THE JOURNAL OF COMPARATIVE NEUROLOGY 291:128-146 (lYY0)

Maturation of the Corpus Callosum of the Rat: I. Influence of Thyroid Hormones on the Topography of Callosal Projections CLAUDE GRAVEL ANDRICHARDHAWKES Department of Biochemistry and Laboratory of Neurobiology, Faculty of Medicine, Lava1 University, Quebec, Canada

ABSTRACT The normal adult rat corpus callosum contains numerous axonal profiles that are immunoreactive for the high molecular weight subunit of the neurofilament triplet (NF-H). NF-H immunoreactivity develops gradually during the first 2 postnatal weeks. The expression of NF-H immunoreactivity is almost completely suppressed in rats rendered hypothyroid by neonatal treatment with propylthiouracil. T o ensure that the cytoskeletal deficit was due to a shortage of thyroid hormones rather than to unspecific, toxic effects of propylthiouracil, hypothyroid animals kept on the propylthiouracil diet were given restorative thyroxine injections daily. Such animals express NF-H a t normal levels. This suggests that the callosal axons may be arrested at an immature stage of development. The immaturity of the hypothyroid corpus callosum can also be revealed by a comparison of the myelin content in the corpus callosum between normal rats, hypothyroid rats, and hypothyroid rats under thyroxine therapy. Hypothyroid rats are severely deficient in myelin, and again this deficit can be corrected by postnatal thyroxine treatment. During normal callosal development, there is a progressive spatial restriction of the transcallosal projection that creates in the adult patches of callosally projecting cortex interposed by acallosal regions. Given the structural immaturity of the hypothyroid callosal axons, it was interesting to investigate the state of development of their topography. For this purpose, multiple injections of wheat germ agglutinin-horseradish peroxidase were made into the occipital and parietal cortices of adult hypothyroid animals. In normal rats, the majority of visual callosally projecting cells are located in three groups-in area 18b, at the boundary of area 17 and 18a, and in the lateral portion of area 18a. Within these areas projecting cells are concentrated in layers 11-111,Va, and Vc-VIa. The callosal axon terminals are concentrated in these same regions, with a laminar distribution as far as the somata plus layer I. In the midportion of areas 17 and 18a, fewer callosal cells are found. and they occupy mainly layers Vc-VIa, as is the case for terminals in these same areas. In the parietal cortex, callosal cells and terminals are disposed in vertical arrays alternating with almost empty zones. Most are concentrated in layers IT1 and V. The topography of the callosal axon terminal fields is unaffected by hypothyroidism. However, there is a dramatic redistribution of the callosally projecting cell somata. Although these adopt the same radial distribution as seen in the normal animals, they are now arranged in continuous tangential laminae throughout both visual and somatosensory areas, and the normal acallosal patches are absent. This arrangement resembles the situation in the early neonate where the somata of callosally projecting cells are also arranged in continuous tangential laminae. The progressive restriction of callosal pro-

0 1990 WILEY-LISS, INC.

Accepted August 9,1989 Address reprint requests to Dr. R.B. Hawkes, Department of Anatomy, University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta T2N4N1, Canada.

129

MATURATION OF THE CORPUS CALLOSUM

jection domains has been thought to occur through selective axon elimination. It therefore appears that hypothyroid animals retain inappropriate immature callosal projections due to the thyroid hormone dependence of the normal elimination mechanisms. Key words: monoclonal antibody, axon elimination, synaptogenesis

Numerous investigations have shown a role for thyroid hormones in the regulation of neuritic growth (for a review on mammals, see Legrand, '83), but little is known concerning thyroid hormones and the establishment of afferent topography. In amphibians, thyroxine has been shown to play a role in axon elimination in the oculomotor nerve and in the appearance of the ipsilateral component of the retinofugal pathway (Hoskins and Grobstein, '85; Schonenberger and Escher, '88). In rats, the stabilization of mature neuronal circuitry often coincides with the expression of the 210 kDa subunit of the neurofilament triplet (NF-H) (Carden et al., '87). By using t h e monoclonal antibody N210 (mabN210) that recognizes an epitope on NF-H, we have previously studied neurofilament expression during postnatal development in rats made hypothyroid from embryonic life (Leclerc et al., '85; Plioplys et al., '86; Gravel and Hawkes, '87a,b). Immunocytochemistry revealed that axons that normally express the mabN210 epitope fall into two broad classes: those in which expression is suppressed by hypothyroidism and those that are insensitive to the thyroid status of the animal. Interestingly, the thyroxine dependency of NF-H expression seems to be correlated with the time the axon reaches its target, with those projections that mature before serum thyroid hormone concentrations reach significant levels (around postnatal day 4 in a rat; see Dussault and Walker '78) being insensitive to the thyroid status of the animal, and those maturing later being sensitive (Leclerc et al., '85; Plioplys et al., '86; Gravel and Hawkes, '87a,b). One thyroid hormone-sensitive tract in the rat is the corpus callosum where neonatal treatment with propylthiouracil (PTU), a synthetic antithyroid compound (Oppenheimer et al., '72), almost completely suppresses the normal postnatal expression of NF-H (Plioplys et al., '86). In placental mammals, the corpus callosum constitutes the main commissural pathway between cortical neurons of the two hemispheres, and its maturation has proved an important model to develop and test theories concerning the development of afferent topography. A crucial feature of the adult callosal projections is that some regions, such as the inferior parietal lobule, have extensive interhemispheric connections, whereas others, such as the visual cortex, have few. The important feature for developmental studies is that in immature animals callosal projections are more widespread: for example, in adult cats callosal connections originate and terminate only near the frontier between areas 17 and 18, whereas in newborn kittens they originate throughout these areas (Innocenti et al., '77). The progressive restriction of callosal connections during development appears to be a general phenomenon that affects many cortical areas in numerous animal species (Innocenti and Caminiti, '80; Chow et al., '81; Ivy and Killackey, '81,82; O'Leary et al., '81; Feng and Brugge, '83). Studies of the cat visual cortex (Innocenti et al., '77; Innocenti, '81) and of the rat somatosensory cortex (O'Leary et al., '81; Ivy and Killackey, '82) have shown that the developmental mechanisms responsible for

restricting the distribution of the callosally projecting neurons during early postnatal development involve the elimination of a set of transitory callosal axon collaterals. Significant neuronal cell death is not involved. In the rat, the corpus callosum matures during the first 2 postnatal weeks. On the day of birth (PO), corticocortical axons have crossed the midline and are queueing in the subcortical white matter but have not yet entered the cortical plate (Wise and Jones, '76; Ivy and Killackey, '81; Valentino and Jones, '82). Invasion of the cortical plate begins a t P3, and the mature adult distribution is attained by the end of the second week in both the somatosensory (Wise and Jones, '76; Ivy e t al., '79; Ivy and Killackey, '81) and the visual cortices (Lund et al., '84; Miller and Vogt, '84; Olavarria and Van Sluyters, '85). During this same period, the distribution of callosally projecting neurons undergoes extensive changes: until P5, callosally projecting somata are confined to two continuous horizontal bands deep to the cortical plate, but as development proceeds there is a noticeable decrease in the number of callosally projecting cells in those areas destined not to be callosally connected in the adult. As a result, by the end of the second postnatal week, the barrelfield area of the parietal cortex and the medial portions of area 17 and 18a of the visual cortex contain very few callosally projecting neurons (Ivy and Killackey, '81; Olavarria and Van Sluyters, '85). The almost complete suppression of N210 immunoreactivity in the corpus callosum of P25 hypothyroid rats (Plioplys et al., '86), together with the temporal congruence of the postnatal reshaping of the callosal organization and the increase of thyroxine serum concentration in normal animals, has encouraged us to explore the relationships between thyroid hormones, axon maturation and the development of corticocortical topography. In this article, we examine three main issues. First, is the blockade in NF-H expression that was seen in hypothyroid animals due specifically to a shortage of thyroid hormones or merely to some unspecific effect of the PTU administration? This is addressed by using NF-H immunocytochemistry to examine thyroxine-treated animals otherwise kept on a P T U diet. Second, given the absence of NF-H immunoreactivity, what is the state of myelination of callosal axons? This question has been approached by light microscopic observation of the hypothyroid corpus callosum. Third, is the failure of the cytoskeletal maturation associated with immaturity of the callosal topography? T o answer this last question, we have used tracer injections to study the distribution of callosal terminals and callosally projecting somata in the adult hypothyroid rat.

MATERIALS AND METHODS Experimental hypothyroidism Sprague-Dawley rats were rendered hypothyroid by inclusion of 0.05 % w/v 6n-propyl-2-thiouracil (PTU; Sigma Inc.) and 10% w/v sucrose in the maternal drinking water

130

C. GRAVEL AND R. HAWKES

(Dussault and Walker, '78). Sucrose is required to make the solution palatable. Treatment commenced prenatally a t the 17th day of gestation (E17; the mating day being considered EO) and was continued throughout postnatal life. At birth, the pups of different hypothyroid litters of the same age were mixed and the families rounded down, as necessary, to 8 pups per mother. The pups remained with their mothers throughout the treatment period. P T U reliably produces hypothyroidism in our experimental animals (Leclerc et al., '85; Plioplys e t al., '86; Gravel and Hawkes, '87a,b); the effectiveness of the treatment was witnessed by such overt signs as reduced body size and weight, motor deficits, and scarcity of fur. The hypothyroid condition was corrected by the subcutaneous injection of 0.035 to 0.2 p g thyroxine/day from birth to P13 and 1pg/lOO g body weight/day thereafter.

Immunocytochemistry For immunocytochemistry, rats were deeply anesthetized with sodium pentobarbital and after surgical exposure of the heart were perfused via the ascending aorta with icecold 4 % paraformaldehyde, 0.2 % glutaraldehyde in phosphate buffer (PB, 0.1 M, pH 7.4) in volumes ranging from 25 to 100 ml depending on the size of the animal. Subsequently, the brain was removed from the cranium and postfixed overnight a t 4OC in 4 57 paraformaldehyde in P13.Fifty pm frontal sections of the brain were cut on a freezing-stage microtome and selected sections incubated overnight in the spent culture medium of the mabNZlO hybridorna secreting line (Leclerc et al., '85) diluted 1:s in 10% normal horse serum in PBS (PB + 0.9% NaC1). T o detect specific antibody binding, sections were subsequently incubated for 2 hours in rabbit antimouse immunoglobulins conjugated to horseradish peroxidase (Dako Inc.) diluted 1:lOO in 10% horse serum. Antibody binding was revealed by using 4-chloro-1-naphthol as substrate (Hawkes et al., '82). The sections were washed in several changes of PBS between incubations. Incubations from which the primary antibody was omitted or replaced by myeloma-conditioned medium gave no staining. Other sections were processed according to the Weigert method to reveal the presence of myelin.

Tracing the callosal projections of the visual cortex T o reveal the callosal projections of the visual cortex, P25 and P35 hypothyroid animals, P25 normal animals, and P35 hypothyroid animals given a regime of thyroxine therapy but otherwise kept on the PTV diet were anesthetized by intraperitoneal injection of pentobarbital (65 mg/kg) and ketamine hydrochloride (10 mg/kg), and the head secured in a stereotaxic apparatus. An opening was made in the left posterior part of the skull and 10 to 15pressure injections of a 3 % WGA-HRP solution in saline (50 nl each) were delivered under visual guidance into the primary and secondary visual cortices with a glass micropipette (tip diameter about 75 pm) fitted to a 1 p1 Hamilton syringe. In one P30 normal animal, 16 injections were made into the parietal cortex following the same protocol. The Zilles ('85) stereotaxic atlas of the rat cortex was used to locate the cortical areas. After a survival time of 48 hours, the animals were deeply anesthetized with pentobarbital and perfused via the ascending aorta with 50 to 100 ml of saline (0.9% NaC1) followed by 100 to 250 ml of an ice-cold fixative containing 1%paraformaldehyde, 1.25% glutaraldehyde in PB. The brain was removed and stored overnight in PBS a t 4OC. To reveal the

presence of HRP, 50 pm alternate sections were processed for H R P histochemistry, using tetramethylbenzidine (TMB) as substrate (Mesulam, '78). Every other section was counterstained with either cresyl violet or neutral red. The ages of the animals given in the text refer to the time of the tracer injections. Primary visual cortical area 17 and secondary areas 18a and 18b were defined according to standard criteria (Krieg, '46; Caviness, '75) and correspond to the ocl, oc2L, and oc2M, respectively, of Zilles ('85).

RESULTS Immunocytochemistry The mabN210 antibody recognizes an epitope associated with the 210kD subunit of the neurofilament triplet, NF-H (Leclerc et al., '85). Immunocytochemistry of adult rat brain reveals that the epitope is strictly associated with axons and that glial cells, neuronal perikarya, and dendrites are devoid of the antigen (Leclerc et al., '85, Plioplys e t al., '86, Gravel and Hawkes, '87a,b). Axons in the corpus callosum normally express NF-H (Plioplys et al., '86). In previous studies we reported that the expression of the mabN210 epitope is blocked in selected axonal tracts in animals rendered hypothyroid by treatment from 17 days of gestation (E17) with the antithyroid compound P T U (Leclerc et al., '85; Plioplys et al., '86; Gravel and Hawkes, '87a,b). In general, it seems that axons that have attained their mature distribution before serum thyroid hormones reach appreciable levels are not affected by hypothyroidism, whereas expression of the mabN210 epitope is suppressed in axons that develop later. Although callosal axons have crossed the midline by the time of birth (Valentino and Jones, '82; Floeter and Jones, '85), they do not start to invade the cortical gray matter until late in P3 when they are restricted to the deepest region of layer VI (Ivy and Killackey, '81; Olavarria and Van Sluyters, '85). In both visual and somatic areas, they penetrate farther into the cortex during the next few days until the superficial layers (I1and I) are attained by P8-P9 (Ivy and Killackey, '81; Miller and Vogt, '84; Olavarria and Van Sluyters, '85). This late developmental profile would then classify callosal axons in the putative thyroid-sensitive category and, indeed, there is a dramatic shortage of NF-H immunoreactivity in the corpus callosum of P25 hypothyroid animals as compared to P25 controls (Fig. lA,B): the NF-H epitope is almost completely absent throughout the rostrocaudal extent of the corpus callosum and remains so a t least until P60 (not shown). To test whether this blockade might be due to unspecific he., nonantithyroid) toxicity of the PTU treatment, a regime of thyroxine therapy was given simultaneously from P30 to P60 to hypothyroid animals otherwise kept on the PTU diet. This resultcd in a significant release of the blockade in NF-H antigen expression (Fig. IC). Sections processed so as to reveal the presence of myelin show a drastic reduction of myelin content in the hypothyroid corpus callosum, but a normal content in hypothyroid animals given thyroxine therapy (not shown, but see Gravel et al., '90).

Tracer experiments The next step was to investigate whether the blockade in NF-H expression and the low myelination in the hypothyroid corpus callosum were correlated with immature and/or abnormal callosal projections. For this, advantage was taken

131

MATURATION OF THE CORPUS CALLOSUM

Fig. 1. Fifty-pm frontal sections through the brain, taken about 7 mm anterior to the interaural line (Paxinos and Watson, '85) from norma1 (A) and hypothyroid (B) P25 rats, and from a P60 PTU-treated rat that received restorative thyroxine treatment from P30 (C). Sections were immunoperoxidase-stainedwith mabN2lO to reveal the presence of the high molecular weight neurofilament subunit (NF-HI. Note the

intense labelling of callosal fibers running in the horizontal plane in the normal P25 animal (A). In the hypothyroid animal, the corpus callosum is almost devoid of immunoreactivity (B). Thirty days of thyroxine treatment restores immunoreactivity to about normal levels in an animal otherwise kept on the PTU regime (C).In each case, dorsal is towards the top and the scale bar 50 pm.

-

132

C. GRAVEL AND R. HAWKES

Fig. 2. Darkfield photomicrograph of a 50-pm-thick frontal section through the occipital cortex of a normal P27 rat that received multiple injections of WGA-HRP into the left visual cortex a t P25. Refractile TMB precipitates revealing the presence of HRY appear as white deposits against the dark background. Ipsilaterally, the tracer was trans-

ported to the superior colliculus (sc). There is also labelling in the cerebral peduncle (cp) owing to the spread of the injection site to snmatosensory cortex. Contralaterally, HRP deposits are concentrated in alternating patches of high and low density in the right visual cortex. Scale bar: 1 pm.

of the fact that the topography of both the occipital and parietal callosal projections of the normal rat is well known and undergoes extensive modifications during postnatal development (Wise and Jones, '76, '78; Ivy et al., '79; Ivy and Killackey, '81; Lund et al., '84; Miller and Vogt, '84; Olavarria and Van Sluyters, '85). The following data are based on 24 cases of WGA-HRP injections into the left visual cortex of young rats. Injections were judged as successful by the analysis of injection sites and by anterograde and retrograde transport of the tracer into the contralateral hemisphere and into subcortical structures. In all 7 P25 normal rats, 9 P25 hypothyroid rats, 4 P35 hypothyroid rats, and 4 P35 hypothydroid rats given a regime of thyroxine therapy from birth but otherwise kept on PTU diet were studied. In most cases the injection site also extended into the parietal cortex, resulting in both retrograde and anterograde transport into the opposite parietal cortex. In one P30 normal animal, the injection site was centered directly on the left parietal cortex. The pattern of both occipital and parietal callosal connections in normal rats displays all the major features of the adult projections from the second postnatal week onwards (Wise and Jones, '76; Ivy and Killackey, '81: Lund et

al., '84; Miller and Vogt, '84; Olavarria and Van Sluyters, '85), and so P25 and P30 normal rats can be considered adequate controls for both groups of hypothyroid animals.

Callosal projections of the occipital cortex Normal animals. Figure 2 exemplifies the anterograde and retrograde transport of WGA-HRP in a P25 control animal following multiple tracer injections that covered the visual areas of the right occipital cortex. Forty-eight hours after tracer injections, HRP had been transported into the contralateral hemisphere, the ipsilateral superior colliculus, the optic tract, the ipsilateral lateral geniculate nucleus (not shown), and various other structures known to be connected with the visual cortex. Labelled fibers are also apparent in the cerebral peduncle a t this level as the injection site extends into the parietal cortex. A higher magnification view of the right (contralateral) occipital cortex (Fig. 3a) shows labeled axon bundles coursing in the white matter, and retrogradely filled neurons and anterogradely filled axon terminals in the gray matter. A comparison of a section stained for HRP activity with the adjacent one stained with neutral red (Fig. 3b) reveals

MATURATION OF THE: CORPUS CALLOSUM

133

Fig. 3. Brightfield photomicrographs of two 50+m consecutive frontal sections through the right occipital cortex of the same animal as in Figure 2. The section in A was processed for HRP; the section in B was stained with neutral red to reveal the cytoarchitectonic boundaries of area 17 (arrowheads in both A and B). Visual areas are numbered. From medial (left) to lateral, the regions containing the highest density of both anterogradely and retrogradely transported tracer (dark deposits

in A) are found in area lab, lateral area 17-medial area Ma, and lateral area Ma. In the midportion of area 17 and 18a, retrogradely labelled perikarya are mainly restricted to the lower third of the cortex. Sparse extraperikaryal deposits are codistributed with the labelled somata. Arrows in A and B point to the same blood vessels in the two sections. Scale bar: 500 wm.

the characteristic distribution of labelled somata and terminals with respect to the cytoarchitectonic boundaries of primary (area 17) and secondary (area 18a and 18b) visual cortices. In area 17 most of the labelled cell bodies and terminals are coextensive, mainly occupying the lateral third,

with a heavy radial band of anterogradely transported material straddling the cytoarchitectonic boundary with area 18a. The retrogradely labelled cells are especially numerous in layers 11-111, Va, and Vc-VIa, with anterograde labelling also concentrated a t these layers and in the lower part of

134 layer I (Fig. 4a,b). In the medial two-thirds of area 17 (Fig. 4c,d), retrogradely labelled cells are less numerous and are concentrated as a tangential band in layer Vc-VIa, with a few scattered in the more superficial layers. The density of extraperikaryal deposits found in this region is low and restricted to the lowest part of the cortex (layers Vc and VI). This distribution of callosal projections in the body of area 17 (and also in the body of area 18a; see below) accords with previous studies in rodents (Olavarria and Van Sluyters, ’83, 85; Rhoades et al., ’87). At the junction of area 17 and area 18b both the anterograde and retrograde staining of supragranular layers is contained within area 18b (Fig. 3). In area H a , lateral to the frontier with area 17, a “weakly callosal” stretch of cortex is also found, with the same characteristics already described for medial area 17: retrogradely-labelled cells mainly restricted to layers Vc-VIa with very low anterograde staining in layer VI. This last zone is bordered at its lateral limit by a heavy column of anterograde and retrograde staining, narrower than that at the 17-18a junction, but with the somata and terminals in the same radial distribution. These results confirm published data on the topography of visual callosal projections in rats (Cusick and Lund, ’81; Olavarria and Van Sluyters, ’83, ’85; Lund et al., ’84; Miller and Vogt, ’84; Olavarria et al., ’87). Hypothyroid animals. Tracer injections into the visual cortex of hypothyroid rats (either P25 or P35) resulted in both anterograde and retrograde transport of the tracer into the opposite hemisphere. Tracer was also transported t o subcortical structures, including the ipsilateral superior colliculus, ipsilateral lateral geniculate nucleus, and the optic tract. Although the hypothyroid cortex is slightly thinner than the normal, layers I to V1 can still be distinguished, and the cytoarchitectonic boundaries of the primary and secondary visual cortices are clear. Thus callosal projections in hypothyroid animals can be related to the same landmarks used for normals animals. The pattern of visual callosal connections seen in both groups of hypothyroid animals is illustrated in Figure 5 for a P35 animal. In the zones of the visual cortex corresponding to the “strongly callosal” regions in the normal animal (i.e., the lateral part of area 18b, the lateral third of area 17, and both the more medial and more lateral regions of area 18a) the anterograde and retrograde staining is very similar to that in normal animals, both in tangential extent and radial distribution. At the border between area 18b and 17, the anterograde staining extends up to the lower part of layer I. Tangentially, it is confined within the limits of area 18b. In the lateral part of area 17 and the medial part of area 18a, anterograde staining of axon terminals is seen in all layers, with some emphasis on I, 11.111, Va, Vc, and Vla. Retrogradely labelled neurons are concentrated in those same layers, except for layer I (Fig. 6a,b). As in normal rats, the heaviest concentration of anterogradely labelled terminals directly straddles the limits of 17 and 18a. As in controls, a relatively narrow column of anterogradely labelled material is present in the lateral part of area 18a, and in this same stretch of cortex retrogradely labelled cells are present, concentrated in layers 11-111, Va, and Vc-VIa. In hypothyroid animals the retrogradely labeled cells are pyramidal in shape as in the normal cases but appear to be more numerous (see below). Significant, reproducible differences between the callosal projection patterns of hypothyroid and normal animals are found when the medial two-thirds of area 17 and midarea

C. GRAVEL AND R. HAWKES

18a are considered. In area 17, three tangential bands of retrogradely labelled cells are present in both P25 and P35 hypothyroid animals, situated in layers 11-111, Va, and VcVIa (Fig. 6c,d) in continuity with the bands found in area 18b and the lateral part of area 17 to either side. This is in contrast to the single band found in layer Vc-Vla of this region in normal animals, where only the occasional labelled cell is found in the more superficial layers. Occasional retrogradely labelled cells in layer Va and supragranular layers of medial area 17 are also reported by Olavarria and Van Sluyters (’83, ’851,but these authors do not refer t o them as forming layers. These “supernumerary” layers of callosally projecting cells in layers 11-111 and Va were observed consistently in all our hypothyroid animals and, in cases where all contralateral visual areas were saturated with the tracer, are continuous from section to section throughout area 17. The extraperikaryal staining in this same region is lower than in area 18b or lat.era1 area 17 on each side and is restricted to the lower part of the cortex (layers V and VI) as in the normal animals (Fig. 6c,d). Area 18a of the left hemisphere was saturated with WGAHRP in three P35 hypothyroid animals belonging to two different litters. In each case, a trilaminar arrangement of retrogradely labelled cells was observed in midarea 18a of the coatralateral hemisphere (Figs. 5, 7; see also Fig, 10). The bands of labelled cells occupy layers 11-111, Va, and VcVIa and are continuous with those found in the “normal” callosal zones to each side. This contrasts with normal animals where most retrogradely labelled cells are concentrated in layers Vc-VIa, and only scattered cells are found more superficially (see above, and Olavarria and Van Sluyters, ’83). Based on tangential sections of the visual cortex, Olavarria and Van Sluyters (’85) and Olavarria et al. (’87) have described several narrow bands of labelling that bridge area 18a at different anteroposterior levels. In these callosal “bridges,” the distribution of both anterogradely and retrogradely transported material is similar to that found at the 17-18a border and in the lateral part of 18a. However, there are three reasons to believe that the distribution of labelled profiles in hypothyroid animals is abnormal in area 18a. First, the trilaminar organization of retrogradely labelled cells found in midarea 18a can be followed from section to section through most of the anteroposterior extent of area Ma, and shows no discontinuities of the type that would be expected for a multibridge organization (Fig. 8). Second, the laminar distribution of anterogradely transported material is different from that found in callosal bridges: the extraperikaryal staining is much lower in midarea 18a than in medial or lateral area 18a and is restricted to the infragranular layers. Third, we have identified in our hypothyroid material what we believe is the equivalent to these “bridges.” As seen in coronal sections, a bridge consists of a region of area 18a where both the anterograde and retrograde staining is tangentially continuous from lateral area 17 to lateral area 18a (Fig. 7). The bridge illustrated in Figure 7 is a few hundred microns thick in the rostrocaudal direction and is situated in the posterior region of the visual cortex, and thus is clearly different from the remainder of area 18a. Although callosal cells are arranged in three continuous tangential sheets all over the visual cortex of hypothyroid rats, their packing density probably varies between different visual areas. This can be seen in camera lucida drawings of serial sections through the visual cortex of an hypothyroid animal (Fig. 8). In all sheets of callosal cells, the

MATURATION OF THE CORPUS CALLOSUM

Figure 4 (Caption on Overleaf)

135

136

C . GRAVEL AND R. HAWKES

Fig. 5. Photomicrographs of frontal sections through the right occipital cortex of a hypothyroid animal that received multiple WGAHRP injections into the left visual cortex at P35. A. Section processed according to the TMB method. B. Adjacent Nissl-stained section. In contrast to the distribution in normal animals (Fig. 3), retrogradely

labelled cells from 3 continuous horizontal laminae extending from area 18b to lateral area 18a. Anterogradely transported HRP is concentrated in area 18b, lateral area 17-medial area 18a and the lateral part of area 18a as in the controls, and is also found in the lower third of midarea 17 and 18a. The symbols are the same as in Figure 3. Scale bar: 500 Fm.

Fig. 4. Photomicrographs of frontal sections through the right occipital cortex of the same animal as Figures 2 and 3 showing the distribution of callosally transported HRP in cortical layers (roman numerals) at the area 17-area 18a border (A) and in the midportion of area 17 (C). B and D are from sections adjacent to A and B respectively that were stained with neutral red. In lateral 17-medial Ma, retrogradely labelled cell bodies are plentiful with the highest densities found in layers 11-111, Va, and Vc-VIa. Anterogradely transported material has a fine, granular

appearance and is present in all layers, with the lowest concentration in the upper half of layer I. As compared to lateral area 17, midarea 17 contains fewer callosally projecting cell bodies, with most occupying layers V and VT. Anterograde staining is scarce and is restricted to Vc and VI. Large arrowheads point to the cytoarchitectonic limits of area 17. Small arrowheads indicate the cortical layers. Arrows point to successive profiles of the same blood vessels. Scale bar: 200 pm.

Figure 6 (Caption on Overleaf)

138

C. GRAVEL AND R. HAWKES

Fig. 7. Photomicrographs of frontal sections taken through the caudal pole of the right occipital cortex of a hypothyroid rat that had received multiple WGA-HRP injections in the left visual cortex a t P35. A was processed according to the TMB method; B was Nissl-stained. Callosally projecting cells are concentrated in 3 tangential laminae throughout the visual areas. The anterograde staining occupies the

whole thickness of the cortex from lateral area 17 to lateral area 18a, although it is mainly restricted to the inferior third in midarea 17. This invasion of supragranular layers by callosal afferents in the midportion of area 18a is a callosal “bridge,” and in this particular case it extends for 300 pm along the rostrocaudal axis. Symbols as in Figure 3. Scale bar: 500 pm.

Fig. 6. Photomicrographs of the right visual cortex taken from the same animal as in Figure 5 to show the distribution of callosally transported HRP a t the areal 7-area 18a border (A) and in the midarea 17

4): retrogradely filled neurons are concentrated in layers 11-111,Va, and Vc-VIa, and anterogradely transported material is concentrated in these same layers and also in layer I. In midarea 17 (C), HRP-filled somata are less numerous than in lateral area 17-medial area 18a but share the same distribution. Anterograde staining is restricted mainly to layers Vc and VI. Symbols as in Figure 4. Scale bar: 200 pm.

(C), together with the corresponding regions from adjacent Nisslstained sections (B,D). In lateral area 17-medial area 18a (A), the HRP distribution is indistinguishable from that seen in normal animals (Fig.

139

MATUKATION OF THE CORPUS CALLOSUM

CG 44 Fig. 8. Camera lucida drawings of serial 50-pm-thick frontal sections taken a t 200-rm intervals, through the right occipital cortex of a hypothyroid rat that received multiple injections of WGA-HRP into the left visual cortex at P35. The sections were processed to reveal callosally transported HRP. Medial is to the left and rostra1 to caudal runs from top to bottom. In these drawings, each large dot represents one retro-

lmm gradely stained neuron and the small dots represent schematically the areas of high anterograde labelling (note: anterograde labelling in midarea 17 has been omitted). A trilaminar arrangement of callosally projecting cells is evident throughout the visual cortex. The density of labelled cells is lower in midarea 17 and midarea 18a.

140 labelled cell density is higher in regions where most anterogradely transported material is present-area 18b, lateral area 17-medial area 18a, and the lateral portion of area 18a. When the hypothyroid visual cortex is compared to the normal (Fig. 9), more callosally projecting cells are found in all areas. Any quantitative estimate from these experiments must be approximate owing to differences in the tracer density and the exact position of the injection sites. As a result the cell density varies considerably from animal to animal (e.g., Fig. 10). Bearing this caveat in mind, when two cases are compared in which the injection site covered both area 17 and Ma, the density of retrogradely labelled cells in midarea 17 is 12.4 times higher in the hypothyroid animal (case CG44, Fig. 8 ) than in the normal (case CG40, Fig. 9), and 1.7 times higher in lateral area 17-medial area 18a.

Callosal projections in the parietal cortex Normal animals. I n several animals, injection sites centered on visual areas were large enough to encompass also a significant part of the neighboring somatic sensory cortex. Both the callosal connections in adult rats and the normal postnatal development of this region are well documented and have many points in common with the visual cortex (Wise and Jones, '76, '78; Ivy et al., "79; Ivy and Killackey, '81, '82; O'Leary et al., '81). Therefore, we have compared the distribution of the somatic sensory callosal projections in hypothyroid rats with that in normal controls. Figure l l a shows a coronal section taken from the right parietal cortex of a normal P30 rat t,hat received multiple WGA-HRP injections that encompassed most of the left parietal cortex. In the parietal cortex contralateral to the injection site, the labelling reveals coextensive vertical arrays of callosally projecting cells and callosal afferent terminals, interposed by stretches of unlabelled cortex. These "acallosal" zones correspond to regions of the cortex that have dense aggregations of granule cells in layer IV. In callosally connected areas, retrogradely labelled cells are found mainly in layers I11 and V, with a few scattered in layers 11, IV, and VI. Callosal afferent terminals are concentrated in layers I to 111, V, and VI. Similar distributions of callosal afferents and cell bodies were found in other normal animals in which the injection site covered only part of the parietal cortex, but in these cases the retrograde labelling did not extend as far rostrocaudally. This distribution of parietal callosal projections is essentially similar to that described previously (Wise and Jones, '76; Ivy et al., '79; Ivy and Killackey, '81, '82; O'Leary et al., '81). Hgpothgroid animals. Figure I l b shows the pattern of anterograde and retrograde labelling in the right parietal cortex of a P35 hypothyroid animal following injections of WGA-HRP that covered part of the contralateral parietal cortex. Retrogradely labelled cells are in two continuous tangential bands instead of the radially oriented columns described for normal animals. A more normal radial columnar organization of the callosal afferent connections is suggested by the concentration of extraperikaryal deposits in restricted tangential stretches of cortex. The columnar organization of the afferent terminal fields is accented by a higher density of retrogradely labelled cells in the same regions. Adjacent sections stained with cresyl violet reveal that the labelled perikarya in regions displaying high anterograde labelling are densely packed in layers 111 and V, with fewer labelled cells in layers 11, IV, and VI. In regions that are virtually free of anterograde labelling and display a thick layer IV, retrogradely labelled neurons are rare out-

C. GRAVEL AND R. HAWKES

side layers I11 and V. The extraperikaryal material has a normal radial and tangential distribution, being concentrated in layers I to 111, V, and VI of regions where layer IV is thinner.

Callosal projections in PTU-treated animals under thyroxine therapy T o test whether these abnormal callosal projections might

be due to unspecific toxicity of the PTU treatment, callosal projections were investigated in 4 P35 hypothyroid animals that were given a daily regime of thyroxine therapy from PO to P35 but were otherwise kept on the PTU diet throughout that period. In these 4 animals, most of the extent of visual areas have been saturated with the tracer, and in 3 of them a significant portion of the parietal cortex is also covered by the injection site. In all 4,the pattern of callosal projections is indistinguishable from that observed in normal controls, for both visual and somatosensory cortices (not shown).

DISCUSSION In the hypothyroid rat cerebral cortex the cell somata are smaller and more densely packed, the pyramidal cells have abnormal dendritic branching patterns, and there are fewer axonal profiles (Eayrs, '71). In the visual cortex, the number of nerve terminals is reduced (Cragg, '70) and in visual and auditory cortices, there is a reduced density and an abnormal distribution of spines on the apical dendrites of pyramidal cells (Ruiz-Marcos et al., '78, '79). These morphological alterations are accompanied by profound deficits in higher mental functions concerning both innate and adaptive behaviors (Eayrs, '71). We now report a novel consequence of neonatal hypothyroidism in the rat central nervous system: an abnormal callosal projection. In hypothyroid animals callosal cells are arranged in continuous tangential layers throughout the visual and parietal cortices, in contrast to the discontinuous arrangement found in the same regions of normal animals. The callosal cells in these areas also seem to be more numerous in hypothyroid animals than in normals. In contrast, the callosal afferent terminals appear to be clustered in the same discrete tangential stretches of the parietal and visual cortices as in normal controls. These data can be interpreted in two ways: either as due to modified callosal axon sprouting in response to the shortage of thyroid hormones, or to a block in development that impedes the pruning of the callosal projection that normally occurs during the first 2 weeks of postnatal development. Excessive axon collateral sprouting would be an unusual consequence of thyroid hormone deprivation-studies both in vivo and in vitro have rather demonstrated impaired neurite development in nerve cells exposed to insufficient levels of thyroid hormones (reviewed in Legrand, '83). For this reason, we favor the hypothesis that neonatal hypothyroidism results in protracted immaturity of the callosal connections. This hypothesis is consistent with the structural immaturity of callosal axons as reflected in the failure to express the NF-H epitope and by the severe reduction in axon myelination. Ultrastructural findings are also consistent with this hypothesis (see the accompanying paper this issue: Gravel et al., '90).

The callosal projections of the parietal cortex During the first 2 postnatal weeks of development in rats, the tangential distribution of callosally projecting neurons

MATURATION OF THE CORPUS CALLOSUM

,

......

141

.

Fig. 9. Camera lucida drawings of 50-pm-thick frontal sections taken a t 200-pm intervals through the right occipital cortex of a normal rat tha t received multiple WGA-HRP injections into the left visual cortex at P25. T he same conventions are used as in Figure 8. Th e lowest

density of retrogradely labelled cells is found in midarea 17 and midarea 18a. On comparing these drawings with those of Figure 8, it is evident t h at the density of callosally projecting cells in the visual cortex is lower in a normal animal than in a hypothyroid.

142

C. GRAVEL AND R. HAWKES

Fig. 10. Photomicrographs of frontal sections through the right occipital cortex of a hypothyroid rat that received multiple WGA-HRP injections into the left visual cortex at P35. The same conventions are used as in Figure 3. In this animal, the density of labelled callosal cells in midarea 17 is low in the superior third of the cortex. We do not know whether differences in callosal cell density in the upper part of midarea

17 of hypothyroid animals represent variations in tracer concentration between injection sites or the extent of interindividual differences. Also note extensive retrograde labelling of cells in the auditory cortex lateral to area 1Xa and the absence of anterograde labelling in this same region. Scale bar: 500 gm.

in the parietal cortex undergoes important changes. From PO to P4 callosally projecting neurons are arranged in two continuous tangential layers deep to the compact cell mass of the upper cortical plate (Wise and Jones, '76; Ivy and Killackey, '81). Subsequently, in areas destined not to be callosally connected in the adult there is a nearly complete disappearance of transcallosal projections and a more modest

decrease in the number of callosal cells in regions destined to remain callosally connected in the adult (Wise and Jones, '76; Ivy and Killackey, '81, '82). By P15, callosally projecting cells are confined to discrete radial arrays, mainly in layers I11 and V with a few in layers Vc-VIa, interspersed by acallosal zones of variable width (Wise and Jones, '76; Ivy and Killackey, '79, '81). In the hypothyroid parietal cortex, retro-

MATURATION OF THE CORPUS CALLOSUM

143

A

B

Fig. 11. Darkfield photomicrographs of frontal sections through the right parietal cortex of normal P30 (A) and hypothyroid P35 (B) rats that received WGA-HRP injections in the left parietal cortex. Medial is to the left. In the normal animal, the callosal projection consists of vertical arrays of codistributed callosal cells and af€erents interposed by

stretches of cortex almost devoid of both. In the hypothyroid animal, these vertical arrays are present, but the intervening patches of cortex contain two tangential laminae of retrogradely labelled cells. Note the absence of anterograde labelling in regions displaying a prominent layer IV (asterisks). Scale bar: 500 fim.

C. GRAVEL AND R. HAWKES

144

gradely labelled callosal neurons are also concentrated in layers I11 and V, but instead of being in radial arrays they form two continuous tangential sheets (Fig. 11).This suggests that after callosal axons have reached the contralateral hemisphere, the lack of thyroid hormones interferes with normal axon collateral elimination and thus no zones form that are free of callosally projecting neurons. This interpretation presupposes that the callosally projecting cells OCcupy the same layers in the immature cortex (PO-P4)as in the adult. According to Floeter and Jones ('85) and Wise and Jones ('76, '781, the callosally projecting cells in the newborn rat parietal cortex are already in two tangential strata corresponding to future layers I11 and V. This being the case, a failure of collateral elimination in layers I11 and V would produce the distribution observed in P35 hypothyroid animals. However, according to Ivy and Killackey ('81) the two laminae of callosally projecting cells found in the perinatal parietal cortex are the forerunners of layers Vc and VIa. According to their developmental scheme, the axons from callosal cells situated in the supragranular layers do not reach the contralateral hemisphere until F'4 or later. This being the case, a blockade of axon elimination affecting all callosally projecting cells would yield adults with a trilaminar retrograde labeling pattern: that is, with cells located in supragranular layers, in layer Va and in layers VcVIa. However, this is not the case. Our data show that callosally projecting cells in the hypothyroid parietal cortex are restricted to layer I11 and layer Va. Only in zones that receive callosal axons from the contralateral hemisphere and are callosally connected in a normal adult have we seen retrogradely labelled neurons in layers Vc-VIa (Fig. 11). The present data do not permit a choice between these two developmental schemes. T o be consistent with parietal callosal development according to Ivy and Killackey ('81), hypothyroidism must preferentially block axon elimination by cells in the supragranular layers and layer Va, whereas axon elimination proceeds normally for cells of layers VcVIa. A selective blockade of axon elimination in the parietal cortex that does not affect layers Vc-VIa is not as implausible as it might appear. We have argued elsewhere that the sensitivity of axons to a shortage in thyroid hormones correlates with the time at which they attain their targets, with those having reached their targets before about P4 being insensitive, those arriving later being sensitive (Plioplys et al., '86). In the parietal cortex, the first callosal axons reached the cortical subplate of the opposite hemisphere a t P3-P4 (Wise and Jones, '76; Ivy and Killackey, '811, and so may fall into the putative thyroid-insensitive class. These originate from callosal cells located in the infragranular layers (Floeter and Jones, '85), presumably layers Vc-VIa. Two other sets of data hint a t a possible differential sensitivity of layer VI callosally projecting neurons. First, neonatal destruction of the optic radiation in hamsters eliminates almost all but layers Vc-VIa callosal projections in the adult visual cortex (Rhoades et al., '87),and second, bilateral eyelid suture in neonatal cats results in a severe loss of supragranular callosal neurons, whereas layer \I1 callosal neurons are unaffected (Innocenti et al., '85).

The callosal projections of the occipital cortex Observations made on the visual cortex are allso consistent with the hypothesis of a thyroid-dependent sharpening of callosal projections during normal development. In the newborn rat, callosally projecting cells in the visual cortex

are arranged in two continuous layers corresponding to future layers Va and Vc-VIa (Olavarria and Van Sluyters, '85). The adult discontinuous pattern is generated subsequently by a progressive loss of callosal connections in regions that will normally have few callosally projecting cells in the adult (eg., the interiors of areas 17 and 18a),together with an increase in the number of callosally projecting cells in layers 11-IV of regions that are densely callosal in the adult (e.g., the 77-18a border, area 18b, and lateral area 18a: Olavarria and Van Sluyters, '85). Layers Vc-VIa contain numerous callosally projecting neurons throughout the visual cortex in adults (Olavarria and Van Sluyters, '83; '85; see also above). In contrast, in visual areas of hypothyroid animals there are two additional bands of callosally projecting cells, in layer 11-111and layer Va (Figs. 5-8, 10). This finding is also consistent with the hypothesis that hypothyroidism leads t o a block in the normal elimination of callosal projections from neurons in the supragranular layers and layer Va.

Mechanisms that refine cdosal topography The deficits leading to the abnormally wide tangential distribution of callosally projecting neurons in hypothyroid animals are unknown. It is important to note that the abnormal topography of the callosal projections in hypothyroidism does not seem to inrlude the axon terminal fields as the areas of high anterograde labelling in the hypothyroid cortex appear entirely normal. Thus it seems that a given set of cortical neurons receive callosal input from an unusually large territory in the contralateral hemisphere. This presupposes that the axons of "supernumerary" callosal cells effectively invade the opposite gray matter. Whether axons of transiently projecting corticocortical neurons significantly invade the gray matter of their target areas is contentious (Innocenti and Clarke, '84b; Dehay et al., '84, '88; Clarke and Innocenti, '86). Nonetheless, the present results indicate that normal axon outgrowth and target recognition mechanisms are not seriously compromised in hypothyroid animals, but rather what is affected is the ability to selectively eliminate projections deriving from topographically inappropriate cortical zones. In the hypothyroid visual cortex, there seems to be a supranormal number of callosally projecting cells even in regions that are strongly callosal in normal adults. This can be interpreted as another sign of the immaturity of the hypothyroid callosal connections since both qualitative (Ivy and Killackey, '81, '82) and quantitative (Koppel and Innocenti, '83) studies suggested a reduction in the callosal projection of strongly callosal zones during normal development. Therefore, it appears that both the selective and general elimination of exuberant callosal projections during normal development requires thyroid hormones. Several other experimental protocols lead to abnormal callosal connectivity, some of which produced an enlargement of callosal projection fields. In several species of rodents (including rat) as well as in cats, neonatal monocular eriucleation results in an expanded callosal projection to the ipsilateral area 17 of the visual cortex (Innocenti and Frost, '79, '80; Rhoades and DellaCroce, '80; Rothblat and Hayes, '82; Rhoades and Fish, '83; Lund e t al., '84; Olavarria and Van Sluyters, '84; Olavarria et al., '87). However, the distribution of callosal afferents in the deprived hemisphere is coextensive with the field of callosally projecting somata, which is also wider (Rhoades and DellaCroce, '80; Cusick and Lund, '82; Rhoades and Fish, '83; Lund et al., '84; Olav-

MATURATION OF THE CORPUS CALLOSUM

arria and Van Sluyters, '84; Olavarria et al., '87). Moreover, these enlarged callosal fields do not occupy the full extent of area 17 (Innocenti and Frost, '79, '80; Cusick and Lund, '82; Lund et al., '84; Olavarria and Van Sluyters, '84; Olavarria et al., '87). Neonatally induced convergent or divergent strabismus also leads to expanded callosal connections of area 17, but again this applies to both cell bodies and terminals, and only to a portion of this area (Lund et al., '78; Innocenti and Frost, '79; Berman and Payne, '83). Similarly, lesion of the neonatal rat visual thalamus leads to a wider distribution of callosal terminals (Cusick and Lund, '82). Other experimental manipulations including dark-rearing, binocular eyelid suture, lesions of the visual cortex, and lesions of the optic tract or optic radiations had no effects, or reduced the callosal connections, or induced distorted topography (Cusick, '78; Innocenti and Frost, '79, '80; Lund and Mitchell, '79; Cusick and Lund, '82; Innocenti e t al., '85; Rhoades et al., '87). Finally, in the Boston Siamese cat, which has genetic abnormalities in the connectivity of the visual system, the visual callosal connections are tangentially expanded but again this applies to both cell somata and terminals (Shatz, '77; Berman and Payne, '83). In summary, it is clear that none of these experimental conditions results in the characteristic topography seen in the visual cortex of hypothyroid animals-a wider distribution of callosally projecting neurons together with a normal distribution of terminals. Given that callosal maturation is sensitive to postnatal experience, it cannot be excluded that the observed callosal abnormalities are secondary. For example, elevated rates of strabismus and nystagmus have been reported in humans suffering from congenital hypothyroidism (Schulman and Crawford, '69; Kirkland et al., '72). We do not know if this is also the case for our hypothyroid animals, but, as discussed above, the consequences of hypothyroidism are quite different from those obtained in natural or surgically induced strabismic animals. There are no reports on the effects of nystagmus on callosal projections and the corollary, the possible effects of abnormal callosal projections on eye alignment or eye movement, is to our knowledge undocumented experimentally. The only previous reports of abnormal topography due to hypothyroidism come from experiments with anurans where a lack of thyroid hormone blocks development a t a premetamorphic stage (e.g., Maclean and Turner, '76). Metamorphosis is accompanied by many changes in neural connectivity (reviewed in Kollros, '81). For example, the ipsilateral retinofugal projections do not appear until late stage 54 in Xenopus, coinciding with a rise in the synthesis of thyroxine (Hoskins and Grobstein, '85). That the timing is thyroid hormone-dependent is supported experimentally: on the one hand, intraocular injections of thyroxine in young tadpoles induce the ipsilateral retinofugal projection prematurely (Beach and Jacobson, '79), and on the other hand, PTU-treatment blocks the normal appearance of the ipsilateral retinofugal connections, a blockade that is overridden by thyroxine treatment (Hoskins and Grobstein, '85). Similarly, hypothyroidism has been shown to block the axon elimination that normally occurs in the oculomotor nerve of Xenopus tadpoles during development, and the pruning of the projection resumes after the goitrogen is removed (Schonenberger and Escher, '88). The axons of the hypothyroid corpus callosum are immature in a t least three respects: their topography, their low level of myelination (see also Gravel et al., 'go), and their failure to express a normal mature cytoskeleton. It is

145

unclear whether these three are causally related (although it seems likely that myelination is not a prerequisite for NF-H expression: Couillard and Hawkes, MS in preparation), or whether each is independently affected by hypothyroidism. The most parsimonious view is that neuronal development is simply arrested a t about the P 5 stage, much as metamorphosis is blocked in amphibians. As a result, the axonal arbors would remain immature, able neither to retract inappropriate collaterals nor to permit a normal rate of myelination. However, as the adult hypothyroid callosal projection is not identical to the normal at P5, this view is necessarily incomplete. The most persuasive model of the maturation of callosal topography is due to Innocenti ('81, '86). An important reduction in the number of axons in the corpus callosum has been reported during the postnatal development of both cats and monkeys (Koppel and Innocenti, '83; LaMantia and Rakic, '84), and so it was proposed that this pruning might act to shape the callosally projecting domains in these animals. Applying this model to the hypothyroid rat might suggest that the immature topography is correlated with the preservation of exuberant collaterals and thus with abnormally high numbers of axons in the adult hypothyroid corpus callosum. This is not the case, and is discussed further in the accompanying manuscript (Gravel et a]., '90).

ACKNOWLEDGMENTS We thank Jamel Rafrafi and Rachel Sasseville for technical assistance. This work was supported by grants from the Medical Research Council of Canada (R.H.), and by a studentship from the F.R.S.Q. (C.G.).

LITERATURE CITED Beach, D.H., and M. Jacobson (1979) Jnfluence of thyroxine on cell proliferation in the retina of the clawed frog a t different ages. J. Comp. Neurol. 183.%15-624. Berman, N.E., and B.R. Payne (1983) Alterations in connections of the corpus callosum following convergent and divergentstrahismus. Brain Res. 274r201-212. J.Q. Trojanowski, W.W. Schlaepfer, and V.M.-Y. Lee (1967) Carden, M.J., Two-stage expression of neurofilament polypeptides during rat neurogenesis with early establishment of adult phosphorylation patterns. J. Neurosci. 7r3489-3504. Caviness, V.S. J r (1975) Architectonic map of neocortex of the normal mouse. J. Comp. Neurol. 164r247-264. Chow, K.L., H.D. Baumbach, and R. Lawson (1981) Callosal projections of the striate cortex in the neonatal rabbit. Exp. Brain Res. 42:122-126. Clarke, S., and G.M. Innocenti (1986) Organization of immature intrahemispheric connections. J. Comp. Neurol. 25r1-22. Cragg, B.G. (1970) Synapses and membranous bodies in experimental hypothyroidism. Brain Res. 18r297-307. Cusick, C.G. (1978) Normal and abnormal connections in the occipital cortex of albino rats. SOC. Neurosci. Abstr. 4,470. Cusick, C.G., and R.D. Lund (1982) Modification of visual callosal projections in rats. J. Comp. Neurol. 212r385-398. Dehay, C., J. Bullier, and H. Kennedy (1984) Transient projections from the frontoparietal and temporal cortex to areas 17, 18 and 19 in the kitten. Exp. Brain Res. 57;20&212. Dehay, C., H. Kennedy, and J. Bullier (1988) Characterization of transient cortical projections from auditory, somatosensory, and motor cortices to visual areas 17,18, and 19 in the kitten. J. Comp. Neurol. 272r68-89. Dussault, J.H., and P. Walker (1978) The effect of iodine deficiency and propylthiouracil on the hypo-thalamo-pituitary-thyroidaxis in the neonatal rat. Can. J. Physiol. Pharmacol. 56.950-955. Eayrs, J.T. (1971) Thyroid and developing brain--anatomical and behavioural effects. In: M. Hamburg and E.J. Barrington (eds.): Hormones and Development. New Y o r k Appleton-Century-Crofts, pp. 345-355. Feng, J.Z., and J.F. Brugge (1983) Postnatal development of auditory callosal connections in the kitten. J. Comp. Neurol. 2I4:41&426.

146 Floeter, M.K., and E.G. Jones (1985) The morphology and phased outgrowth of callosal axons in the fetal rat,. Dev. Brain Res. 22:7-18. Gravel, C., and R. Hawkes (1987a) Neuronal maturation in the normal and hypothyroid rat cerebellar cortex studied with monoclonal antibody MIT-23. J. Comp. Neurol. 258:447462. Gravel, C., and R. Hawkes (1987b) Thyroid hormone modulates the expression of a neurofilament antigen in the cerebellar cortex: premature induction and overexpression by basket cells in hyperthyroidism and a critical period for the correction of hypothyroidism. Brain Res. 422:327-335. Gravel, C., R. Sasseville, and R. Hawkes (1990). Maturation of the corpus callosum of the rat. 11. Influence of thyroid hormones on the number and maturation of axons. J. Comp. Neurol. 291:147-161 Hawkes, R., E. Niday, and J. Gordon (1982) A dot immunobinding assay for monoclonal and other antibodies. Anal. Biochem. 119r142-147. Hoskins, S.G., and P. Grobstein (1985) Development of the ipsilateral retinothalamic projection in the frog Xenopus laevis. The role of thyroxine by thyroxine. J. Neurosci. 5t930-940. Innocenti, G.M. (1981) Growth and reshaping of axons in the establishment of visual callosal connections. Science 212r824-827. Innocenti, G.M. (1986) What is so special about callosal connections? In: F. Lepore, M. Ptito, and H.H. Jasper (eds): Two Hemispheres-One Brain: Functions of the Corpus Callosum. New York Alan R. Liss Inc.. pp. 7 s 81. Innocenti, G.M., and S. Clarke (1984a) Bilateral transitory projection to visual areas from auditory cortex in kittens. Dev. Brain Res. 14:143-148. Innocenti, G.M., and S. Clarke (1984b) The organization of immature callosal connections. J. Comp. Neurol. 230:287-309. Innocenti, G.M., and Caminiti, R. (1980) Postnatal shaping of callosal connections from sensory areas. Exp. Brain Res. 38:381-394. Innocenti, G.M., Fiore, L., and Caminiti, R. (1977) Exuberant projection into the corpus callosum from the visual cortex of newborn cats. Neurosci. Lett. 4237-242. Innocenti, G.M., and D.O. Frost (1979) Effects of visual experience on the maturation of the efferent system to the corpus callosum. Nature 2811:231-234. Innocenti, G.M., and D.O. Frost (1980) The paatnatal development of visual callosal connections in the absence of visual experience or of the eyes. Exp. Brain Res. 39t365-375. Innocenti, G.M., D.O. Frost, and J. Illes (1985) Maturation of visual callosal connections in visually deprived kittens: A challenging critical period. J. Neurosci. 5955-267. Ivy, G.O., and H.P. Killackey (1981) The ontogeny of the distribution of callosal projection neurons in the rat parietal cortex. J. Comp. Neurol. 195r367-389. Ivy, G.O., and H.P. Killackey (1982) Ontogenetic changes in the projections of neocortical neurons. J. Neurosci. 2:73&743. Ivy, G.O., R.M. Akers, and H.P. Killackey (1979) Differential distribution of callosal projection neurons in the neonatal and adult rat. Brain Res. 173:532-537. Jacobson, S. (1963) Sequence of myelination in the brain of the albino rat. A. Cerebral cortex, thalamus and related structures. J. Comp. Neurol. 121:529. Kirkland, R.T., J.L. Kirkland, MC. Roberson, L. Librik, and G.W. Clayton (1972) Strabismus and congenital hypothyroidism. J. Pediat. 80.648650. Kollros, J.J. (1981) Transitions in the nervous system during amphibian metamorphosis. In L.I. Gilbert and E. Frieden (eds): Metamorphosis. A Problem in Developmental Biology, 2nd ed. New York Plenum Press, pp. 445-459. Koppel, H., and G.M. Innocenti (1983) Is there a genuine exuberancy of callosal projections in development? A quantit.ative electron microscopic study in the cat. Neurosci. Lett. 41:33-40. Krieg, W.J.S. (1946) Connections of the cerebral cortex. I. The ,albino rat. B. Structure of the cortical areas. J. Comp. Neurol. 84:277-324. LaMantia, A-S., and P. Rakic (1984) The number, size, myelination, and regional variation of axons in the corpus callosum and anterior commissure of the developing rhesus monkey. Soc. Neurosci. Abstr. lOr1081. Leclerc, N., C. Gravel, A. Plioplys, and R. Hawkes (1985) Basket cell development in the normal and hypothyroid rat cerebellar cortex studied with a monoclonal anti-neurofilament antibody. Can. J. Biochem. Cell Biol. 63:564-576. Legrand, J. (1983) Hormones thyroidiennes et maturation du systeme nerveux. J. Physiol. (Paris) 78:603-652. Lund, R.D., F.-L.F. Chang, and P.W. Land (1984) The development of callosal projections in normal and one-eyed rats. Dev. Brain Res. 14:139-142. Lund, R.D., and D.E. Mitchell (1979) Asymmetry in the visual callosal con-

C. GRAVEL AND R. HAWKES nections of strabismic cats. Brain Res. 167:17&179. Lund, R.D., D.E. Mitchell, and G.H. Henry (1978) Squint-induced modifications of callosal connections in cats. Brain Res. 144:169-172. MacLean, N., and M. Turner (1976) Adult hemoglobin in developmentally retarded tadpoles of Xenopus laevis. J. Embryol. Exp. Morphol. 35261266. Mesulam, M.M. (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: A non-carcinogenic blue reaction product with superior sensitivity for visualizing neural af€erents and efferents. J. Histochem. Cytochem. 26:10&117. Miller, M.W., and B.A. Vogt (1984) The postnatal growth of the callosal connections of primary and secondary visual cortex in the rat. Dev. Brain Res. 14:30&309. Olavarria, J., and R.C. Van Sluyters (1983) Widespread callosal connections in infragranular cortex of the rat. Brain Res. 279:233-237. Olavarria, J., and R.C. Van Sluyters (1984) Callosal connections of the posterior neocortex in normal-eyed, congenitally anophthalmic, and neonatally-enucleated mice. J. Comp. Neurol. 230r249-268. Olavarria, J., and R.C. Van Sluyten (1985) Organization and postnatal development of callosal connections in the visual cortex of the rat. J. Comp. Neurol. 239:l-26. Olavarria, d., R. Malach, and R.C. Van Sluyters (1987) Development of visual callosal connections in neonatally enucleated rats. J. Comp. Neurol. 260:321-348. O'Leary, D.D.M., B.B. Stanfield, and W.M. Cowan (1981) Evidence that the early postnatal restriction of the cells of origin of the callosal projection is due to the elimination of axonal collaterals rather than to the death of neurons. Dev. Brain Res. lr607-617. Oppenheimer, J.H., H.L. Schwartz, and M.I. Surks (1972) Propylthiouracil inhibits the conversion of 1-thyroxine to 1-triiodothyronine. J. Clin. Invest. 51:2493-2497. Paxinos, G., and C. Watson (1986) The Rat Brain in Stereotaxic Coordinates. 2nd ed. Sydney: Academic Press. Plioplys, A.V., C. Gravel. and R. Hawkes (1986) Selective suppression of neurofilament antigen expression in the hypothyroid rat cerebral cortex. J. Neurol. Sci. 75:53-68. Rhoades, R.W., and D.D. Dellacroce (1980) Neonatal enucleation induces an asymmetric pattern of visual callosal connections in hamsters. Brain Res. 202:189-195. Rhoades, R.W., and S.E. Fish (1983) Bilateral enucleation alters visual callosal but not corticotectal and corticogeniculate projections in hamster. Exp. Brain Res. 51:451-462. Rhoades, R.W., S.E. Fish, R.D. Mooney, and N.L. Chiaia (1987) Distribution of visual callosal projection neurons in hamsters subjected to transection of the optic radiations on the day of birth. Dev. Brain Res. 32.217-232. Rothblat, L.A., and L.L. Hayes (1982) Age-related changes in the distribution of visual callosal neurons following monocular enucleation in the rat. Brain Res. 246,146149. Ruiz-Marcos, A,, J. Sala, F. Sanchez-Toscano, and G. Morreale De Escobar (1978) Influence of neonatal hypothyroidism on the pyramidal cells of the rat auditory cortex. Ann. Endocrinol. (Paris) 39r15A. Ruir-Marcos, A., F. Sanchez-Toscano, F. Escobar Del Rey, and G. Morreale De Escobar (1979) Severe hypothyroidism and the maturation of the rat cerebral cortex. Brain Res. 162:315-3B9. Schonenberger, N., and Escher, G. (1988) Excessive numbers of axons after early enucleation and blockade of metamorphosis in the ocnlomotor nerve of Xenopus laevis. Dev. Brain Res. 40.253-260. Schulman, J.D., and J.D. Crawford (1969) Congenital nystagmus and hypothyroidism. New Engl. J. Med. 280:70R-710. Seggie, J., and M. Berry (1972) Ontogeny of interhemispheric evoked potentials in the rat: significance of myelination of the corpus callosum. Exp. Neurol. 35215-232, Shntz, C.J. (1977) Anatomy of interhemispheric connections in the visual system of Boston Siamese and ordinary cats. J. Comp. Neurol. 173:497-518. Valentino, K.L., and E.G. Jones (1982) The early formation of the corpus callosum: a light and electron microscopic study in foetal and neonatal rats. J. Neurocytol. JJt583-609. Wise, S.P., and E.G. Jones (1976) The organization and postnatal development of the commissural projection of the rat somatic sensory cortex. J. Comp. Neurol. 168:313-344. Wise, S.P., and E.G. Jones (1978) Developmental studies of thalamocortical and commissural connections in the rat somatic sensory cortex. J. Comp. Neurol. 178:187-208. Zilles, K. (1985) The Cortex of the Rat. A Stereotaxic Atlas. New York: Springer-Verlag.

Maturation of the corpus callosum of the rat: I. Influence of thyroid hormones on the topography of callosal projections.

The normal adult rat corpus callosum contains numerous axonal profiles that are immunoreactive for the high molecular weight subunit of the neurofilam...
3MB Sizes 0 Downloads 0 Views