocortex in the albino rat, Brain Res., 66 (1974) 23-38. 13 Heumann, D., Leuba, G. and Rabinowicz, T., Postnatal development of the mouse cerebral neocortex. II. Quantitative cytoarchitectonics of visual and auditory areas, J. Hirnforsch., 18 (1977) 483-500. 14 Heumann, D., Leuba, G. and Rabinowicz, T., Postnatal development of the mouse cerebral neocortex. IV. Evolution of the total cortical volume, of the population of neurons and glial cells, J. Hirnforsch., 19 (1978) 385-393. 15 Hollander, H., Tietze, J. and Distel, H., An autoradiographic study of the subcortical projections of the rabbit striate cortex in the adult and during postnatal development, J. Comp. Neurol., 184 (1979) 783-794. 16 Horder, T.J. and Martin, K.A.C., Morphogenetics as an alternative to chemospecificity in the formation of nerve connections, Symp. Soc. Exp. Biol., 32 (1978) 275-358. 17 Huerta, M.F. and Harting, J.K., The mammalian superior colliculus: studies of its morphology and connections. In H. Vanegas (Ed.), Comparative Neurology of the Optic Tecturn, Plenum, New York, 1984, pp. 687-773. 18 Ivy, G.O. and Killackey, H.P., Corticospinal and corticotectal neurons: mechanisms of ontogenetic changes in distribution, Soc. Neurosci. Abstr., 7 (1981) 178. 19 Jeffery, G., Retinal ganglion cell death and terminal field retraction in the developing rodent visual system, Dev. Brain Res., 13 (1984) 81-96. 20 Jones, E.G., Development of connectivity in the cerebral cortex. In W.M. Cowan (Ed.), Studies in Developmental Neurobiology, Oxford University Press, New York, 1981, pp. 354-394. 21 Juraska, J.M. and Fifkova, E., A Golgi study of the early postnatal development of the visual cortex of the hooded rat, J. Comp. Neurol., 183 (1979) 247-256. 22 Kaas, J.H., A comparative survey of visual cortex organization in mammals. In S.O.E. Ebbesson (Ed.), Comparative Neurology of the Telencephalon, Plenum, New York, 1980, pp. 483-502. 23 Kalil, K., Development and regrowth of the rodent pyramidal tract, Trends Neurosci., 7 (1984) 394-398. 24 Kalil, K. and Reh, T., Establishment of cortical topography in hamster pyramidal tract neurons during development,

Soc. Neurosci. Abstr., 7 (1981) 177. 25 Keefer, D.A., Horseradish peroxidase as a retrogradely transported detailed dendritic marker, Brain Res., 140 (1978) 15-32. 26 Killackey, H.P. and Erzurumlu, R.S., Trigeminal projections to the superior colliculus of the rat, J. Comp. Neurol., 201 (1981) 221-242. 27 Krieg, W.J.S., Connections of the cerebral cortex. I. The albino rat. A. Topography of the cortical areas, J. Comp. Neurol., 84 (1946) 221-275. 28 Laemle, L.K. and Labriola, A.R., Retinocollicular projections in the neonatal rat: an anatomical basis for plasticity, Dev. Brain Res., 3 (1982) 317-322. 29 Lam, K., Sefton, A.J. and Bennett, M.R., Loss of axons from the optic nerve of the rat during early postnatal development, Dev. Brain Res., 3 (1982) 487-491. 30 Land, P.W. and Lund, R.D., Development of the rat's uncrossed retinotectal pathway and its relation to plasticity studies, Science, 205 (1979) 698-700. 31 Ling, E.A. and Leblond, C.P., Investigation of glial cells in semithin sections. II. Variation with age in the numbers of the various glial cell types in rat cortex and corpus callosum, J. Comp. Neurol., 149 (1973) 73-82. 32 Ling, E.A., Paterson, J.A., Privat, A., Mori, S. and Leblond, C.P., Investigation of glial cells in semithin sections. I. Identification of glial cells in the brain of young rats, J. Comp. Neurol., 149 (1973) 43-72. 33 Lithgow, T. and Barr, G.A., A method for stereotaxic implantation in neonatal rats, Dev. Brain Res., 2 (1982) 315-320. 34 Lund, R.D. and Bunt, A.H., Prenatal development of central optic pathways in albino rats, J. Comp. Neurol., 165 (1976) 247-274. 35 Lund, R.D. and Mustari, M.J., Development of the geniculocortical pathway in rats, J. Comp. Neurol., 173 (1977) 289-306. 36 Martin, P.R., Nichol, K., Sefton, A.J. and Nisbett, G., No evidence for loss of cells in the rat's lateral geniculate nucleus projecting to visual cortex during early postnatal development, Proc. Aust. Physiol. Pharmacol. Soc., 15 (1984) 217P. 37 Martin, P.R., Sefton, A.J. and Dreher, B., The retinal location and fate of ganglion cells which project to the ipsilateral superior colliculus in neonatal albino and hooded rats, Neurosci. Lett., 41 (1983) 219-226. 38 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem. Cytochem., 26 (1978) 106-117. 39 Miller, M., Maturation of rat visual cortex. I. A quantitative study of Golgi-impregnated pyramidal neurons, J. Neurocytol., 10 (1981) 859-878. 40 Montero, V.M., Comparative studies on the visual cortex. In C.N. Woolsey (Ed.), Cortical Sensory Organization, Vol. 2, Multiple Visual Areas, Humana, Clifton, NJ, 1981, pp. 33-81. 41 Olavarria, J. and Van Sluyters, R.C., The projection from striate and extrastriate cortical areas to the superior colliculus in the rat, Brain Res., 242 (1982) 332-336. 42 Paxinos, G. and Watson, C., The rat brain in stereotaxic coordinates, Academic Press, Sydney, 1982. 43 Perry, V.H., Henderson, Z. and Linden, R., Postnatal changes in retinal ganglion cell and optic axon populations

238

44

45

46

47

48

49

50

51

in the pigmented rat, J. Comp. Neurol., 219 (1983) 356-368. Potts, R.A., Dreher, B. and Bennett, M.R., The loss of ganglion cells in the developing retina of the rat, Dev. Brain Res., 3 (1982) 481-486. Raedler, A. and Sievers, J., The development of the visual system of the albino rat, Adv. Anat. Embryol. Cell. Biol., 50 (1975) 5-87. Rakic, P., Genesis of visual connections in the rhesus monkey. In R.D. Freeman (Ed.), Developmental Neurobiology of Vision, Plenum, New York, 1979, pp. 249-276. Rakic, P., Developmental events leading to laminar and areal organization of the neocortex. In F.O. Schmitt, F.G. Worden, G. Adelman and S.G. Dennis (Eds.), The Organization of the Cerebral Cortex, MIT Press, Cambridge, MA, 1981, pp. 7-28. Sanderson, K.J., Welker, W. and Shambes, G.M., Reevaluation of motor cortex and of sensorimotor overlap in cerebral cortex of albino rats, Brain Res., 292 (1984) 251-260. Sefton, A.J. and Lam, K., Quantitative and morphological studies on developing optic axons in normal and enucleated albino rats, Exp. Brain Res., 57 (1984) 107-117. Sefton, A.J., Mackay-Sim, A., Baur, L.A. and Cottee, L.J., Cortical projections to visual centres in the rat: an HRP study, Brain Res., 215 (1981) 1-13. Shatz, C.J. and Rakic, P., The genesis of efferent connec-

52

53

54

55

56

57

58

tions from the visual cortex of the fetal rhesus monkey, J. Comp. Neurol., 196 (1981) 287-307. Sherwood, N.M. and Timiras, P.S., A stereotaxic atlas of the developing rat brain, University of California Press, Berkeley, 197(I. So, K.-F., Schneider, G.E. and Frost, D.O., Postnatal development of retinal projections to the lateral geniculate body in Syrian hamsters, Brain Res., 142 (1978) 343-352. So, K.-F., Woo, H.H. and Jen, L.S., The normal and abnormal postnatal development of retinogeniculate projections in golden hamsters: an anterograde horseradish peroxidase tracing study, Dev. Brain Res., 12 (1984) 191-205. Sretavan, D. and Shatz, C.J., Prenatal development of individual retinogeniculate axons during the period of segregation, Nature (London), 308 (1984) 845-848. Stanfield, B.B., O'Leary, D.D.M. and Fricks, C., Selective collateral elimination in early postnatal development restricts cortical distribution of rat pyramidal tract neurones, Nature (London), 298 (1982) 371-373. Thong, I.G. and Dreher, B., The development of the cortico-tectal pathway in the albino rat, Proc. Physiol. Pharmacol. Soc., 15 (1984) 211P. Tsumoto, T., Suda, K. and Sato, H., Postnatal development of corticotectal neurons in the kitten striate cortex: a quantitative study with the horseradish peroxidase technique, J. Comp. Neurol.. 219 (1983)88-99.