A critical period for reduced brain vulnerability to developmental injury

Developmental Brain Research 105 Ž1998. 325–337

Research report

A critical period for reduced brain vulnerability to developmental injury II. Volumetric study of the neocortex and thalamus in cats Troy D. Schmanke, Jaime R. Villablanca ) , Vladimir Lekht, Hetul M. Patel Departments of Psychiatry and BiobehaÕioral Sciences and of Neurobiology, Mental Retardation Research Center and Brain Research Institute, UCLA School of Medicine, UniÕersity of California, Los Angeles, CA 90024, USA Accepted 21 October 1997

Abstract Groups of young adult cats with a left hemineodecortication at postnatal ŽP. ages Žin days. 5–15 ŽP10., 30 ŽP30., 60 ŽP60., 90 ŽP90., 120 ŽP120. and in adulthood, were used to measure the volume of the thalamus, bilaterally, and of the remaining neocortex Žright hemisphere.. The same subjects were employed for the behavioral studies reported in the preceding paper. There was a bilateral, age-dependent, thalamic volume decrease. Ipsilateral to the resection, the thalamic shrinkage was the largest for the adult-lesioned cats Žby 56.7%. and it was the smallest for the P30 group Ž43.3%., with a tendency towards a greater atrophy as the age at lesion increased. A similar pattern of atrophy was seen for the contralateral thalamus but the volume reduction was much less pronounced such that it was significant only for the four older age-at-lesion groups Žranging from 18.2% to 11.2% for the P120 and P90 groups respectively.. Once again, the shrinkage was the smallest for the P30 group Ž5.3%.. The remaining neocortex also shrunk in these animals, but the volume decrease was significant only for the adult-lesioned Ž17.8%. and the P120 group Ž15.4%., while the P30 group had practically no shrinkage Ž2.4%.. The frontal cortex had no atrophy or it was minimal but the shrinkage gradually increased caudally such that all lesioned groups had some size reduction of the occipital cortex. The present results, together with the main conclusion of the preceding paper, indicate that there is a critical maturation period ŽCMP. of reduced forebrain vulnerability to neocortical injury which, in cats, tends to end between 30 to 60 days postnatally. The implications for developmental brain damage in other higher mammal species as well as the possible morphological ontogenetical underpinnings of this period are discussed. q 1998 Elsevier Science B.V. Keywords: Neocortex; Thalamus; Cat; Developmental injury of brain

1. Introduction During the last seventeen years we have used a feline model of cerebral hemispherectomy w56x to study differential age-at-lesion effects upon the size of the remaining structures as well as upon the numberrpacking density and soma size of cells in selected brain areas. Initially, we reported w55x that after hemispherectomy in adult cats, there is a bilateral shrinkage of the thalamus as determined by measuring cross-sectional areas at two coronal levels. This atrophy is dramatic in the thalamus ipsilateral to the ablation, reaching almost a 60% reduction, and coexists with extensive neuron loss, decrease in size of the largest neurons and increase in glial cells packing density in the ventrobasal complex. These regressive events are signifi) Corresponding author. Department of Psychiatry, 760 Westwood Plaza Žroom 58-258., UCLA, Los Angeles, CA 90024, USA. Fax: q1-310-206-5060; E-mail: [email protected]

0165-3806r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 7 . 0 0 1 8 8 - 0

cantly attenuated in cats sustaining a similar resection between postnatal ŽP. 5 to 15 days of age. Later, we reported similar results in the dorsal lateral geniculate w52x and in the medial geniculate w18x nuclei. We have not performed studies on the size of the remaining neocortex in hemispherectomized cats. However, we have determined that adult cats with a unilateral removal of the sensorimotor cortex sustained prenatally show a substantial decrease in the volume Žby 26.5%. of the remaining neocortex ipsilateral to the lesion w30x while such atrophy does not occur in cats with a similar ablation sustained neonatally at age P9–14 w24x. In the fetal-lesioned cats the ipsilateral thalamus also shrinks Žby 25.7%. with no substantial cytoarchitectural changes Žventrobasal complex. while the contralateral thalamus only tends to a volume decrease Žby 11.1%.. The thalamic shrinkage also occurs in the neonatal-lesioned cats, albeit to a lesser extent Ž14% ipsilateral reduction., and is concomitant with a tendency to neuron soma shrinkage Žby 8.0%..

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

326

In view of the above findings, the goals of the present study were to use cats with hemineodecortication ŽHDec. at different postnatal ages to: Ža. determine the volume of the neocortex of the remaining intact hemisphere w48x; Žb. more precisely evaluate the thalamic shrinkage by measuring total thalamic volumes bilaterally w64x; and Žc., determine if there is a transitional age-at-lesion during which the neonatal pattern of relatively reduced atrophy shifts into the adult pattern of greater thalamic and cortical shrinkage. We selected hemineodecortication rather than hemispherectomy in order to preserve the caudate nucleus Žn.. for analysis w31x. The neostriatum is usually spared in human hemispherectomy w13,44x and thus our model is very similar to the clinical procedure.

2.2. Surgery and maintenance The surgical procedure has been described in the preceding companion paper w57x and, in more detail, in previous publications w19,55x. The resection run through the internal capsule, lateral to the caudate n. and thalamus both of which were fully spared. The precise extent of the ablation in the present cats will be described in Section 3. After surgery all animals were rewarmed and the kittens were maintained for the night in a thermostatically controlled incubator. When spontaneously active, cats were returned to the colony and closely monitored thereafter. 2.3. Brain anatomy At the end of the behavioral studies Žsee Ref. w57x., all animals were euthanized with an overdose of pentobarbital Ž100 mgrkg., the brains were quickly removed and frozen in dry ice Žmost brains were used for cerebral metabolism studies which have been partially reported; w19,62x.. The frozen brains were inspected and photographed from several angles at this time and a preliminary description of the gross anatomy of the lesion was recorded. All brains were sectioned coronally at 20 m m thickness Žwith a 220 m m interval between sections. and the sections were serially stained with thionin or processed for cytochrome oxidase histochemistry ŽCYO. as described in Ref. w17x. To verify the extent and gross topography of the HDec, the lesion in each brain was reconstructed at the level of eight regularly spaced coronal planes from the atlas of Reinoso Suarez ´ w46x. Histological sections corresponding to these planes were projected upon copies of matching planes of the atlas and the lesion was drawn on these pages.

2. Materials and methods 2.1. Subjects We used 36 young adult cats of which six were intact control cats and 30 were animals which sustained a left hemineodecortication at six different postnatal ages. The HDec cats included most of the same animals used for the behavioral studies reported in the preceding paper w57x. For the sake of brevity these six age-at-lesion groups will be called P10, P30, P60, P90, P120 and adult-lesioned groups ŽTable 1.. The number of brains in each group was 6, 4, 5, 5, 5 and 5, respectively Žthe other three available brains from the behavioral study could not be used due to scattered defects in histology processing.. The exact mean days Žand ranges. at which the animals sustained HDec, as well as other characteristics of these animals, are listed in Table 1. Although most cats were males, each group had one to two females. All animals were born and reared in the kitten colony of the UCLA Mental Retardation Research Center under conditions described in the preceding paper w57x. Efforts were made to minimize suffering and to reduce the number of animals used. All procedures were approved by the UCLA Chancellor’s Committee for Animal Research.

2.4. Measurements of thalamic and neocortical Õolumes and of cortical cross-sectional areas Using a projection microscope, for each brain the outline of the thalamus and the neocortex was drawn at a magnification of =15 using preferentially the CYO sec-

Table 1 Important features from the groups of hemineodecorticated and control cats used in this study. P s postnatal age in days Žapproximate. Group

N

P10 P30 P60 P90 P120 Adult Control

6 4 5 5 5 5 6

a

s age at euthanasia.

Age at lesion Ždays.

Postlesion survival Ždays.

Weight at euthanasia Žkg.

Mean

Range

Mean

Range

Mean

Range

9.8 30.5 60.2 90.2 120.8 619.0 NA

6–13 30–32 60–61 90–91 120–122 387–1220 NA

485.7 248.0 382.2 401.6 418.6 303.6 658.0 a

335–680 200–369 330–410 373–419 406–431 210–526 244–1825a

4.4 3.8 4.7 5.2 4.3 4.3 3.6

2.3–6.3 3.5–4.9 4.3–5.8 4.2–6.8 2.8–6.0 3.0–5.9 2.4–6.1

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

327

tions, in which the contrast between the gray and white matter is striking ŽFig. 1. or, occasionally, the thionin stained sections. For the thalamus, all sections available were used Žrange, 17–23 per brain. and the drawings excluded the epithalamus, habenula Žand tract., medial geniculate nucleus, anterior and posterior paraventricular nuclei, zona incerta, and fields of Forel. For measurements of the neocortex, the drawings stopped at the rhinal sulcus in order to exclude the paleocortex. Every other coronal section available per brain Žrange 40–63. was used. A calibration slide ŽReichert. was used to draw calibration marks on each of the drawings. The surface area of each of the structures was calculated with a digitizing tablet ŽJandel Scientific, CA. and computer software ŽSigma Scan, Jandel Scientific.. Thereafter thalamic and neocortical volumes were estimated using the formula derived by Madarasz et al. w35x: V s Õ Ž a1r2 q a2 q a3 q . . . qa ny1 q a1r2.; where V is the volume of the structure in mm3 , Õ is the distance between two consecutive sections Žplus 1 section thickness., a is the structure surface area measured at each section, and n is the number of sections. 2.5. Statistical analysis For volumes, the group means for each hemisphere were compared using a univariate ANOVA w15x. For the thalamus the left and right side volumes were compared with the volumes of the same thalamic side in intact controls. Comparisons of cross-sectional areas Žneocortex only. were made using a two-factor, repeated measures ANOVA with lesion grouping as the between factor and coronal section as the within or repeated factor. For the post-hoc analysis, we used a Tukey’s test w38x in both cases. 3. Results The growth of the HDec kittens was uneventful. As shown in Table 1, at the time of euthanasia all animals were young adults, i.e., between 230 Žthe youngest, P30 group. and 1746 Žthe oldest, adult-lesioned group. days. The mean weights were similar between groups. 3.1. Gross anatomy of the brains There were only minor differences in the extent of the ablation between the age at lesions groups. In all animals almost the entire neocortex was removed ŽFig. 1, as well as Fig. 1 of the preceding paper.. However, since the caudate was spared, portions of the neocortex ventral and lateral to the head of this nucleus were not resected. These included the lower 1r3 to 1r2 of the anterior and posterior sylvian sulci, the anterior and posterior sylvian gyri, and the lower 1r3 to 1r2 of the medial sylvian gyrus Žwith these remnants being similar across groups.. The thalamus and the caudate nucleus were grossly intact in all

Fig. 1. Low power photomicrographs of cytochrome oxidase-stained coronal sections at about the same level of the thalamus for Ža. and Žb., to illustrate the extend of the hemineodecortication in adult-lesioned Ža. and in neonatal-lesioned Žb. cats. Note that the decortication is extensive and yet it entirely spares the thalamus. Also note the smaller size of the left thalamus in Ža. as compared to Žb..

brains. Other than the above, the ablation was similar to the hemispherectomy described elsewhere w52,55x. Thus, other telencephalic areas which were spared included the olfactory bulb and tract, the ventral aspect of the g. proreus, the g. presylvius, areas of the ventral surface medial to the olfactory tract and s. rhinicus anterior–posterior, amygdala, n. accumbens, the claustrum and putamen. Dorsally, there were cortical remnants only in the midline, rostral to the genu of the corpus callosum, including variable portions of the gyri rectus and cinguli. The septum and midline fornix were not directly damaged, but most fimbria fibers were interrupted andror removed together with the dorsal hippocampus. 3.2. Volume of the thalamus 3.2.1. Thalamus ipsilateral to the lesion (left) There was a marked volume shrinkage for all lesioned groups as compared to intact controls Ž P - 0.01. and the magnitude of this decrease was age related ŽFig. 2a.. Thus, on the one hand, the volume loss reached over 50% of the normative value for the left thalamus Ž363 mm3 . for both

328

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

Fig. 2. Means ŽS.E.. volume of the ŽA. left side of the lesion, and ŽB. right Žopposite to the lesion side. thalami for each of the age-at-lesion groups and for the control cats. For the left thalamus note that: Ža. all hemidecorticated groups showed a highly significant thalamic atrophy Ž ) ) P - 0.01.; Žb. only the groups lesioned at P10, P30 and P60 showed significantly less atrophy than the cats lesioned in adulthood ŽqP - 0.05.; and, Žc. there was a gradient of increasing atrophy as the age-at-lesion also increased. For the right thalamus note that: Ža. the thalamic atrophy was less pronounced than ipsilaterally for all age-at-lesion groups Ž P - 0.0001.; Žb. compared to controls, only the groups lesioned at P60 or at an older age showed a significant thalamic shrinkage Ž ) ) P - 0.01.. Note that for the intact, control cats, the volume of the right thalamus was larger than that of the left Ž P - 0.02.. P s postnatal age in days.

the P120 Ž177 mm3, equivalent to a 51.3% loss. and adult-lesioned Ž157 mm3 , or a 56.7% loss. cats. On the other hand, for the groups P10 through P90 the volume ranged from 200 to 206 mm3 such that the decreases were smaller Žof the order of 43–45%.. Therefore, for the groups between P10 through P120, the differences were small with the largest increment in shrinkage Ža 6.3%, Fig. 2a. occurring between the P90 and P120. Comparisons

between lesioned groups showed that only the P10, P30 and P60 groups exhibited significantly less thalamic shrinkage than the adult-lesioned cats Ž P - 0.05.. Of these groups, the P30 had the smallest shrinkage Ž43.3%.. The ANOVA values were Ž F w6,27x s 46.61, P - 0.000.. Finally, for the brains with the largest shrinkage ŽP120 and adult-lesioned., the number of coronal sections available for measurements was reduced Žto a mean of 17 sections

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

329

per brain as compared to 23 for intact brains., suggesting that the thalamic atrophy also occurred across the rostal– caudal dimension of this structure.

ANOVA Ž F w1,10x s 7.19, P - 0.02., than that of the left thalamus Ž363 mm3 ..

3.2.2. Thalamus contralateral to the lesion (right) The outstanding finding was that there was also atrophy in the thalamus contralateral to the lesion ŽFig. 2b.; however, this was much less than ipsilaterally. Compared to the normative value for the right thalamus Ž394 mm3 ., only the P60, P90, P120 and adult-lesioned groups showed a significant atrophy, ranging from 11.2% for the P90 group Ž350 mm3, P - 0.05. through 18.2% for the P120 group Ž323 mm3, P - 0.01.. The thalami of the P10 and P30 groups shrank by only 8.35% and 5.3%, respectively. Thus, with the thalamus of the P30 groups showing the least shrinkage, it was interesting to note that, among the P10 through P120 the largest jump Žan 8.8% increment in shrinkage. occurred between the P30 and P60 Žvolume, 338.0 mm3 . groups. The ANOVA values were Ž F w6,28x s 6.89, P - 0.0001.. As suggested by the slope of the bars in Fig. 2, for both sides of the thalamus there was a tendency to an increase in atrophy with age at lesion across all six groups.

The outstanding finding was that the neocortex of the remaining intact hemisphere also atrophied relative to controls Ž4165 mm3 .. However, this effect was significant only for the adult-lesioned and P120 groups ŽFig. 3.. The adult-lesioned cats showed the largest decrease Ž17.8%. with a mean volume of 3423.8 mm3 Ž P - 0.01. while the P120 animals had a 15.4% shrinkage with a mean volume of 3525.7 mm3 Ž P - 0.05.. As seen in Fig. 3, the other groups showed a lesser neocortical atrophy with the P30 group exhibiting practically none Ž2.4%, 4066 mm3 ., followed by the P10 group with a 10.1% decrease Ž3799.5 mm3 .. The differences between lesioned groups were not significant although in some cases they were relatively large Že.g., a 15.6% between the P30 and adult-lesioned groups, P - 0.1.; in addition, it was interesting that, among the P10 through P120 groups, the largest increase in atrophy Ža 9.2% difference; Fig. 3. occurred between the P30 and P60 Ž3683 mm3 . groups. The ANOVA values were Ž F w6,27x s 4.17, P - 0.004.. As seen in Fig. 3, there was also a tendency towards a parallel increase of the cortical shrinkage and the age-at-lesion across all six lesioned groups. As with the thalamus, but to a greater extent, the number of coronal sections available for measurements was significantly less for both the P120 Ž P 0.05. and adult lesioned Ž P - 0.05. groups suggesting a size reduction along the rostral caudal axis of these brains.

3.2.3. Interhemispheric comparisons The larger magnitude of the left versus the right thalamic atrophy was highly significant for all age-at-lesion groups ŽANOVAŽs. range: P - 0.028 to P - 0.0000.. In addition, it was interesting that in intact cats, the volume of the right thalamus Ž399 mm3 . was bigger by 9.0%,

3.3. Volume of the neocortex

Fig. 3. Mean ŽS.E.. volume of the remaining neocortex for each of the age-at-lesion groups and for the control cats. Note that although there was a tendency to cortical atrophy for all age-at-lesion groups, the atrophy was significant only for the P120 Ž ) P - 0.05, Tukey’s test. and adult-lesioned Ž ) ) P - 0.01. groups. P s postnatal age at lesion in days.

330

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

3.4. Neocortex cross-sectional areas This analysis was mainly used to further define differences between the lesioned groups. Coronal sections rostral to A23 and caudal to P5 were not used for this analysis because not all brains had enough available sections outside this range. However, since there were still more sections available than coronal planes illustrated in the

atlas of Reinoso Suarez ´ w46x, the sections were allocated to the planes of the atlas corresponding the closest to the levels of the sections. This also allowed to study if the differences were clustered at specific A–P planes as well as permitted planes comparisons with the control brains. For comparisons between lesioned brains, the ANOVA values showed high significance Ž F w6,27x s 7.18, P 0.001.. The post-hoc comparisons showed significant dif-

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

ferences Žless atrophy. for selected A–P planes mainly of the P30 group. Thus, the P30 group had 7, 5 and 16 planes showing significantly less atrophy than the same planes of the adult-lesioned, P120 and P90 groups respectively ŽFig. 4, P - 0.01 to P - 0.05.. Among the other groups only the P60 exhibited four planes with significantly less shrinkage than the same levels of group P90 Ž P - 0.01 to P - 0.5.. However, as seen in the graphs ŽFig. 4., there was a general tendency for the P10, P30 and P60 groups to have less atrophy than the three older groups. Thus, although the cortical volumes did not show significant differences between the age-at-lesion groups, the cross-sectional area analysis showed that selected coronal sections of brains of the P30 group shrunk significantly less than the corresponding sections of the adult-lesioned, P120 and P90 groups. These results coincided with the tendency to smaller decreases in volume for the P30 group described above suggesting that this was the group with the least neocortical atrophy, Regarding comparisons with the intact animals, the P30 group had only one coronal plane, located caudally in the brain ŽP5., showing significant shrinkage compared the same level of control cats, but on the other hand, it also had two anterior planes ŽA16 and A17. showing higher values Ž P - 0.01. than controls. In contrast, the adult-lesioned, P120, P90, P60 and P10 groups had 7, 4, 14, 3 and 3 planes, respectively ŽFig. 4a and b., showing a significant atrophy compared to the same planes of intact brains. Overall, it was interesting that, as seen in Fig. 4a and b, compared to control cats, the frontal sections of the brain did not show atrophy in any of the groups, including the adult-lesioned cats. It was only behind plane A16, by the end of the caudate n. head and close to the rostral thalamus, that shrinkage began to be seen for the adult-lesioned, P120 and P90 groups Žin fact, for the adult-lesioned group, all sections showing significant shrinkage were behind A16.. For the P60 group atrophy began at about A12, and for the P30–P10 groups some shrinkage was seen only behind P2. Therefore, considering all lesioned groups a pattern emerged suggesting that the visual cortex was affected in all groups, that the parietal–temporal areas were progressively involved as the age at

331

surgery increased, and that the frontal cortex did not shrink.

4. Discussion The present results document the following conclusions. Following unilateral neodecortication there was a bilateral shrinkage of the thalamus but the ipsilateral atrophy was substantially greater. The remaining contralateral neocortex was also reduced in volume. We previously reported this bilaterality of effects for subcortical structures of hemispherectomized cats w18,52,55,59x, but this is the first time that we report an effect upon the neocortex of the intact hemisphere. The effects were age-dependent such that, for the ipsilateral thalamus, although all lesioned groups showed significant atrophy relative to intact cats: Ža. maximal atrophy occurred after hemineodecortication in adulthood; Žb. significantly less shrinkage relative to the adult-lesioned animals was seen for cats lesioned at P10, P30 and P60; and, Žc. the volume for the P90 and P120 groups fell in between. For the contralateral thalamus, the atrophy was minimal Žnot significant. for the P10 and P30 groups and then increased with age-at-lesion such that, starting with the P60 and including all older groups, the atrophy became significant relative to intact cats. For the contralateral cortex, there was a significant atrophy for the P120 and adult-lesioned groups with a tendency to shrinkage for the younger groups Žwhich was minimal for the P30 group.. For a number of coronal planes of the P30 group, there was significantly less shrinkage than for the same levels of the adult-lesioned, P120 and P90 groups. Considering the goals of this paper, two tendencies are worth highlighting, i.e., that for the thalamus bilaterally and for the contralateral neocortex, the group showing the least atrophy was the P30 while for the contralateral thalamus and neocortex of the P10 through P120 groups, the largest increment in shrinkage was seen between the P30 and P60 groups. Regarding the cross-sectional areas, there was an interesting tendency to the same pattern of rostrocaudal distri-

Fig. 4. Mean ŽS.E.. cross sectional areas of coronal sections of the contralateral neocortex at the stereotaxic w46x anterior–posterior planes indicated in the abscissa. ŽA. shows the values for the P10, P30, adult-lesioned and control groups. ) P - 0.05 and ) ) P - 0.01 indicate significantly higher values Žless shrinkage. for the P30 group compared to the adult-lesioned group; the P30 brains also showed significantly Ž". higher values Ž P - 0.01. compared to controls for two anterior planes ŽA16 and A17.. q P - 0.05 and qq P - 0.01 indicate significant shrinkage for adult-lesioned cats compared to controls. The only other significant differences Ž P - 0.01. were for three posterior planes of the P10 Ž;. and one of the P30 Ž". groups which showed shrinkage relative to controls. Thus, while the adult-lesioned cats had six sections showing a significant atrophy relative to controls, the P30 had only one of such sections Žbut also two which were larger.. ŽB. shows the cross-sectional area values for the P60, P90, P120 and control groups. ) P - 0.05 and ) ) P - 0.01 indicate significant shrinkage for the P90 groups compared to the control group. qP - 0.05 and qq P - 0.01 indicate significant shrinkage for the P120 group compared to the control group. The only other significant Ž". difference Ž P - 0.01. was for three posterior planes of the P60 group which showed shrinkage relative to controls Žand also four planes showing less shrinkage relative to the P90 group, not depicted in the graph.. In both ŽA. and ŽB., note that for all lesioned groups there is no atrophy in the frontal sections and that the reduction in size of the cross-sectional areas begins to appear behind A16 for the older age-at-lesion groups and is seen only in the most posterior planes for the P10 and P30 groups Žfor details, see Section 3.4.. The groups were split between the ŽA. and ŽB. graphs only to avoid excessive cluttering of the values. Note that for both graphs. the group with the less shrinkage is shown in gray, while the group with the greater shrinkage is shown in black.

332

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

bution of neocortical atrophy in all lesioned groups. While in the sections in front of the thalamus there was no or very little atrophy Žincluding a tendency to an increased size in the P10 and P30 groups., there was a gradient towards increased atrophy in the caudal regions such that in all groups, the greater shrinkage was for the occipital cortex. 4.1. The nature of the thalamic and cortical atrophy In previous work, we demonstrated that the thalamic atrophy in postnatal hemispherectomized cats was typically due to retrograde–anterograde direct or transsynaptic degeneration w34,52,55x, and it is most likely that the same mechanisms apply in the present hemidecorticated animals. In the ipsilateral thalamic ventrobasal complex of adulthemispherectomized cats w55x there was a dramatic decrease in packing density Žby about 80.0%. of the large projection neurons together with a shrinkage of the soma Žcross-sectional area. of the remaining large neurons Žby about 35%.. In contrast, the glial cells packing density almost doubled in these animals relative to controls. In cats sustaining hemispherectomy at around P10 the decrease in packing density of the large neurons was 52.0% Žwith the reduction on their soma size being significantly less than in adult-lesioned cats., and that of glial cells was slightly lower than in intact cats. Similar changes were seen in the dorsal lateral w52x and in the medial geniculate w18x nuclei and to a lesser extent, transsynaptically, in the mammillary bodies and superior colliculi w59x of P10 versus adulthemispherectomized cats. The low neuronal packing density that occurred concomitantly with thalamic shrinkage, particularly in adult-hemispherectomized cats, reflected a dramatic decrease in neuron numbers. This interpretation has been confirmed in our recent study of the caudate n. w12x in HDec w31x cats, where the total number of neurons was calculated using stereology. Even though the predominantly anterograde degeneration in the ipsilateral caudate of the adult lesioned cats Ž18.1%. was much smaller than the mostly retrograde degeneration seen in the thalamus, the reduction in neuron numbers was also matched by a corresponding decrease in total caudate volume and neuronal cell packing density. The present results for the thalamus in the P10 and adult-lesioned groups confirmed our above report in cats of the same age-at-lesion but in which the resection also included the caudate n. In those cats w55x we did not calculate volumes but only measured the cross-sectional area of the thalamus at two coronal levels, A 8.5 and A 7.5. For these two planes taken together the mean crosssectional area shrinkage of the ipsilateral thalamus was 47.4% and 60.3% for the P10 and adult-hemispherectomized cats and 35.5% and 48.1% for the present P10 and adult-hemidecorticated animals, respectively. Thus, although only two planes were available for comparisons, the results suggest that including the caudate n. in the

resection further adds to the thalamic atrophy. This makes sense since in absence of the caudate the ipsilateral thalamus is further deprived of afferents, transneuronally, due to elimination of the striato–pallidal–thalamic projections w63x. Contralaterally, the effects were similarly different since the volume decrease was 14.1% and 7.0% for the thalamus of the adult- hemispherectomized and adulthemidecorticated cats, respectively. Several factors may be involved in the atrophy of the contralateral thalamus. First, there must be a transsynaptic degeneration component mediated by the atrophy of the remaining neocortex Žsee below.; second, it is known that some neocortical areas project bilaterally to the thalamus, although the contralateral projections are sparse w47x; and, third, it is also known that there are bilateral connection between ventrobasal nuclei of the thalamus w9,34x, such that if one side heavily degenerates, there may be contralateral effects. Degenerative processes similar to the above in the thalamus should be responsible for the shrinkage of the remaining neocortex in the present cats since the removal of the left neocortex deprived, via corpus callosum interruption, the contralateral cortex of targets, contributing to a larger retrograde degeneration, and of axon terminals, contributing to a smaller anterograde degeneration component w31x. The shrinkage that occurred preferentially in caudal and central areas of the neocortex with little or no atrophy in frontal areas, was unexpected. However a similar trend has been reported by Kolb et al. w26x, Kolb and Tomie w27x, and Kolb and Whishaw w28x Žsee Section 4.2. in hemidecorticated rats. Since in normal cats there is a systematic relationship between the position of neurons in the neocortex and the location of their axons in the corpus callosum, such that the A–P axis of the callosum matches the A–P origin of the axons in the mantle w29x, a likely explanation is that there may be more interhemispheric fibers towards the splenium than towards the genu of the callosum Žwhich is supported by the finding w29x that the density of fibers in the splenium is greater than in the body.. 4.2. A critical maturational period (CMP) for reduced brain Õulnerability to injury The main contribution of the present experiments was finding an age-at-lesion limit for the protective effect of young age upon thalamic atrophy. In the preceding paper w57x, which presented an analysis of the behavioral results in the same cohorts of cats, we reported that for the majority of the behaviors tested there was a significant, abrupt increment in the impairments during the transition from the P30 to P60 groups. We interpreted that transition as the postnatal cut off point for a critical maturational period ŽCMP. for optimal post injury behavioral restoration w61x. Here, as for the behavioral results, we found an age-dependent progressive ten-

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

dency towards structure atrophy but there was no salient cut-off time point. It was still relevant though that for the P30 group the ipsilateral thalamic atrophy was among the smallest, while contralaterally the thalamus and the neocortex practically did not shrink. In addition, for the thalamus and neocortex contralateral to the resection the greatest increment in shrinkage among the P10 through P120 groups still occurred in the transition from the P30 to the P60 age groups. Therefore, the present morphometric data generally concur with the behavioral results in suggesting a cut-off point—between 30 through 60 postnatal days for postinjury restorative plasticity in the cat brain. The data also confirms that this favorable period for brain self repair does not end abruptly but is protracted in duration. In our previous analysis of cats with a unilateral frontal cortex resection during the last third of gestation w31x, we reported a 25.7% and a 26.5% reduction in volume for the ipsilateral thalamus and remaining neocortex, respectively, while, contralaterally, the thalamus tended to be smaller by 11.1% and the neocortex volume did not change. This contrasted with smaller atrophies in cats sustaining a similar lesion postnatally, at an age equivalent to that of the present P10 group w24x. In the latter cats the ipsilateral thalamus shrank by only 14.0% with no neocortical atrophy while contralaterally the thalamus shrank by 13.0% with no cortical change either. Differential behavioral w60,61x and cerebral metabolic w20x effects between these two age-at-lesion groups correlated well with the morphometry. ŽAs previously explained w60x, fetal cats do not tolerate hemispherectomy such that a smaller lesion had to be performed for the fetal versus postnatal comparisons.. In fetal-lesioned cats the thalamic cytoarchitecture was almost entirely preserved w30x such that the volume shrinkage could not be explained on the basis of retrograde–anterograde degeneration, as in our postnatal lesioned animals. Therefore, we concluded that there were cell losses Žboth of neuronal and glial cells. but that they were due to volume reduction of the structures involved and not to degeneration. Therefore, the morphological data in fetal-lesioned cats taken together with the results in postnatal subjects reported here, strongly suggest that for the long term postinjury structural atrophy and degeneration, there exists a discrete period of the total life span of the cat which is special, in that the outcome of neocortical brain lesions is better than that for damage occurring at any other life point. Because of the linkage with morphogenetic processes occurring at the same time, we have labeled this time span the Critical Maturational Period Žor CMP. for long term post injury brain restoration. By the time of our earliest postnatal lesions Žaround P10., the cat’s forebrain has already undergone peak neurogenesis and morphogenesis w33,50,67x, but morphogenesis is unfinished. For example, neurons of layers 2 and 3 of the visual cortex are generated between E45–58 but are still migrating at birth, reaching their final position by P21.

333

Thalamic morphogenesis is also protracted starting as early as E35 and finishing at birth w16,68x. However, two other maturational processes become more important during the early postnatal period, naturally occurring Žor programmed. cell death following overproduction of neurons, and excess synaptogenesis, followed by or overlapping with, pruning of synapses. Information on natural cell death, or apoptosis, in the cat is not abundant but the data suggest a postnatal time course w10,54x. Thus, in the suprasylvius gyrus and adjoining sulci, there is a 16% decrease in neuron numbers between P15–25; a slower loss occurs thereafter such that by E180 the decrease has reached about 21.0% w10x; the same happens in the temporal cortex w54x where apoptosis seems to peak between P7–12. There is abundant information regarding synaptogenesis and pruning in cats, particularly for the visual system w4,7,39,43,50,60,67x. Typically the number of synapses per neuron remains low during the late fetal period and first postnatal week, e.g., 7.0% of the adult in the supra sylvian gyrus at birth w4x; thereafter, there is an increase in synapse density lasting for about 3 weeks with a peak between P28 and P35 during which the synaptic density exceeds the adult values by 53 to 82% Žduring a similar period, in the lateral geniculate nucleus the synapse density increases by about four-fold w7x.. A slow decrease starts after this age reaching stable levels by P70. Another closely associated event which occurs at this time, is the development of dendritic fields. For example, by 2 weeks postnatally the basal dendrites of layer three pyramidal cells in the kitten visual cortex have all emerged from the soma but they have few branches; by 4 weeks dendrites have finished branching but continue elongating until reaching adult size by about week 5 w69x. Dendritic growth takes place earlier in layer 6 Žby more than 2 weeks, w32x. and also in layer 4 Žby about 1 week; w36x.. We have proposed w61x that the above stage of ontogenesis, i.e., when synapses and neurons numbers are being actively adjusted, provides the adequate plasticity required for an effective long-term restorative response of the brain to focal loss of tissue. On the one hand, based on our postnatal studies we have proposed w57,61x that the CMP for reduced vulnerability to brain injury, tends to end between P30 and P60. On the other hand, according to our studies in fetal versus neonatal cats, the beginning of the CMP occurs between E55, the older age of our fetal-lesioned animals, and P5–10, the younger age of our postnatal-lesioned cats. Other plasticity-related events also occur during this period, including Ža. peak sensitivity to modification of the ocular dominance properties of neurons of the visual cortex Žbetween 3 and 6 postnatal weeks; w21,41x.; Žb. increases in receptor binding for a number of neurotransmitters w51,66x; and Žc. increase in the level of several intracellular second messengers w2,8x. Thus, this discrete period may be of fundamental importance for a number of ‘critical’ cerebral ontogenetic events.

334

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

At this time of development ŽCMP. thalamic neurons may have axon branches which project collaterally to other sites than those included in the hemidecortication. Under normal conditions the exuberant projections become eliminated w22,54x, but they might persist if the cortically projecting axon are deprived of their targets and retract. Similarly, in the remaining neocortex, neurons with branched transcallosal and subcortical Žor intra cortical. projections Že.g., in the sensorimotor w23x and visual w22x cortices of the cat., may be sustained by their ipsilateral andror bilateral subcortical projections regardless the loss of their callosal branches. Furthermore, the process of reinnervation of subcortical targets by sensorimotor and visual cortical neurons, which we have shown to be prominent after neonatal hemispherectomy w1,58x, may help sustain both the reinnerÕated as well as the reinnerÕating neuron w37x. For example, after visual cortex removal in kittens, we found w1x that neurons of the remaining primary visual cortex, which develop novel decussated axon collaterals to reinnervate superior colliculus neurons on the side of the lesion, tend to increase in soma size. Ultimately, the very process of natural cell death might be directly down regulated if lesion-triggered neuron death occurs during this maturational period. 4.3. Comparisons with studies by other inÕestigators Although other investigators have attempted to assess cortical and subcortical degeneration after neocortical damage, reliable comparisons with our work are not possible because, as far as we know, measurements of structure volume or total number of cells have not been done. However, some of these studies merit comment. Kolb et al. w26x, Kolb and Tomie w27x, and Kolb and Whishaw w28x have extensively studied the effects of hemidecortication and of bilateral restricted cortical lesions in rats, with mixed results. Following hemidecortication at 1, 5 and 10 days of age and in adulthood, they found w27x that the thickness of the contralateral neocortex Žmeasured at five coronal planes. was normal for the P1 rats but was significantly smaller for two or three of the posterior planes for the other three groups. The cross-sectional area Žthree planes. showed that P1 animals had larger areas at two planes while the adult-lesioned had the smaller areas and the P5–P10 groups did not differ from controls. In all cases, the shrinkage was greater at the posterior plane, which coincided with our present results. Since cortical volumes were not estimated, the authors could only suggest that the shrinkage might indicate actual loss of cortical neurons or that, alternatively, it could ‘‘reflect a redistribution of cortical neurons as the remaining cortex is stretched . . . ’’. The results in the thalamus ipsilateral to the lesion were different. The cross-sectional areas showed a smaller thalamus for all groups lesioned at a young age with the thalamus of the adult-lesioned rats being larger compared to the neonatal groups Žalbeit smaller than controls.. As stated by the authors, these results were ‘‘exactly

opposite of our previous findings for animals with bilateral brain damage’’ w26,27x. In a recent paper w26x, the authors extended their studies of extensive bilateral cortical lesions to the effects of the same lesions but of smaller size and they reported similar results regardless of the lesion size. Contrary to hemidecortication, bilateral resections w26,28x resulted in greater thinning of the remaining cortex and a larger shrinkage of the thalamus in the rats lesioned closer to birth, while these effects were attenuated in the animals sustaining the lesions between P9 and P15 or between P7 and P9. The behavioral outcome generally matched the anatomical findings w26,28x. Perhaps considering that these results were obtained from a large number of rats, the authors did not mention w26x their hemidecortication results but emphasized the better outcome for lesions within the P7–P15 period. Consequently they proposed that in rats there is also a ‘critical’ life period for recovery, and that this appears to be about 7–10 days postnatal; they also stated that ‘‘earlier lesions allow for very little recovery’’ w26x. Thus, Kolb et al. agreed with our interpretation w61x that, since the rat brain is less mature at birth than cats’, the CMP in rats starts about a week after birth. The question of why their results were so different after hemidecortication as compared to bilateral lesions, remains elusive. It could be due to methodological problems. For example, we have shown that thickness is not a sufficient measure to evaluate cortical size w49x; in addition, if only cross-sectional areas of a structure are measured, it is still hard to make statements regarding actual degeneration unless total number of neurons or packing densities are measured w18,52,55,59x. It is also possible that the reorganizational brain processes following unilateral versus bilateral lesions might be different. In another study w53x hypoxia–ischemia were used to induce extensive hemisphere damage in 8 days old rats w53x. However, Ža. there was no description of what exactly was destroyed in each of the animals, Žb. only 3 weeks survival were allowed and it is known that degenerative effects continue after this time w34x; and Žc. only cortical thickness and cross-sectional areas were determined in two brain sections. The main reason to cite this study is their conclusion that the contralateral hemisphere was not affected by the lesion and that, therefore, that side is usable as a control. As a rule, this is not correct as attested by all our results and those of Kolb et al. Žincluding the asymmetries between the hemispheres in normal animals. and it is a notion that we repeatedly attempted to dispel w18,24,52,55,59x. As already mentioned, at birth the brain of rats is less mature than that of cats. Neurogenesis and morphogenesis are still active processes in forebrain structures and in some these continue through P5–10 w3x. Apoptosis in the neocortex w10x starts at about P2, peaks at P7 and declines during the second postnatal week to practically end by P30 Žsee also Ref. w42x.. Regarding synaptogenesis and prun-

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

ing, peak connectivity for transcallosal axons occurs during the first postnatal week with pruning continuing into the second week w23x, while geniculocortical visual projections are established between E19 and P3 with pruning continuing thereafter w5,25x. Thus, the morphological underpinnings of the CMP as defined earlier in Section 4 occur entirely postnatal in the rat and roughly peak at the time ŽP7–15. when the best outcome for brain lesions appears to occur for this species. Developmental studies of thalamic degeneration have also been conducted in monkeys. Goldman and Galkin w14x reported that, in monkeys with a prenatal uni- or bilateral resection of the presumptive dorsolateral prefrontal cortex at E102 or E106 Žduration of gestation, 165 days., the parvicellular division of the thalamic mediodorsal nucleus had a normal packing density of neurons while this was reduced in animals with postnatal lesions at P50 or later. The earlier lesioned animals also showed remarkable sparing of function during performance of learning tasks. There were no volume measurements and comparisons were done with the contralateral nucleus. Regardless of these problems, the authors concluded that degeneration did not occur in the earlier lesioned animals. In our studies in fetal cats w30x, we concluded that the thalamus had lost neuronal and glial cells and that this loss was proportional to the volume decrease and evolved in absence of signs of retrograde degeneration. In postnatal cats with a similar lesion at P9–14, there were minor cortical and thalamic morphological changes. Behaviorally, the neonatal-lesioned animals exhibited a superior performance compared to the fetal-lesioned cats. Thus, monkeys and cats with cortical lesions sustained within about the same fetal period Žend of second third of gestation, monkeys; beginning of last third of gestation, cats. showed quite opposite patterns of morphological and behavioral effects. We believe that the above differences in results in fetal cats versus fetal monkeys are trivial and can be easily explained on the basis of the different brain maturation states in the two species at the time of the lesion. The brain of rhesus monkeys is considerably more mature than that of cats during the time frame considered. By E102–106 peak neurogenesis and morphogenesis have already occurred in the forebrain w45x. Apoptosis in the visual cortex occurs relatively late w40x such that the numerical density of neurons decreases by about 36% between birth and 6 month of age. However, in the lateral geniculate nucleus apoptosis takes place much earlier since the neuronal population falls by about 30% during the middle third of gestation and ceases to diminish by E103 w65x. Regarding synaptogenesis and pruning, in the prefrontal cortex of monkeys w6x there is a rapid phase of synapse loss between E104 and P60 with a plateau between P60 and 3 years of age and a protracted loss thereafter Žup to 20 years of age.. In the visual cortex, rapid loss occurs between E60 and 2.5 yrs with a protracted further loss stopping at about 4 yrs of age w6x. Thus, the time for best outcome of cortical lesions

335

in monkeys appears to be within the same period of ontogenesis that we have called CMP in cats and rats, but whereas such a period takes place mostly after birth in cats, it occurs mostly during fetal life in monkeys. Cerebral hemispherectomy has become a routine operation during childhood as a therapy for a number of developmental pathologies w13,44x. It is, therefore, regrettable that, as far as we know, no systematic studies of subcortical or transhemispheric degeneration have been performed in the numerous available patients. However, given the similar patterns of ontogenesis and neural organization of the brain of higher mammals w3,11x, we would expect that same patterns of degeneration as seen in cats will be also found not only after hemispherectomy but after extensive unilateral stroke in patients. Indeed, based on the time course of ontogeny of the human brain and on behavioral results of human hemispherectomy, we have predicted Žsee preceding paper. a CMP for humans as well. Therefore, a thorough radiological morphometry of these patients’ brains would be of great help to understand the pathophysiology, time course and prognosis of extensive hemisphere damage in man, and it could also serve to assess the efficacy of neuroprotective agents and procedures. Acknowledgements We thank Louise D. Loopuijt, PhD and Marta Zagrebelsky, MD for their participation in the early stages of this study. Supported by grants NRSA T32HDO7416 ŽT. Schmanke., USPHS NS-25780, HD-04612, and HD-05958. References w1x P.D. Adelson, D.A. Hovda, J.R. Villablanca, K. Tatsukawa, Development of a crossed corticotectal pathway following cerebral hemispherectomy in cats: A quantitative study of the projection neuron, Dev. Brain Res. 86 Ž1995. 81–93. w2x C. Aoki, P. Siekevitz, Ontogenetic changes in the cyclic adenosine X X 3 ,5 -monophosphate stimulable phosphorylation of cats visual cortex proteins, particularly of microtubules associated protein 2 ŽMAP. 2: Effects of normal and dark rearing and of exposure to light, J. Neurosci. 2 Ž1982. 2465–2483. w3x Sh.B. Bayer, J. Altman, R.J. Russo, X. Zhang, Time tables of neurogenesis in the human brain based in experimentally determined patterns in the rat, Neurotoxicol. 14 Ž1993. 83–144. w4x C. Benhamida, Quantitative analysis of synaptogenesis in the cerebral cortex of the cat suprasylvian gyrus, Brain Res. Bull. 19 Ž1987. 567–579. w5x M.E. Blue, J.G. Parbavelas, The formation and maturation of synapses in the visual cortex of the rat, J. Neurocytol. 12 Ž1983. 599–616. w6x J.P. Bourgeois, P.S. Goldman-Rakic, P. Rakic, Synaptogenesis in the prefrontal cortex of rhesus monkeys, Cereb. Cortex 4 Ž1994. 78–96. w7x B.G. Cragg, The development of synapses in the visual system of the cat, J. Comp. Neurol. 160 Ž1975. 147–168. w8x S.M. Dudek, M.F. Bear, A biochemical correlate of the cortical period for synapse modification in the visual cortex, Science 246 Ž1989. 673–675.

336

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337

w9x E. Fadiga, T. Manzoni, Relationships between the somatosensory thalamic relay nuclei of the two sides, Arch. Ital. Biol. 107 Ž1969. 604–632. w10x M. Ferrer, M. Hernandez-Marti, E. Bernet, S. Alcantara, C. Auladell, ´ ´ Cell death and removal in the cerebral cortex during development, Prog. Neurobiol. 39 Ž1992. 1–43. w11x B.L. Finlay, R.B. Darlington, Linked irregularities in the development and evolution of mammalian brains, Science 268 Ž1995. 1578–1584. w12x R.S. Fisher, M.K. Boylan, C.D. Hull, N.A. Buchwald, M.S. Levine, Branched projections of cat sensorimotor cortex: Multiple retrograde labeling via commissural corticortical, decussated corticostriatal and undecussated corticostriatal axons, Brain Res. 384 Ž1986. 395–400. w13x J.M. Freeman, E.P.G. Vining, Hemispherectomy—the ultimate focal resection, in: H. Luders ŽEd.., Epilepsy Surgery, Raven Press, New York, 1991, pp. 111–118. w14x P.S. Goldman, T.W. Galkin, Prenatal removal of frontal association cortex in the rhesus monkey: Anatomical and functional consequences in postnatal life, Brain Res. 152 Ž1978. 451–485. w15x W.L. Hays, Statistics for the Social Sciences, Holt, Rinehart and Winston, New York, 1988. w16x T.L. Hickey, P.F. Hitchcock, Genesis of neurons in the dorsal lateral geniculate nucleus of the cat, J. Comp. Neurol. 228 Ž1984. 186–199. w17x D.A. Hovda, H.t. Chugani, J.R. Villablanca, B. Badie, R.L. Sutton, Maturation of cerebral oxidative metabolism in the cat: A cytochrome oxidase histochemistry study, J. Cereb. Blood Flow Metab. 12 Ž1992. 1039–1048. w18x D.A. Hovda, J.R. Villablanca, Quantitative study of neural degeneration following adult or neonatal cerebral hemispherectomy in cats: I. Retrograde effects in the medial geniculate thalamic nucleus, Brain Dysfunct. 2 Ž1989. 221–236. w19x D.A. Hovda, J.R. Villablanca, H.T. Chugani, M.E. Phelps, Cerebral metabolism following adult or neonatal hemineodecortication in cats: I. Effects on glucose metabolism using w 14 Cx2-deoxy-D-glucose autoradiography, J. Cereb. Blood Flow Metab. 16 Ž1996. 134–146. w20x D.A. Hovda, J.R. Villablanca, T. Schmanke, A neocortical lesion sustained during the late fetal period induced a widespread enduring reduction of cerebral metabolism in cats, Soc. Neurosci. Abstr. 22 Ž1996. 305. w21x D.H. Hubel, T.N. Wiesel, The period of susceptibility to the physiological effects of unilateral eye closure in kittens, J. Physiol. 206 Ž1970. 419–436. w22x G.M. Innocenti, Transitory structures as substrates for developmental plasticity of the brain, Dev. Neurosci. 13 Ž1981. 305–333. w23x J.O. Ivy, H.P. Killackey, Ontogenetic changes in the projections of neocortical neurons, J. Neurosci. 2 Ž1982. 725–743. w24x J.F. Jackson III, J.R. Villablanca, L.D. Loopuijt, Few neocortical and thalamic morphological changes after a neonatal frontal cortical ablation contrast with the effects of similar lesions in fetal cats, Dev. Brain Res. 90 Ž1995. 62–72. w25x G.H. Kageyama, R.T. Robertson, Development of geniculocortical projections to the visual cortex in rats: Evidence for early ingrowth and synaptogenesis, J. Comp. Neurol. 335 Ž1993. 123–148. w26x B. Kolb, B. Petri, J. Cioe, Recovery from early cortical damage in rats: VII. Comparison of the behavioral and anatomical effects of medial prefrontal lesions at different ages of neural maturation, Behav. Brain Res. 79 Ž1996. 1–13. w27x B. Kolb, J.A. Tomie, Recovery from early cortical damage in rats: IV. Effects of hemidecortication at 1, 5 or 10 days on cerebral anatomy and behavior, Behav. Brain Res. 28 Ž1988. 254–259. w28x B. Kolb, I.Q. Whishaw, Plasticity of the neocortex: Mechanisms underlying recovery from early brain damage, Prog. Neurobiol. 32 Ž1989. 235–276. w29x S.G. Lomber, B.R. Payne, A.C. Rosenquist, The spatial relationship between the cerebral cortex and fiber trajectory through the corpus callosum of the cat, Behav. Brain Res. 64 Ž1994. 25–35. w30x L.D. Loopuijt, J.R. Villablanca, D.A. Hovda, Morphological changes

w31x

w32x

w33x w34x

w35x

w36x

w37x

w38x w39x

w40x

w41x w42x

w43x w44x

w45x w46x w47x

w48x

w49x

w50x

w51x

in the thalamus and neocortex of the cat brain after a restricted unilateral fetal cortical lesion, Dev. Brain Res. 85 Ž1995. 259–272. L.D. Loopuijt, J.R. Villablanca, E. Huang, S. Mancuso, D.A. Hovda, The effects of neocortical lesions in the number of cells of the neonatal or adult feline caudate nucleus: Comparisons to fetal lesions, Neuroscience 77 Ž1997. 403–418. J. Lubke, K. Albus, The postnatal development of layer VI pyramidal neurons in the cat’s striate cortex as visualized by intracellular Lucifer yellow injections in aldehyde-fixed tissue, Dev. Brain Res. 45 Ž1989. 29–38. M.B. Luskin, C.J. Shatz, Neurogenesis of the cat’s primary visual cortex, J. Comp. Neurol. 242 Ž1995. 611–631. G. Macchi, A. Quattrini, P. Chinzari, G. Marchesi, G. Capocci, Quantitative data on cell loss and cellular atrophy of intralaminar nuclei following cortical and subcortical lesions, Brain Res. 89 Ž1975. 43–59. M. Madarasz, J. Somogyi, V.L. Silakor, J. Hamori, Residual neurons in the lateral geniculate nucleus of adult cats following chronic disconnection from the neocortex, Exp. Brain Res. 52 Ž1983. 363– 374. G. Meyer, R. Ferres-Torres, Postnatal maturation of non-pyramidal neurons in the visual cortex of the cat, J. Comp. Neurol. 228 Ž1984. 226–244. E.H. Murphy, R. Kalil, Functional organization of lateral geniculate cells following removal of the visual cortex in the newborn kitten, Science 206 Ž1979. 713–716. J.L. Myers, Fundamentals of Experimental Design, Allyn and Bacon, Boston, 1979. J.R. O’Kusky, Synapse elimination in the developing visual cortex: A morphometric analysis in normal and dark-reared cats, Dev. Brain Res. 22 Ž1985. 81–91. O’Kusky, M. Colonnier, Postnatal changes in the number of neurons and synapses in the visual cortex Žarea 17. of the macaque monkey: Stereological analysis in normal and monocularly deprived animals, J. Comp. Neurol. 210 Ž1962. 291–306. C.R. Olson, R.D. Freeman, Profile of the sensitive period for monocular deprivation in kittens, Exp. Brain Res. 39 Ž1980. 17–21. M. Oudega, S. Varon, T. Hagg, Distribution of corticospinal neurons in the postnatal rat: Quantitative evidence for massive collateral elimination and modest cell death, J. Comp. Neurol. 347 Ž1994. 115–126. B.R. Payne, H.A. Pearson, P. Cornell, Neocortical connections in fetal cats, Neurosci. Res. 5 Ž1988. 513–543. W.J. Peakock, M. Wheby-Grant, W.D. Shields, D.A. Shewmon, H.T. Chugani, R. Sankar, H.V. Vinters, Hemispherectomy for intractable seizures in children: A report of 58 cases, Child’s Nerv. Syst. 12 Ž1996. 376–384. P. Rakic, Specification of cerebral cortical areas, Science 241 Ž1988. 176–179. F. Reinoso-Suarez, Topographyscher Hirnatlas der Katze, E. Merck, ´ Darmstadt, 1961. E. Rinvik, The corticothalamic projection from the pericruciate and coronal gyri in the cat. An experimental study with silver impregnation methods, Brain Res. 100 Ž1968. 79–119. T.D. Schmanke, J.R. Villablanca, V. Lekht, Differential shrinkage of the remaining neocortex in cats with hemineodecortication sustained neonatally or in adulthood, Soc. Neurosci. Abst. 22 Ž1996. 306. T.D. Schmanke, J.R. Villablanca, Age-dependent effects of hemineodecortication upon contralateral cortical thickness: Comparison with other measures of cortical size, submitted. C.J. Shatz, M.B. Luskin, The relationship between the geniculocortical afferents and their cortical targets cells during development of the cats primary visual cortex, J. Neurosci. 6 Ž1986. 3655–3668. C. Shaw, M.C. Needler, M. Cynander, Ontogenesis of muscarinic acetylcholine binding sites in cat visual cortex: Reversal of the specific laminar distribution during the critical period, Dev. Brain Res. 14 Ž1984. 295–299.

T.D. Schmanke et al.r DeÕelopmental Brain Research 105 (1998) 325–337 w52x B.L. Shook, J.R. Villablanca, Quantitative cytoarchitectural analysis of cellular degeneration in the dorsal lateral geniculate nucleus of cats and kittens with cerebral hemispherectomy, Exp. Neurol. 111 Ž1991. 80–94. w53x J. Towfight, C. Housman, R.C. Vannucci, D.F. Heitjan, Effect of unilateral hypoxicischemic brain damage on the gross development of the opposite cerebral hemisphere, Biol. Neonate 65 Ž1994. 108– 118. w54x F. Valverde, M.V. Facal-Valverde, Transitory population of cells in the temporal cortex of kittens, Dev. Brain Res. 32 Ž1987. 283–288. w55x J.R. Villablanca, J.W. Burgess, F. Benedetti, There is less thalamic degeneration in neonatal-lesioned than in adult-lesioned cats after cerebral hemispherectomy, Brain Res. 368 Ž1986. 211–225. w56x J.R. Villablanca, J.W. Burgess, B.J. Sonnier, Neonatal cerebral hemispherectomy: A model for postnatal reorganization of the brain, in: S. Finger, C.R. Almli ŽEds.., The Behavioral Biology of Early Brain Damage, Academic Press, New York, 1984, pp. 179–210. w57x J.R. Villablanca, P. Carlson-Kuhta, T. Schmanke, D.A. Hovda, A critical maturational period for reduced brain vulnerability to developmental injury: I. Behavioral studies in cats, Dev. Brain Res. 105 Ž1998. 309–324. w58x J.R. Villablanca, F. Gomez-Pinilla, Novel crossed corticothalamic ´ projections after neonatal cerebral hemispherectomy. A quantitative autoradiography study in cats, Brain Res. 410 Ž1987. 219–231. w59x J.R. Villablanca, D.A. Hovda, Quantitative study of neural degeneration following neonatal or adult cerebral hemispherectomy in cats: II. Transsynaptic effects in the superior colliculus and mammillary nuclei, Brain Dysfunct. 2 Ž1989. 237–254. w60x J.R. Villablanca, D.A. Hovda, G.F. Jackson, R. Gayek, Neurological and behavioral effects of a unilateral frontal, cortical lesion in fetal

w61x

w62x

w63x w64x

w65x

w66x

w67x

w68x

w69x

337

kittens: I. Brain morphology, movements, posture and sensorimotor tests, Behav. Brain Res. 57 Ž1993. 63–77. J.R. Villablanca, D.A. Hovda, J.F. Jackson, J. Infante, Neurological and behavioral effects of a unilateral frontal cortical lesion in fetal kittens: II. Visual tests and proposing an ‘Optimal Developmental Period’ for lesion effects, Behav. Brain Res. 57 Ž1993. 79–92. J.R. Villablanca, D.A. Hovda, T.D. Schmanke, Regional cerebral metabolic rates reflect age-at-lesion differences in hemineodecorticated cats, Soc. Neurosci. Abstr. 23 Ž1997. 1311. J.R. Villablanca, Ch.E. Olmstead, The striatum: A fine tuner of the brain, Acta Neurobiol. Exp. 42 Ž1962. 225–297. J.R. Villablanca, H.M. Patel, T.D. Schmanke, Age-at-lesion reduction in bilateral thalamic volumes in hemineodecorticated cats, Soc. Neurosci. Abstr. 22 Ž1996. 305. R.W. Williams, P. Rakic, Elimination of neurons from the rhesus monkey’s lateral geniculate nucleus during development, J. Comp. Neurol. 272 Ž1989. 424–436. M. Wilkinson, C. Shaw, I. Khan, M. Cynander, Ontogenesis of beta-adrenergic binding sites in kitten visual cortex and the effects of visual deprivation, Dev. Brain Res. 7 Ž1983. 349–352. D.A. Windfield, The postnatal development of synapses in different laminae of the visual cortex in the normal kitten and in kittens with eyelid suture, Dev. Brain Res. 9 Ž1983. 155–169. S.P. Wise, S.H.C. Hendry, E.G. Jones, Prenatal development of sensorimotor cortical projections in cats, Brain Res. 38 Ž1977. 538–544. N. Zec, S. Bliss-Tieman, Development of the dendritic fields of layer 3 pyramidal cells of kitten’s visual cortex, J. Comp. Neurol. 339 Ž1994. 288–300.