Chapter 8 The development of the rat prefrontal cortex : Its size and development of connections with thalamus, spinal cord and other cortical areas

H.B.M. Uylings. C.G. Van Eden, J.P.C. De Bruin, M.A. Corner and M.G.P. Feenstra (Eds.) Progress in Brain Research, Vol. 85

0 1990 Elsevier Science Publishers B.V. (Biomedical Division)

169

CHAPTER 8

The development of the rat prefrontal cortex Its size and development of connections with thalamus, spinal cord and other cortical areas C.G. van Eden, J.M. Kros* and H.B.M. Uylings Netherlands Institute for Brain Research, Amsterdam, The Netherlands

Introduction

The development of the brain has always attracted much interest, not only because of the intriguing processes which lead to a persons mental development, but also because of the intellectual challenge of understanding the biological mechanisms behind the development of such a complex and crucial organ as the brain. Many of these questions concern the generation, migration and differentiation of the constituent elements (the neurons and glial cells), and how these cells come to establish the “appropriate” contacts. Another kind of question concerns the functional development. When does a certain structure begin to operate, and when does its functional maturation stop (e.g. Goldman, 1971, 1972;Diamond and Goldman-Rakic, 1989). The following two chapters deal with the development of the cortical layers, the development of cortical neurons and ingrowth of afferent fibers in the

* Present address: Department of Pathology, Erasmus University, Dr. Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. Correspondence: Dr. C.G. van Eden, Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands.

human fetus (KostoviC, and Mrzljak et al., this volume). The scope of the present paper is to give a description of some of these processes in the rat. On the basis of this description we will then focus upon the question of whether or not the prefrontal cortex has a later or prolonged development in comparison with other cortical areas. For a general review of the pre- and postnatal development of the rat cerebral cortex, the reader is referred to Uylings et al. (1990). Development of the prefrontal cortex

The idea that the prefrontal cortical areas develop later than other cortical areas goes back to the work of the German neuroscientist Paul Flechsig. In 1898,Flechsig (see Flechsig, 1901)published his studies on the development of myelinization of the human cerebral cortex, which demonstrated that regional differences in the developmental myelinization fall into 3 groups. First of all, those cortical areas which attain an adult-like myelinization pattern already prior to birth all belong to the sensoric/motoric areas (areas 1 - 10 in Fig. 1). The second group, containing areas l l - 30, also starts to develop before birth but continues myelinization into the first months of postnatal life. This group comprises, among others, the premotor areas and

170

higher order sensory areas. The third group is formed by areas which do not show any signs of myelinization until 1 month after birth. All of the areas in this group belong to the so-called “association” areas. Fig. 1 summarizes these results and demonstrates that the prefrontal areas are among the latest of all to develop with only the superior temporal area being later. Furthermore, Yakovlev and Lecours (1967) claim that myelinization of the prefrontal areas has not stabilized until the fourth decade of human life. Another indication for a late and/or prolonged development of the prefrontal cortex comes from more recent studies investigating the agedependent effects of prefrontal lesions. Lesion

studies have demonstrated that early lesions of the prefrontal cortex have little effect on those functions which in adulthood depend upon the integrity of the prefrontal cortex. In rats and monkeys the ability to functionally recover from prefrontal ablations persists until relatively late into postnatal life. In the monkey, for example, these investigations suggest that orbital prefrontal cortex may become funtionally mature by 1 year of age, whereas the dorsolateral cortex does not approach functional maturity until well into the second year of life (Goldman, 1971, 1972). In the rat, Kolb and Nonneman (1976) demonstrated that orbital prefrontal cortex lesions do not produce signs of adipsia or aphagia when inflicted before 60 days of age. Later studies showed that also the medial prefrontal cortex has a prolonged potential for recovery (Kolb and Nonneman, 1978); Nonneman and Corwin, 1981; Kolb, 1984). This prolonged ability to recover functionally has been taken as being indicative of a relatively late maturation of these cortical areas, since the brain is considered to retain its plasticity as long as its connections have not yet become stabilized. In the immature brain, therefore, the function of the lesioned prefrontal cortex could be taken over by other parts of the (prefrontal) system (see also Kolb, this volume). Lesions performed at a later age have more serious effects because the prefrontal system would have lost its flexibility in the course of maturation (e.g. Goldman, 1972; Goldman-Rakic and Galkin, 1978; Finger and Amli, 1985).

All these examples, the old myelinization study as well as more recent neonatal lesion studies, point to a late or prolonged development of the prefrontal areas in rodents as well as in primates. This chapter will describe the structural development of these prefrontal areas in the rat in comparison with other cortical areas. B

Fig. 1 . Cortical areas on the lateral (a) and medial (b) aspects on the cerebral hemisphere. Numbering in sequence of myelogenesis, after Flechsig (1901).

Cytoarchitectonic development

The postnatal development of the rodent prefrontal cortex is characterized by the transformation of

171

the cortical plate into the laminated adult cortex (see Fig. 2) (Van Eden and Uylings, 1985a). At birth, the cortical plate contains most of the neurons that will contitute the adult cortex but, until postnatal day (P) 2, some of the last generated neurons are still migrating towards the surface of the cortical plate (Raedler and Sievers, 1975; Uylings et al., 1990). On the first postnatal day (PI) several developing layers can be distinguished in the cerebral wall (Fig. 2). The outermost layer (layer I) is the former marginal zone, while most of the cerebral wall is formed by the cortical plate. Within the latter a certain degree of lamination can already be discerned. The most superficial part is the youngest. This dense part of the cortical plate contains immature neuronal elements of the future layers 11-V. In the more mature, and less cell dense, lower part of the cortical plate, the developing layer VI and part of V can be seen. Underneath the cortical plate the future white matter can be found with in it the remnant neurons of the subplate zone as we know from human and cat studies (Kostovid and Molliver, 1974; Kostovid and Rakic, 1980; Valverde and Facal-Valverde, 1987;

Chun et al., 1987; for rat see also Uylings et al., 1990). By the end of the first postnatal week (P6 - P7), the upper dense part of the cortical plate is still present but is greatly reduced in width. Underneath the dense part of the cortical plate the CP has developed layers VI and V and, in the dorsolateral cortex, the granular layer IV can be recognized. In contrast to the human cortical development where a granular layer is initially formed even in areas that in adulthood are agranular, in the agranular rat prefrontal cortex no layer IV is ever formed during development. By P7 all layers 11- VIb can recognized in the frontal pole of the cerebral hemisphere, although layers I1 and 111 are still very immature and tightly packed. Furthermore, by day 6, development, of the frontal cortex has progressed so far that using slightly adapted criteria the prefrontal areas can be distinguished (Van Eden and Uylings, 1985a). At P10 the cortical plate has completely disappeared and layer I1 and I11 can clearly be recognized as two distinct layers. From day 10 through day 18 the cytoarchitectonic features of the cortical

PL I

n+m I

v

CP

ma

SP

YIb 1

6

10

14

18

24

30

60

90 days

Fig. 2. Cyotarchitectonic development of the prefrontal cortex. Cross-sections through the prelimbic area of the medial PFC at various stages of postnatal development CP = cortical plate; SP = subplate layer; Roman numbers indicate cortical layers.

172

layers mature into the adult agranular pattern. Around day 18 the cortical laminae are most clearly delineated. Although between day 18 and 30 the laminar pattern faints somewhat, there is little change in the overall cytoarchitectonic pattern during this period. However, important changes do occur during this period (see “volumetry” below). Later, from day 30 onwards until day 90 the acquired characteristics blur further and the individual variation in the cytoarchitecture increases. These developmental changes in the appearance of the laminae are caused by a further dispersal of the cells, presumably as a result of differentiation and growth of the dendrites of the neurons present. If we compare this account of the development of the prefrontal cortex with the descriptions of cytoarchitectonic development in other regions of the cortex, we must conclude that there is much similarity in the rate of development of cortical lamination in different cortical areas (e.g. Eayrs and Goodhead, 1959; Raedler and Severs, 1975; Rice and van der Loos, 1977). These studies described the postnatal cytoarchitectonic development of the visual and somatosensory cortex in the rat. According to their observations the CP has differentiated into layers I1 and I11 at P7 and all cortical layers have attained their adult features by P14. Further changes, after P14, are only quantitatively dependent on a further decrease in cell density through continued increase of the neuropil. Leuba et al. (1976) and Heumann et al. (1976) give a similar rate of development for the maturation of a whole range of cortical areas. In all these areas (viz. areas 2, 3, 4, 10 an 17 and 18) layers 11-VI become formed between P5 and P10. It is clear, therefore, that the differentiation of cortical lamination occurs not later in the PFC than in other (neo)cortical areas. Development of connections between the mediodorsal thpIamie uuch!us and the prefrontal cortex The development of what is considered the most important afferent system for the prefrontal cortex

(Leonard, 1969; Krettek and Price, 1977), the mediodorsal nucleus of the thalamus (MDT) was studied in the postnatal period by means of an anterograde tracer (Van Eden, 1986). The development of other important afferent prefrontal fiber systems, such as the dopaminergic projection from the ventral tegmentum and the basolateral nucleus of the amygdala will be presented elsewhere (Kalsbeek et al., this volume; Verwer and van Vulpen, 1989). From other studies it is known that the specific thalamic fibers from, e.g., the lateral geniculate nucleus reach the visual cortex around ED18 (Lund and Mustari, 1977). They remain in the subplate layer (see also KostoviC, this volume) until the day of birth (ED22), at which point they enter the cortical plate. After entering the cortex these fibers stay within the basal layers of the cortical plate until postnatal day 3. Only on day 4, when layer IV neurons have evolved from the dense part of the cortical plate, do they finally make contact with their adult target cells (Lund and Mustari, 1977; Wise and Jones, 1978). The ingrowth of the mediodorsal fibers into the postnatal prefrontal cortex differs only in details from this scheme (see Fig. 3). In contrast to the thalamocortical fibers in the granular sensory cortices, not all of the mediodorsal fibers in the prefrontal area wait in either the subplate layer or the basal cortical laminae until day 3. The majority of the MDT fibers waits but already by the day of birth a substantial number of the MDT fibers have penetrated the superficial, dense part of the cortical plate and some are observed in the future layer 1. During the following 3 - 4 days an increasing number of fibers accumulates in the upper cortical plate but it will take until P7 before layer I1 can be distinguished. The ingrowth of fibers from the mediodorsal thalamic nucleus, therefore, is not later in the prefrontal cortex than the thalamocortical fibers in the primary sensory areas. They are, in fact, relatively earlier in penetrating the immature cortical plate (dense part) and could thereby exert a potentially greater influence than do other thalamocortical projections on the development of the cortex. By P7, when the cortical plate is no longer distinguish-

173

ed and layers I1 and I11 have become visible as separate layers, the distribution of MDT fibers is essentially the same as in the adult. These fibers terminate densely in layers I, 111 and upper part of

layer V of the medial and orbital prefrontal cortex. This scheme of ingrowth of thalamic fibers into the prefrontal area is comparable with the ingrowth into the visual or the somatosensory cortex, where

Fig. 3. Ingrowth of mediodorsal thalamic fibers into the medial prefrontal cortex. The thalamic fibers are labeled by iontophoretic WGA-HRP injection in the mediodorsal nucleus at PO (a, b), P4 (c, d) and P7 (e, 0;Dark-field illumination (a, c, e) and counterstained section (b, d, f). SP = subplate layer; UCP = upper part of the cortical plate; w = white matter; Roman numbers indicate cortical layers.

174

these fibers attain an adult-like termination pattern between P6 and P7 (Lund and Mustari, 1977; Wise and Jones, 1978). The reciprocal projection from the prefrontal cortex to the mediodorsal thalamus is largely formed during the second week of postnatal life. The first retrogradely labeled cells are observed after injections of WGA-HRP in the mediodorsal thalamus on P4. Between P4 and P10 the number of pyramidal cells in layer VI increases and, by P10, retrogradely labeled cells begin to be observed in the hemisphere opposite to the injection site (Van Eden, 1986).

Volumetric developmeut Another way to monitor the development of the PFC is to measure its increase in size during the first postnatal months. Volumetry has advantages over measurement of cortical thickness because it is a parameter which is rather insensitive to minor deviations of the sectioning plane. Furthermore, it takes into account any developmentally occuring changes in the shape of the object (Uylings et al., 1984). The volumetric analysis of prefrontal cortex development reveals a different growth pattern than for cytoarchitectonic development. Since we were able to distinguish between prefrontal subareas from day 6, the areas could reproducibly be delineated (Van Eden and Uylings, 1985a), and, therefore, the volumes could be determined. The for shrinkage corrected, volumetric growth of the prefrontal areas (Fig. 4) shows a distinctly different pattern than that of the brain as a whole (Van Eden and Uylings, 1985b). All prefrontal areas irrespective of whether they are located in the medial or the lateral parts, show a period of transient overgrowth. At first, during the first weeks of postnatal life, the volume increases. For the medial PFC this increase lasts until P24, whereas in the orbital PFC growth continues until at least P30. After this period the volume is higher than in the adult rat. In the medial PFC the volume overshoot is approximately 30% higher than the adult value, whereas in the orbital PFC this is approximately 80%.

Comparing the individual subareas within the medial and orbital PFC, two kinds of growth patterns can be distinguished. The shoulder area of the hemisphere (i.e., the PFC subareas which are innervated by the lateral subnucleus of the MD) shows a different volumetric growth pattern than does the rest of the prefrontal areas. These PFC subareas, ACd and PrC,, differ from the PL and the orbital PFC subareas with respect to both, the maximum volume, and the rate of volumetric decline thereafter. In the PFC subareas of the “shoulder region”, this decline is relatively fast and the adult volume is attained already at P30, whereas the other PFC subareas continue to decline until at least day 90. Since day 90 is usually regarded as the beginning of the adult stage in the rat, a developmental process which continues until this age seems to point to a rather protracted maturation of at least some of the prefrontal areas. There are no volumetric data available on the development of other neocortical areas in the rat, although a transient volumetric overgrowth has been demonstrated to also occur in different brain regions in Tzipaia belatzgeri (viz., striatum, visual cortex and subcor-

6

10 14 16

24

M

age (days)

60

90

Fig. 4. Volumetric development of the medial and orbital prefrontal cortex. Upper 3 lines: growth curves of medial cortical areas, viz. prelimbic area (PL), dorsal anterior cingular area (ACd) and medial precentral area (PrCm). Lower lines: growth curves of the orbital prefrontal areas, viz. the ventral and dorsal agranular insular areas (AIvand AI,, respectively).

175

tical nuclei of the visual system (Zilles, 1978)) and area striata of the marmoset monkey (Fritschy and Garey, 1986). Therefore, a direct comparison with other cortical areas cannot be made in the rat. At the moment it is not clear which elements of the cortex cause the transient volumetric overgrowth. Counting of cells in the shoulder region of the PFC showed no significant changes in the neuron number during the period of overgrowth (Zijlstra er al., 1988). The volume of the cortex is mainly determined by the dendrites and the somata of the constituent neurons. In the rat visual cortex these elements also demonstrate a transient overgrowth. The neuronal somata reach their maximal volume between postnatal days 18 and 22 (Werner et al., 1981), and also the dendrites of pyramidal and non-pyramidal cells have their maximal extension around P l 3 (Parnavelas and Uylings, 1980; Uylings et al., 1990). Compared to these parameters the volumetric overgrowth in the prefrontal cortex seems to indicate a relatively late and prolonged maturation. Since the transient volumetric overgrowth probably coincides with a transient extension of dendritic trees and neuronal somata, and not to a reduction in cell number, the possibility exists that reduction of volume is, somehow, related to interneuronal connectivity, e.g. a reduction or elimination of axonal branches or terminals. This possibility is further corroborated by reduction of plasticity of the mPFC after P25, i.e. during the period of volumetric decline (Kolb and Nonneman, 1978; Nonneman and Corwin, 1981).

Development and regression in the cortical afferents and spinal efferent systems Regression and degeneration are processes which are frequently reported to occur during brain development. Programmed cell death, and the elimination of collateral axons and synapses have been observed in many parts of the developing brain and seem to represent a fundamental principle. A restriction in the projection pattern of neurons has been reported to occur during normal

development in various areas of the CNS in a variety of spieces (Dehay and Kennedy, 1984; Jeffery et al., 1984; Innocenti, 1981, 1986; Ivy et al., 1984; Killackey and Chalupa, 1986; Olivarria and Sluyters, 1985; Provis et al., 1985). For example, developmental restriction has been demonstrated in retinal projections (Jeffrey et al., 1984; Provis et al., 1985), and in ipsilateral (Dehay and Kennedy, 1984), as well as commissural, corticocortical projections of the visual, parietal, temporal and somatosensory cortices (Innocenti, 1981, 1986; Ivy et al., 1984; Killackey and Chalupa, 1986; Olivarria and Sluyters, 1985). In corticocortical projections the developmental restriction probably serves to refine the initial distribution of connections between neurons. The development of corticocortical projections has been studied in different species: rat (Crandal and Caviness, 1984; Ivy et al., 1984), cat (Innocenti, 1981) and monkey (Swartz and Goldman-Rakic, 1982). In the rat the commissural corticocortical projections cross the corpus callosum soon after the first bridge between the hemispheres, the so-called “glial sling”, is formed (Hankin and Silver, 1986). The corpus callosum is generated between embryonic days (E) 17 and 20 (Crandal and Caviness, 1984; Floeter and Jones, 1985; Uylings et al., lM),and on El8 the cortical afferents can be found in the subplate layer underneath the cortical plate. In rat (Ivy et al., 1984), cat (Innocenti, 1981, 1986), and monkey (Killackey and Chalupa, 1986) it was shown that these initial pioneer fibers are more in number and have a more widespread distribution in the contralateral hemisphere than is found in the adult animal. Subsequently, during the early postnatal period (2nd postnatal week), these projections from the visual and somatosensory cortex are refined by elimination of axon collaterals as could be demonstrated by using fluorescent retrograde tracers as either long or short term markers. However, within the prefrontal corticocortical projection systems of the monkey no reduction could be demonstrated during development (Schwartz and Goldman-Rakic, 1982). We have compared the neonatal and adult

176

distribution of cortical cells projecting to the medial prefrontal cortex in the rat, either through the corpus callosum or ipsilaterally. Fig. 5 shows cells retrogradely labeled -after an injection with the fluorescent tracer Fast Blue made on P5. The injection site included the entire medial prefrontal cortex including the subareas PL, ACd and PrCm (Fr2). Comparison of the distribution of the retrogradely labeled cells in the contra- and ipsilateral hemispheres with cells labeled by an injection of Fast Blue labeling the same mPFC subareas in the adult rat (Fig. 6) shows no obvious restrictions of the neonatal pattern, either in the commissural or in the ipsilateral projections to the prefrontal cortex. Therefore, it does not seem likely that elimination of transient collaterals plays an important role in development of prefrontal corticocortical connections in the rat. On the other hand, a developmental reduction of fibers has also been described to occur in another cortical projection system: the cor-

Fig. 5. Ipsilateral distribution of retrogradely labeled cortical cells after injection of 0.5 pl of Fast Blue into the medial PFC of a neonatal rat pup (PS). Injection site is comparable with Fig. 6.

ticospinal tract (CST) (Bates and Killackey, 1984; O’Leary et al., 1981; O’Leary and Stanfield, 1985, 1986; Stanfield and O’Leary, 1985; Stanfield et al., 1982). O’Leary and Stanfield showed that the layer V cells of the occipital rat cortex could be retrogradely labeled after second postnatal day injections with fluorescent tracers into the pyramidal decussation of the lower medulla (O’Leary et al.,

A

Fig. 6 . Ipsilateral distribution of retrogradely labeled cortical cells after a 1 4 injection of Fast Blue into the medial PFC. (a) Injection site; (b) distribution of FB-labeled cells at an anterior level comparable to that of Fig. 5. One symbol stands for 5 - 10 labeled cells.

177

1981). In adulthood this area of the cortex does not contribute to the CST. The fibers which constitute the CST in the adult rat originate in the rostral two thirds of the cortex. A major part of the CST fibers from the occipital cortex are retracted during the second postnatal week (Stanfield and O’Leary, 1985). As in the commissural systems of certain cortical areas, the restriction in CST projections from the occipital cortex is caused by selective collateral elimination rather than by cell death (O’Leary et al., 1981; Stanfield et al., 1982). By using retrogradely transported fluorescent dyes as either short term or long term markers, occipital neurons, which had a transiently extended pyramidal tract axon, appeared to have concurrent projections to either the superior colliculus or the pons (O’Leary and Stanfield, 1985). After retraction of the pyramidal tract collaterals, the occipital neurons maintain their connections with one of these subcortical nuclei. In the adult rat, injections of anterograde tracers into prefrontal areas fail to identify fibers projecting toward the spinal cord (Beckstead, 1979). Recent reports suggest that more rostral areas of the cortex may also contain neurons which project to the CST during a limited period of postnatal development (Schleyer and Jones, 1988). By injecting the retrograde tracer Fast Blue into the cervical spinal cord, between C5 and C7, in newly born and young adult animals we have investigated a possible reduction of CST fibers from the prefrontal areas. Since Fast Blue is taken up by damaged as well as undamaged axons passing through the injection site, cortical neurons that extend axons to the level C5 - C7 or further caudally are labeled by such injections which include the corticospinal tract (examples of injections are shown in Fig. 7). Animals in which the tracer spread through the central canal to more rostral levels were not used for analysis. Injections in the cervical spinal cord of neonatal rats result in a continuous band of cortical neurons in the frontal pole of the hemisphere (Fig. 8). Labeled cells are medium-sized and large pyramidal neurons. In the ventro-medial and the

Fig. 7. Injectionsites in the cervical spinal cord of (a) neonatal (P5)rat pup and (b) adult rat. Retrogradely labeled cells in the PFC are shown in Figs. 8 and 9, respectively.

medio-orbital parts of the frontal pole (areas: IL, MO, VO, VLO,) where layer V is thinner than in the dorsolateral and dorsomedial parts of the cortex, the labeled cells appear to be smaller and fewer in number. The medial half of VLO contains very few labeled cells. Further caudally, the band of labeled cells extends from the fundus of the rhinal sulcus to the ventral border of the medial cortex (Fig. 8c). At this level, however, the band is interrupted: no labeled cells are observed in the insular cortex immediately dorsal to the rhinal sulcus, i.e. the ventral subdivision of the agranular insular cortex (AIv). The posterior agranular insular area (AXp), which lies caudal of AI,, contains many labeled cells. At this level, the band of labeled neurons is again uninterrupted extending from the

178

rhinal sulcus to the cingulate cortical areas at the medial aspect of the hemisphere. This pattern of labeling continues in caudal direction, until at least the level of the anterior commissure. .

a

In adult rats similar injections label cells in a much more restricted area than CST injections in the neonatal rat. In the adult rat, the orbital PFC areas in the frontal pole are completely devoid of

b

Fig. 8. Transverse sections through the frontal pole of the left hemisphere after neonatal injection with Fast Blue in the cervical spinal cord (see Fig. 7A). (a) FB-labeled cells in layer V at level A 7.5 mm (Sherwood and Timiras, 1970), (b) Same section as (a) cresyl-violet counterstained. (c) FB-labeled cells in at level A 7.0 mm. (d) Same section as (c) cresyl-violet counterstained.

179

labeled cells after CST injections (Fig. 9). No labeled cells can be observed in MO, VO, VLO, LO, AI,, or AId. In dorsolateral parts of the frontal cortex, containing the motor cortices (Frl, Fr3),

a

C

and in the medial prefrontal areas (PrC,, AC,), a great number of labeled cells is still found within layer V. Few cells are located in the prelimbic (PL) and infralimbic (IL) areas on the medial wall of the

b

d

Fig. 9. Transverse sections through the frontal pole of the lefit hemisphere after injection with Fast Blue in the cervical spinal cord of an adult rat (see Fig. 7b). (a) FB-labeled cells in la)rer Vat lev el comparable to Fig. 8a. (b) Same section as (a) cresyl-violet cowiterstained. (c) FB-labeled cells in at level comparable t o Fig. Sc. (d) Same section as (c) cresyl-violet counterstained.

180

cortex, especially in the most rostral parts. Few labeled cells are also observed in the frontal parts of the parietal area (Parl). Another group of labeled cells is situated in Alp lining the shallow parts of the rhinal sulcus. Comparison of the labeling pattern after neonatal and adult corticospinal tract (CST) injections shows that several areas that have a corticospinal (CS) projection in neonatal stages fail to be labeled in adulthood. In the frontal cortex, this phenomenon is particularly evident in the orbital areas, where, after neonatal CST injections, numerous labeled layer V cells are found in all these orbital cortical areas, whereas the adult labeling pattern in the frontal cortex is similar to that previously reported (Hicks and D’Amato, 1977; Leong, 1983; Wise and Jones, 1977; Miller, 1987; Schleyer and Jones, 1988). Similar to these reports we failed to observe labeled cells in the following orbital areas: MO, LO, VO, VLO, AI, or AId. The orbital prefrontal area AIv is an exception in this respect since neither adult nor neonatal CST injections result in labeled CST neurons in this area. Thus, neurons in this area may never sustain a CST projection or, alternatively, they may grow a CS axon which fails to reach the level of the caudal cervical spinal cord. The present study can not be conclusive on this point. The cells which can no longer be labeled in the adult rat either disappear in the course of development, or survive but do not longer project toward the spinal cord. In an attempt to make a distinction betwee these two mechanisms, we made use of the property that retrogradely transported FB remains visible in the cell soma for extended periods without leakage (e.g. O’Leary et al., 1981). Six pups received CST injections on day 5 and were allowed to survive until 35 days of age. This time interval was chosen because it is well beyond the period in which the transient CS neurons retract their spinal projections (Schleyer and Jones, 1988; Stanfield and O’Leary, 1985; Joosten and Van Eden, 1989). This experiment demonstrates that the cells which are labeled by CST injections during the first postnatal week are not eliminated by

cell death (Fig. lo), and, therefore, are likely to retract their spinal collaterals similar to the developing neurons in the occipital cortex. Recent investigations (Joosten and Van Eden, 1989) using anterograde tracers support this observation by indicating that the number of axons in the spinal cord capable of beiag labeled by injections in the medial prefrontal cortex (ACd, PL, IL) initially increases between day 3 and day 7. At day 14, however, very few, if any axons running down the CST can be labeled, Therefore, a number of projections from PL and IL also seem to be transient. Furthermore, the extension and subsequent retraction of axons from this part of the prefrontal cortex seems to follow a simifw time course as those from the occipital cortex which are retracted during the second postnatal week.

Conclusions These data on the structural development of the prefrontal cortex in the rat show that the prefron-

Fig. 10. Transverse section through the frontal pole of a 35-day-

old rat, which received an FB-injection in the spinal cord at P5, showing a neonatal distribution of retrogradely labeled cells (see Fig. 7a).

181

tal association cortex develops according to a similar time schedule as do other (neo)- cortical areas such as the visual and parietal cortex. The development of the cortical layers, ingrowth of thalamic and dopaminergic fibers follow a scheme of development which is comparable to that of other cortical areas. Only the volumetric development seems to point to a delayed maturation of the prefrontal areas especially the orbital PFC. The reduction of volume which is observed after the maxima volume is attained at P24 and P30 for the medial and orbital PFC, respectively, is accompanied by a reduced potential for functional plasticity. However, this late volumetric decline and reduced functional plasticity are probably not correlated with an elimination of exuberant cortical afferents or corticospinal efferents. In the callosal and ipsilateral cortical afferent systems towards the prefrontal cortex, no such elimination seems to be present, whereas in the prefrontal CST system, where a considerable pruning of transient axons was demonstrated, this elimination of collateral projections takes place in the second postnatal week. This corresponds to the period, in which the CST collaterals are pruned from the occipital cortex, where the reduction of volume and plasticity occurs much earlier.

References Bates, C.A. and Killackey, H.P. (1984) The emergence of a discretely distributed pattern of corticospinal projection neurons. Dev. Brain Rex, 13: 265-273. Beckstead, R.M. (1979) An autoradiographic examination of corticocortical- and subcortical projections of the mediodorsal projection (prefrontal) cortex in the rat. J. Comp. Neurol., 184: 43 - 62. Chun, J.J.M., Nakamura, M.J. and Shatz, C.J. (1987) Transient cells of the developing mammalian telencephalon are peptide-immunoreactive neurons. Nature, 325: 617 - 620. Crandal, J.E. and Caviness, V.S. (1984) Axon strata of the cerebral wall in embryonic mice. Dev. Brain Rex, 14: 185 - 195.

Dehay, C. and Kennedy, H. (1984) Transient projections from the fronto-parietal and temporal cortex to areas 17, 18 and 19 in the kitten. Exp. Bruin Res., 57: 208-212. Diamond, A. and Goldman-Rakic, P.S. (1989) Comparison of

human infants and rhesus monkeys on Piaget’s AB task: Evidence for dependence of dorsolateral prefrontal cortex. Exp. Bruin Res., 74: 261 - 268. Eayrs, J.T. and Goodhead, B. (1959) Postnatal development of the cerebral cortex in the rat. J. Anut., 93: 385-402. Finger, S. and Amli, C.R. (1985) Brain damage and neuroplasticity: Mechanisms of recovery or development Bruin Res. Rev., 10: 177- 186. Flechsig, P. (1901) Developmental (myelogenetic) localisation of the cerebral cortex in the human subject. Lancet, 2: 1027- 1029.

Floeter, M.K. and Jones, E.G. (1985) The morphology and phased outgrowth of callosal axons in the fetal rat. Dev. Brain Res., 22: 7 - 18. Fritschy, J.M. and Carey, L.J. (1986) Quantitative changes in morphological parameters in the developing visual cortex of the marmoset monkey. Dev. Brain Res., 29: 173 - 188. Goldman, P.S. (1971) Functional development of the prefrontal cortex in early life and the problem of neuronal plasticity. Exp. Neurol., 32: 366 - 387. Goldman, P.S. (1972) Developmental determinants of cortical plasticity. Acta Neurobiol. Exp., 32: 495 - 5 11. Goldman-Rakic, P.S. and Galkin, Th.W. (1978) Prenatal removal of frontal association cortex in the fetal rhesus monkey: Anatomical and functional consequences in postnatal life. Bruin Res., 152: 451-485. Hankin, M.H. and Silver, J. (1986) Mechanisms in axonal guidance. The problem of intersecting fiber systems. In: L.W. Browder (Ed.), Developmental Biology, Vol. 2 , Plenum Press, New York, pp. 565 - 604. Heumann, D., Leuba, G. And Rabinowicz, Th (1976) Postnatal development of the mouse cerebral neocortex. 11. Quantitative cytoarchitectonics of visual and auditory areas. J. Hirnforsch., 18: 483 - 500. Hicks, S.P. and D’Amato, C.J. (1977) Locating corticospinal neurons by retrograde axonal transport of horseradish peroxidase. Exp. Neurol., 56: 410-420. Innocenti, G.M. (1981) Growth and reshaping of axons in the establishment of visual callosal connections. Science, 212: 824 - 827. Innocenti, G.M. (1986) General organization of callosal connections in the cerebral cortex. In: E.G. Jones and A. Peters (Eds.), Cerebral Cortex, Vol. 5 , Plenum, New York, pp. 291 - 353.

Ivy, G.O., Could 111, H.J. and Killackey, H.P. (1984) Variability in the distribution of callosal projection neurons in the adult rat parietal cortex. Bruin Res., 306: 53 - 61. Jeffery, G., Arzymanow, B.J. and Lieberman, A.R. (1984) Does the early exuberant retinal projection to the superior colliculus in the neonatal rat develop synaptic connections? Dev. Brain Rex, 14: 135 - 138. Joosten, E.A.J. and Van Eden, C.G. (1989) An anterograde tracer study on the development of corticospinal projections from the medial prefrontal cortex in the rat. Dev. Bruin Res.,

182 45: 313-319. Kalsbeek, A., De Bruin, J.P.C., Feenstra, M.G.P. and Uylings, H.B.M. (1990) Age-dependent effects of lesioning the mesocortical dopamine system upon prefrontal cortex morphometry and PFC-related behaviors. This Volume, Ch. 12. Killackey, H.P. and Chalupa, L.M. (1986) Ontogenetic change in the distibution of callosal projecting neurons in the postcentral gyms of the fetal rhesus monkey. J. Comp. Neurol., 244: 33 1 - 348. Kolb, B. (1984) Functions of the frontal cortex of the rat: A comparative review. Bruin Res. Rev., 8: 65 - 98. Kolb, B. and Gibb, R. (1990) Anatomical correlates of behavioural change after neonatal prefrontal lesions in rat. This Volume, Ch. 11. Kolb, B. and Nonneman, A.J. (1976) Functional development of prefrontal cortex in rats continues into adolescence. Science, 193: 335 - 336. Kolb, B. and Nonneman, A.J. (1978) Sparing of function in rats with early prefrontal cortex lesions. Bruin Res., 151: 135 - 148. Kostovii, I (1990) Structural and histochemical reorganization of the human prefrontal cortex during perinatal and postnatal life. This Volume, Ch. 10. Kostovic, I. and Molliver, M.E. (1974) A new interpretation of the laminar development of the cerebral cortex: Synaptogenesis in different layers of the neopallium in the human fetus. Anut. Rec., 178: 395. KostoviC, I and Rakic, P. (1980) Cytology and time of origin of interstitial neurons in the white matter in infant and adult human and monkey telencephalon. J. Neurocytol., 9: 219- 242. Krettek, J.E. and Price, J.L. (1977) The cortical projections of the mediodorsal nucleus and adjacent thalamic nuclei in the rat. J. Comp. Neurol., 171: 157-192. Leonard, C.M. (1969) The prefrontal cortex of the rat. 1. Cortical projection of the mediodorsal nucleus. 11. Efferent connections. Bruin Res., 12: 321 -343. Leong, S.K. (1983) Localizing the corticospinal neurons in neonatal, developing and mature albino rat. Bruin Res., 265: 1-9. Leuba, G., Heumann, D. and Rabinowicz, Th. (1976) Postnatal development of the mouse cerebral neocortex. 11. Quantitative cytoarchitectonics of some motor and sensory areas. J. Hirnforsch., 18: 461 -481. Lund, R.D. and Mustari, M.J. (1977) Development of the geniculocortical pathways in rats. J. Comp. Neurol., 173: 289 - 306. Miller, M.W. (1987) The origin of corticospinal projection neurons in rat. Exp. Bruin Res., 67: 339-351. Mrzljak, L., Uylings, H.B.M., Van Eden, C.G. and JudlS, M. (1990) Neuronal development in human prefrontal cortex in prenatal and postnatal stages. This Volume, Ch. 9. Nonneman, A.J. and Corwin, J.V. (1981) Differential effects of prefrontal cortex ablation in neonatal, juvenile, and

young adult rats. J. Comp. Physiol. Psychol. ,95: 588 - 602. O’Leary, D.D.M. and Stanfield, B.B. (1985) Occipital cortical neurons with transient pyramidal tract axons extend and maintain collaterals to subcortical but not intracortical targets. Bruin Res., 336: 326-333. O’Leary, D.D.M. and Stanfield, B.B. (1986) A transient pyramidal tract projection from the visual cortex in the hamster and its removal by selective collateral elimination. Dev. Bruin Res., 27: 87-99. O’Leary, D.D.M., Stanfield, B.B. and Cowan, W.M. (1981) Evidence that the early postnatal restriction of the callosal projection is due to the elimination of axonal collaterals rather than to the death of neurons. Dev. Bruin Res., 1: 607 - 617. Olivarria, J. and Van Sluyters, R.C. (1985) Organization and postnatal development of callosal connections in the visual cortex of the rat. J. Comp. Neurol., 239: 1 - 26. Parnavelas, J.G. and Uylings, H.B.M. (1980) The growth of nonpyramidal neurons in the visual cortex of the rat: a morphometric study. Bruin Res., 193: 373 - 382. Provis, J.M., Van Driel, D., Billson, F.A. and Russell, P. (1985) Human fetal optic nerve: Overproduction and elimination of retinal axons during development. J. Comp. Neurol., 238: 92 - 100. Raedler, A. and Severs, J. (1975) The development of the visual system of the albino rat. Adv. Anut. Embryo/. Cell Biol., 50 (Fasc.3): 5 - 88. Rice, F.L. and Van der Loos, H. (1977) Development of the barrels and barrelfield in the somatosensory cortex of the mouse. J. Comp. Neurol., 171: 545-560, Schleyer, D.J. and Jones, E.H.G. (1988) Topographic sequence of outgrowth of corticospinal axons in the rat: a study using retrograde axonal labeling with Fast Blue. Dev. Bruin Res., 38: 89- 101. Schwartz, M.L. and Goldman-Rakic, P.S. (1982) Single cortical neurons have axon collaterals to ipsilateral and contralateral cortex in fetal and adult primates. Nature, 299: 154- 156. Sherwood, N.M. and Timiras, P.S. (1970) A Siereotuxic Atlas of the Developing Rut Bruin, University of California Press, Berkeley, CA. Stanfield, B.B. and O’Leary, D.D.M. (1985) The transient corticospinal projection from the occipital cortex during the postnatal development of the rat. J. Comp. Neurol., 238: 236 - 248. Stanfield, B.B., O’Leary, D.D.M. and Fricks, C. (1982) Selective collateral elimination in early postnatal development restricts cortical distribution of rat pyramidal tract neurons. Nature, 298: 371 -373. Uylings, H.B.M., Van Eden, C.G. and Hofman, M.A.H. (1984) Morphometry of size/volume variables and comparison of their bivariate relations in the nervous system under different conditions. J. Neurosci. Meth., 18: 19- 37. Uylings, H.B.M., Van Eden, C.G., Parnavelas, J.G. and

183 Kalsbeek, A. (1990) The pre- and postnatal development of rat cerebral cortex. In: B. Kolb and R. Tees (Eds.), The Cerebral Cortex of the Rat, MIT Press, Cambridge. Valverde, F. and Facal-Valverde, M.V. (1987) Transitory population of cells in the temporal cortex of kittens. Dev. Brain Res., 32: 283 - 288. Van Eden, C.G. (1986) Development of connections between the mediodorsal nucleus of the thalamus and the prefrontal cortex in the rat. J. Comp. Neurol., 244: 349- 359. Van Eden, C.G. and Uylings, H.B.M. (1985a) Cytoarchitectonic development of the prefrontal cortex in the rat. J. Comp. Neurol., 241: 253 - 267. Van Eden, C.G. and Uylings, H.B.M. (1985b) Postnatal volumetric development of the prefrontal cortex in the rat. J. Comp. Neurol., 241: 268-215. Verwer, R.W.H. and van Vulpen, E.H.S. (1989) Development of prefrontal cortex innervation from the amygdala in the rat. Proc. 16th Summerschool Brain Res., N.I.B.R., Amsterdam, p. 84. Wahlsten, D. (1981) Prenatal schedule of appearance of mouse brain commissures. Dev. Brain Res., 1: 461-473. Werner, L., Himmel, W. and Lohman, W. (1981) Postnatale Volumenentwicklung der Neurosomata in dorsalen Anteil des Corpus geniculaturn laterale (Cgld) und in der Area striata der albino Ratte. In: K. Hecht, W. Ruediger, K. Seidel and M. Poppei (Eds.) Zentralnervensystem, Entwickelung, Stoerungen und Motivation, Deutches Verlag, Berlin, pp. 34 - 89. Wise, S.P. and Jones, E.G. (1977) Cells of origin and terminal distribution of descending projections of the rat somatic sensory cortex. J. Comp. Neurol., 175: 129- 158. Wise, S.P. and Jones, E.G. (1978) Developmental studies of thalamocortical and commissural connections in the rat somatic sensory cortex. J. Comp. Neurol., 178: 187 - 208. Yakovlev, P.I. and Lecours, A.-R. (1967) The myelogenetic cycles of regional maturation of the brain. In: A. Minkowski (Ed.), Development of the Brain in Early Life, Blackwell, Oxford, pp. 3-70. Zilles, K.J. (1978) Ontogenesis of the visual system. Adv. Anat. Embryol. Cell Biol., 54 (Fasc.3): 5 - 138. Zijlstra, C., Verwer, R.W.H. and Van Eden, C.G. (1986) Comparison of numerical density of neurons before, during and after the volume overshoot of the medial prefrontal cortex. Research report Neth. Inst. Brain Res. 86- 1. (In Dutch.)

Discussion P. Goldman-Rakic: Do you think that different principles govern development of callosal and corticospinal projections, or between prefrontal callosal as opposed to visual callosal neurons? C.G. van Eden: From the work that I presented here today, I have no indications that callosal PFC fibers project to other areas in the neonatal period than in adulthood. This does not exclude a developmental reduction of callosal fibers within these areas. So far, attempts to use fluorescent tracers for a double labeling experiment separated in time in order to investigate such a developmental reduction in callosal PFC projections have failed because the form of the PFC changes markedly during neonatal development. This makes a second, overlapping injection with the second tracer almost impossible. Therefore, I do not know whether the development of callosal and CS projections are governed by a different principle. But when the mechanisms in the rat are comparable with those in the monkey and cat, it might be that a pruning of collateral axons occurs in the CS fiber systems and in the visual callosal system but not in the PFC callosal system (e.g., O’Leary and Stanfield, 1985; Innocenti, 1986; Schwartz and GoldmanRakic, 1984). R.W.H. Venver: I would like to comment on the idea to make a first injection and later subsequent injection in exactly the same place. The first injection will damage the cortex and it is afterwards not comparable with a normal cortex at the same age. C.G. van Eden: That is true; nevertheless it has been proved that it is possible to double label many cells in this way, indicating that necrosis is not preventing uptake of the second tracer in a large part of the initial injection site. I agree that a number of cells that fail to transport the second tracer may act like this because of this phenomenon. Therefore, a restriction of the projection area should also be demonstrated by other (anterograde) tracer methods.

References Innocenti, G.M. (1986) General organization of callosal connections in the cerebral cortex. In: E.G. Jones and A. Peters (Eds.), Cerebral Cortex, vo1.5, New York, pp. 291 - 353. O’Leary, D.D.M. and Stanfield, B.B. (1985) Occipital cortical neurons with transient pyramidal tract axons extend and maintain collaterals to subcortical but not intracortical targets. Brain Res., 336: 326 - 333. Schwartz, M.L. and Goldman-Rakic, P.S. (1984) Callosal and intrahemispheric connectivity of the prefrontal association cortex in rhesus monkey: relation between intraparietal and principal sulcal cortex. J. Comp. Neurol., 226: 403 - 420.