Adjustment of connectivity in rat neocortex after prenatal destruction of precursor cells of layers II-IV

Prenatal exposure to permethrin influences vascular development of fetal brain and adult behavior in mice offspring

Pyrethroids are one of the most widely used classes of insecticides and show neurotoxic effects that induce oxidative stress in the neonatal rat brain. However, little is still known about effects of prenatal exposure to permethrin on vascular development in fetal brain, central nervous system development, and adult offspring behaviors. In this study, the effects of prenatal exposure to permethrin on the development of cerebral arteries in fetal brains, neurotransmitter in neonatal brains, and locomotor activities in offspring mice were investigated. Permethrin (0, 2, 10, 50, and 75 mg/kg) was orally administered to pregnant females once on gestation day 10.5. The brains of permethrin-treated fetuses showed altered vascular formation involving shortened lengths of vessels, an increased number of small branches, and, in some cases, insufficient fusion of the anterior communicating arteries in the area of circle of Willis. The prenatal exposure to permethrin altered neocortical and hippocampus thickness in the mid brain and significantly increased norepinephrine and dopamine levels at postnatal day 7 mice. For spontaneous behavior, the standing ability test using a viewing jar and open-field tests showed significant decrease of the standing ability and locomotor activity in male mice at 8 or 12 weeks of age, respectively. The results suggest that prenatal exposure to permethrin may affect insufficient development of the brain through alterations of vascular development.

Postsynaptic currents prior to onset of epileptiform activity in rat microgyria

Structural malformations of the cortex, arising as a result of genetic mutation or injury during development are associated with dyslexia, epilepsy, and other neurological deficits. We have used a rat model of a microgyral malformation to examine mechanisms of epileptogenesis. Our previous studies showed that the frequency of miniature excitatory postsynaptic currents (mEPSCs) recorded in neocortical layer V pyramidal neurons is increased in malformed cortex at a time when field potential epileptiform events can be evoked. Here we show that the increase occurs at an age before onset of cortical epileptiform activity and at a time when the frequency of mEPSCs in control layer V pyramidal neurons is stable. An increase in the frequency of spontaneous (s)EPSCs in layer V pyramidal neurons of malformed cortex occurs earlier than that for mEPSCs, suggesting that there may additionally be alterations in intrinsic properties that increase the excitability of the cortical afferents. Frequencies of EPSC bursts and late evoked activity were also increased in malformed cortex. These results suggest that a hyperinnervation of layer V pyramidal neurons by excitatory afferents occurs as an active process likely contributing to subsequent development of field epileptiform events.

N-cadherin regulates ingrowth and laminar targeting of thalamocortical axons

Thalamocortical axons are precisely targeted to cortical layer IV, but the identity of specific molecules that govern the establishment of laminar specificity in the thalamocortical projection has been elusive. In this study, we test the role of N-cadherin, a homophilic cell adhesion molecule, in laminar targeting of thalamocortical axons using cocultured thalamic and cortical slice explants exposed to N-cadherin function-blocking antibodies or inhibitory peptides. In untreated cocultures, labeled thalamocortical axons normally grow to and stop in layer IV, forming terminal-like arbors. In the N-cadherin-blocked cocultures, thalamic axons reach layer IV by growing through deep layers at the same rate as those in the untreated cocultures, but instead of terminating in layer IV, they continue growing uninterruptedly through layer IV and extend into supragranular layers to reach the outermost cortical edge, where some form terminal-like arbors in this aberrant laminar position. In cocultures in which the cortical slice is taken at an earlier maturational stage, one that corresponds to a time when thalamic axons are normally growing through deep layers before the emergence of layer IV from the cortical plate, thalamic axon ingrowth through deep layers is significantly attenuated by N-cadherin blocking reagents. These data indicate that N-cadherin has multifaceted roles in establishing the thalamocortical projection, governing aspects of both thalamic axon ingrowth and laminar targeting by acting as a layer IV stop signal, which progressively change in parallel with the maturational state of the cortex.

A radialization factor in normal cortical plate restores disorganized radial glia and disrupted migration in a model of cortical dysplasia

Treatment of pregnant ferrets on embryonic day 24 (E24) with the antimitotic methylazoxy methanol (MAM) leads to a specific constellation of effects in newborn kits, which include a very thin and poorly laminated neocortex, disruption of radial glial cell morphology with early differentiation into astrocytes, and abnormal positioning of Cajal-Retzius cells. We suggest that MAM treatment on E24 results in this model of cortical dysplasia by eliminating a population of cells that produce a factor capable of maintaining radial glia in their normal morphology. The abnormal radial glia, either alone or in combination with other abnormal features, are likely to prevent proper migration into the cortical plate. To test the possibility that normal cortex can provide the missing substance that influences radial glia, slices of E24 MAM-treated cortex were removed at postnatal day 0 (P0) and cultured adjacent to explants of P0 normal cortical plate. By labelling a small number of cells with injections of fluorescent dextrans into the cultured slices, we found that abnormal radial glia in MAM treated slices cocultured adjacent to normal cortical plate were restored toward normal, in comparison to E24 MAM treated slices cultured alone and in other control conditions. We also found that abnormally positioned Cajal-Retzius cells move into the marginal zone and that neurons are able to migrate into the cortical plate more effectively in the coculture condition. These data indicate that normal cortical plate of ferrets contains a factor causing radial glia to maintain their elongated morphology; the improved position of radial glia encourages repositioning of Cajal-Retzius cells and improved neuronal migration into the cortical plate.

NPY sensitivity and postsynaptic properties of heterotopic neurons in the MAM model of malformation-associated epilepsy

Neuronal migration disorders (NMDs) can be associated with neurological dysfunction such as mental retardation, and clusters of disorganized cells (heterotopias) often act as seizure foci in medically intractable partial epilepsies. Methylazoxymethanol (MAM) treatment of pregnant rats results in neuronal heterotopias in offspring, especially in hippocampal area CA1. Although the neurons in dysplastic areas in this model are frequently hyperexcitable, the precise mechanisms controlling excitability remain unclear. Here, we used IR-DIC videomicroscopy and whole cell voltage-clamp techniques to test whether the potent anti-excitatory actions of neuropeptide Y (NPY) affected synaptic excitation of heterotopic neurons. We also compared several synaptic and intrinsic properties of heterotopic, layer 2-3 cortical, and CA1 pyramidal neurons, to further characterize heterotopic cells. NPY powerfully inhibited synaptic excitation onto normal and normotopic CA1 cells but was nearly ineffective on responses evoked in heterotopic cells from stimulation sites within the heterotopia. Glutamatergic synaptic responses on heterotopic cells exhibited a comparatively small, D-2-amino-5-phosphopentanoic acid-sensitive, N-methyl-D-aspartate component. Heterotopic neurons also differed from normal CA1 cells in postsynaptic membrane currents, possessing a prominent inwardly rectifying K(+) current sensitive to Cs(+) and Ba(2+), similar to neocortical layer 2-3 pyramidal cells. CA1 cells instead had a prominent Cs(+)- and 4-(N-ethyl-N-phenylamino)-1,2-dimethyl-6-(methylamino) pyrimidinium chloride-sensitive I(h) and negligible inward rectification, unlike heterotopic cells. Thus heterotopic CA1 cells appear to share numerous physiological similarities with neocortical neurons. The lack of NPY's effects on intra-heterotopic inputs, the small contribution of I(h), and abnormal glutamate receptor function, may all contribute to the lowered threshold for epileptiform activity observed in hippocampal heterotopias and could be important factors in epilepsies associated with NMDs.

Disruption of layers 3 and 4 during development results in altered thalamocortical projections in ferret somatosensory cortex

The precision of projections from dorsal thalamus to neocortex are key toward understanding overall cortical organization and function. To identify the significance of layer 4 cells in receiving the bulk of thalamic projections in somatosensory cortex, we disrupted layer 4 genesis and studied the effect on thalamic terminations in ferrets. Second, we ascertained the result of layer 4 disruption on functional responses and topographic organization. Methylazoxy methanol (MAM) was injected into pregnant ferrets on embryonic day 33 (E33), when most layer 4 neurons of somatosensory cortex are generated. This treatment resulted in dramatic reduction in the thickness of targeted layer 4. E38 MAM treatment was used as a control, when layer 2-3 neurons are generated. The projections of ventrobasal thalamus into somatosensory cortex were studied using DiI injections. We found only subtle differences between groups (normal, E33, or E38 MAM-treated) in the thalamic afferent pattern on postnatal day 1 (P1) and P7. On P14, thalamic terminations distribute almost equally throughout the remaining cortical layers in the E33 MAM-treated group compared with normal and E38 MAM-treated animals, in which the ventrobasal thalamus projects primarily to central layers. Electrophysiological recordings conducted on mature ferrets treated with MAM on E33 demonstrated that somatotopic organization and receptive field size are normal. These findings emphasize the importance of layer 4 in determining the normal laminar pattern of thalamic termination and suggest that, although its absence is likely to impact on complex neocortical functional responses, topographic organization does not arise from the influence of layer 4.

In utero irradiation of rats as a model of human cerebrocortical dysgenesis: a review

Certain developmental abnormalities of the cerebral cortex are closely associated with epilepsy in humans. Exposure of fetal rats to external gamma-irradiation produces diffuse cortical dysplasia and neuronal heterotopia. These abnormalities are the result of radiation-induced cell death coupled with continued cortical development in an altered cellular environment. In vivo electroencephalography studies in these animals have revealed an increased propensity for electrographic seizures in the presence of the sedating agents, acepromazine and xylazine. In vitro neocortical slices containing dysplastic cortex demonstrate enhanced excitability, as compared to control neocortex, when inhibition that is mediated by the A-type gamma-amino butyric acid receptor is blocked with bicuculline methiodide. In utero irradiation of rats produces structural changes that mimic some aspects of cerebral dysgenesis in humans and results in physiologic changes that increase the animals' propensity for seizures. Similarities and differences between the animal model and the human syndromes are discussed.

Altered connections between neocortical and heterotopic areas in methylazoxymethanol-treated rat

We are currently investigating various treatments which could determine, in the rat brain, structural abnormalities mimicking those reported in human brain dysgeneses. We can induce the formation of neuronal heterotopia in the progeny of rats by means of a double injection of the cytotoxic agent methylazoxymethanol acetate (MAM) on embryonic day 15. We have now investigated the anatomical connections of these heterotopia by means of anterograde and retrograde tract tracing techniques. The induced heterotopia along the border of the lateral ventricles shared common anatomical features with the periventricular nodules in human periventricular or subcortical nodular heterotopia (PNH). The tract tracing data demonstrated the existence of reciprocal connections between the neuronal heterotopia and the ipsilateral and contralateral cortical areas, and the presence of abnormal cortico-hippocampal and cortico-cortical connections. On the basis of the connectivity patterns, it may be speculated that some cells in the heterotopia could be neurons originally committed to the cortex, that were interrupted in their migration by the MAM treatment. Given the common morphological features seen in human PNH and MAM-induced brain heterotopia, the anatomical and developmental analysis of MAM-treated rats may shed light on the mechanisms by which human brain dysgeneses develop in human patients.

Neocortex in the hippocampus: an anatomical and functional study of CA1 heterotopias after prenatal treatment with methylazoxymethanol in rats

Migration disorders cause neurons to differentiate in an abnormal heterotopic position. Although significant insights have been gained into the etiology of these disorders, very little is known about the anatomy of heterotopias. We have studied heterotopic masses arising in the hippocampal CA1 region after prenatal treatment with methylazoxymethanol (MAM) in rats. Heterotopic cells were phenotypically similar to neocortical supragranular neurons and exhibited the same temporal profile of migration and neurogenesis. However, they did not express molecules characteristic of CA1 neurons such as the limbic-associated membrane protein. Horseradish peroxidase injections in heterotopia demonstrated labeled fibers not only in the neocortex and white matter but also in the CA1 stratum radiatum and stratum lacunosum. To study the pathophysiological consequences of this connectivity, we compared the effects of neocortical and limbic seizures on the expression of Fos protein and on cell death in MAM animals. After metrazol-induced seizures, Fos-positive cells were present in CA1 heterotopias, the only hippocampal region to be activated with the neocortex. By contrast, kainic acid-induced seizures caused a prominent delayed cell death in limbic regions and in CA1 heterotopias. Together, these results suggest that neocortical heterotopias in the CA1 region are integrated in both the hippocampal and neocortical circuitry.

Effect of chronic amitriptyline administration on serotonergic receptors in rats with methylazoxymethanol-induced microencephaly

Methylazoxymethanol (MAM)-induced cortical hypoplasia resulted in a 20% decrease in the Bmax of 5-HT2A receptors in the frontal cortex with no change in the Bmax of 5-HT1A receptors. Chronic treatment with amitriptyline did not further decrease the Bmax of 5-HT2A receptors in the MAM-lesioned cortex, suggesting that the persistent down-regulation of cortical 5-HT2A receptors in MAM-lesioned rats was induced by serotonergic hyperinnervation.

Increased seizure susceptibility in adult rats with neuronal migration disorders

Recent data show that neuronal migration disorders (NMD) lower the seizure threshold in the immature brain. To assess if this is an age-related phenomenon, kainic acid (KA) was administered to induce status epilepticus in adult rats with NMD. Results of the present study demonstrate that adult rats with NMD had a shorter latency to seizures and longer duration of status epilepticus compared to age-related controls. Furthermore, in rats with NMD seizures were more severe and status epilepticus-induced mortality was worse than in age-matched controls. These data confirm that NMD lower the seizure threshold in the adult rat. The results of the present study combined with our previous studies in the immature rat, suggest that the facilitating effects of NMD on seizures are not age dependent.

Radiation-induced, lamina-specific deletion of neurons in the primate visual cortex

To examine the early determinants of cortical cytoarchitecture, we deleted specific neuronal classes in the primate visual cortex by ionizing irradiation at selected prenatal stages. Multiple doses of X-rays were delivered to the macaque monkey brain between embryonic day (E) 80 and E90 to block the division of cells destined to populate the superficial cortical layers, between E70 and E79 to eliminate neurons destined for the middle layers; and between E33 and E40 to delete neurons destined for the lateral geniculate nucleus (LGN) that project to the cortex. All animals were killed after birth, and their brains were processed for histological and electron microscopic analyses. Cell density and number in the LGN and visual cortex were determined by using three-dimensional, computer-aided morphometry. In animals irradiated with low doses (total of approximately 200 cGy) during the genesis of the LGN but before the onset of corticogenesis (E33-40), the LGN was reduced in both volume and number of neurons. Area 17 in these animals displayed only slight changes in cortical thickness, cell density, and area-specific cytoarchitectonic features, whereas the total surface devoted to area 17 was significantly diminished. In contrast, animals irradiated with low doses during the period of corticogenesis, after the completion of the LGN genesis, showed no significant change in the volume of the LGN or in the number of its cells. Moreover, in these animals, the surface of area 17 was not significantly altered, although the cortical layers generated at the time of irradiation had a significantly lower density and total number of cells, whereas the layers generated before and after the period of irradiation were spared. In contrast, cases exposed to high doses of X-ray (total > 300 cGy) showed more severe effects, including all layers. However, layers normally generated during irradiation were depleted and consisted of cell-sparse strata populated by densely packed neuropil (axons, small dendrites, dendritic spines, and synaptic boutons). These cell-sparse strata were situated deeper in the early irradiated animals than in the later irradiated animals, and their laminar position changed abruptly at the area 17/18 border. These results show that low doses of irradiation in a slowly developing primate brain can be used effectively to eliminate targeted classes of neurons before they reach their final position, providing an opportunity to examine the role of cell-cell interactions in the formation of circuitry and the role of specific cell classes in cortical development.

Short exposure to methylazoxymethanol causes a long-term inhibition of axonal outgrowth from cultured embryonic rat hippocampal neurons

Methylazoxymethanol (MAM) is an alkylating agent that is used to induce microencephaly by killing mitotically active neuroblasts. We found that at later developmental times, MAM exposure can result in abnormal fiber growth in vivo. However, there have not been any previous studies on the effects of MAM on differentiating neurons. We examined the outcome of short exposure to MAM on postmitotic embryonic hippocampal cultures during the establishment of axonal polarity. At 0, 1, or 2 days in vitro (DIV), neurons were treated with 0.1 nM-1 microM MAM for 3 hr and then transferred to glial conditioned media. At 3 DIV, the cells were fixed and analyzed by immunofluorescent staining for neuron viability and differentiation. Control cells initiate several minor processes; one process elongates rapidly at about 1 DIV eventually becoming an axon, while extensive dendritic growth occurs after 3-4 DIV. Neurons treated with 1 microM MAM at 0 or 1 DIV showed a marked inhibition of neurite growth and withdrawal of axons without affecting cell viability. These cells continued to show minimal neurite outgrowth at 7 DIV, even when transferred to a glial coculture. In contrast, cells treated initially with MAM, after neuronal polarity is established at 2 DIV, showed no effect on axonal growth. To determine the effects of MAM on the neuronal cytoskeleton, we examined the in vitro assembly of brain microtubules in a one cycle assay. Exposure to MAM depleted the soluble pool of proteins, including microtubule-associated protein 1B (MAP1B) and MAP2, which are required for neurite outgrowth, through a nonspecific process. Under non-saturating conditions, there were no changes in the total amount of microtubules assembled or the coassembly of MAP1B and MAP2 in the presence of MAM. These results demonstrate that MAM can directly affect differentiating neurons, indicating that an early disruption of axonal outgrowth may have long-term effects.

Trends in the anatomical organization and functional significance of the mammalian thalamus

The last decade has witnessed major changes in the experimental approach to the study of the thalamus and to the analysis of the anatomical and functional interrelations between thalamic nuclei and cortical areas. The present review focuses on the novel anatomical approaches to thalamo-cortical connections and thalamic functions in the historical framework of the classical studies on the thalamus. In the light of the most recent data it is here discussed that: a) the thalamus can subserve different functions according to functional changes in the cortical and subcortical afferent systems; b) the multifarious thalamic cellular entities play a crucial role in the different functional states.

Wiring, dysmorphogenesis and epilepsy: a hypothesis

Cerebral cortical dysgenesis has been found by magnetic resonance imaging to be the second most common pathology underlying medically refractory chronic partial epilepsy. Patients with the latter condition form the largest group in specialist epilepsy clinics. The pathogenesis of the epilepsy in cortical dysgenesis remains largely obscure. The most popular current hypothesis holds neuronal misconnection secondary to neuronal malpositioning culpable for seizure activity. However, a review of the published literature of cortical dysgenesis and an analysis of newer magnetic resonance and histopathological data, suggests that this view is no longer tenable. A modified hypothesis is proposed in which neuronal connectivity itself is postulated to be the primary motive force in both cerebral morphogenesis and epileptogenesis in cases of cortical dysgenesis. This hypothesis leads to the generation of a model for cortical development and directly testable predictions of intercellular connectivity, as well as a potential tool for the prediction of the possibility of freedom from seizure activity after surgical resection of dysgenetic lesions in individual cases.

Seizure susceptibility in immature rats with micrencephaly induced by prenatal exposure to methylazoxymethanol acetate

The administration of the alkylating neurotoxin methylazoxymethanol acetate (MAM) to pregnant rats on day 15 of gestation induces, in the offspring, a marked micrencephaly, characterized by an impaired formation of interneurons at cortical, hippocampal and striatal levels. Since in man developmental CNS malformations are often associated with severe epileptogenic encephalopathies with seizures appearing in the first months or years of life, we have studied the development of kainic-acid- and bicuculline-induced seizures in 15- and 30-day-old rats, prenatally exposed to MAM. Compared to controls, a higher susceptibility to seizures has been found in micrencephalic rats aged 15 days, while no significant differences have been observed in those aged 30 days. It is hypothesized that the cerebral global anatomical dysgenesis caused by MAM underlies the higher seizure susceptibility shown by animals during the first periods of life. Successively, the processes of adjustment occurring between the cerebral regions affected by the neurotoxic action of MAM and the afferent and efferent pathways spared by the substance may re-establish adequate interneuronal relationships and, therefore, a normal convulsive susceptibility.

Immunohistochemical localization of calbindin-D28K in telencephalic regions of microencephalic rats

The localization of calbindin neurons was studied in different brain areas of rats rendered microencephalic by gestational methylazoxymethanol acetate (MAM) treatment. In layers VI and V of the cortex, the only recognizable layers in MAM-treated rats, a higher density of calbindin interneurons and an apparent increase in protein expression was observed. In the hippocampus, calbindin pattern was essentially preserved, despite the dramatic decrease in size. In other telencephalic regions, calbindin distribution was not changed except for the septum, where a large increase of calbindin neurons was observed. The present results suggest that the MAM model may be used to investigate the role of calbindin.

Stages and patterns of centrifugal arrest of diffuse neuronal migration disorders

Centrifugal migration of newly generated neuroblasts toward the cortical surface can be arrested at different levels, resulting in anatomically different disorders. This is illustrated by three patients with diffuse or generalized neuronal migration disorders (NMDs): pachygyria, subcortical laminar heterotopia ('double-cortex' syndrome), and periventricular laminar heterotopia. All had medically intractable partial and generalized epileptic seizures, but there was no close correlation between the type of dysplasia and intelligence or clinical pattern. The low intelligence found in these patients may relate to the epileptic syndrome, rather than to the NMDs alone. MRI suggested that in this spectrum of disorders the migration process was arrested at different stages, depending on the extent, timing and site of damage to radial glial fibres.

The fates of cells in the developing cerebral cortex of normal and methylazoxymethanol acetate-lesioned mice

We are interested in the mechanisms that generate the mature cerebral cortex. We used bromodeoxyuridine (BrdU) to label cortical cells as they were being born. We followed the fates of specific sets of cortical precursors in normal mice and in mice in which other groups of cortical progenitors had been destroyed with the antimitotic agent methylazoxymethanol acetate (MAM Ac). In normal mice, most cells destined for the cerebral cortex were produced from embryonic day 12 (E12) to E16 in the expected inside-to-outside sequence (deep layers first, superficial layers last). Injection of MAM Ac at E13 killed cells that would normally have contributed to the deep cortical layers. As a consequence, the cortex was thinned by approximately 25% at postnatal day 21 (P21). However, all laminae were present and had normal connections with subcortical structures, although all were proportionately thinner. BrdU injected on E16 labelled a normally sized complement of cells that spanned a larger proportion of the depth of the thinned cortex. Thus, the deep cortical layers comprised many cells that were born several days later than normal. At embryonic ages prior to E12, a transient set of cells is produced in the early telencephalon. After injection with MAM Ac at E10, the cortex appeared histologically and histochemically normal at P21. However, many cells that would normally have contributed to superficial cortex (born on E15) were significantly deeper than normal. These results suggest that, during the early stages of cortical development, the nervous system is sufficiently plastic to compensate to some extent for the destruction of specific precursor cells by altering the fates of neurons born later. They indicate that the embryonic date on which a cortical cell is born does not necessarily determine its eventual phenotype.

Cellular expression of somatostatin in MAM-induced microencephaly in the rat

Methylazoxymethanol acetate (MAM) is a mitotic inhibitor that has been used to selectively destroy neuroblasts at specific times during gestation. The administration of MAM results in a dose-dependent microencephaly. Following MAM treatment at 15 days of gestation, we have noted an increase in the level of SS immunoreactivity in the neocortex, as determined by radioimmunoassay. Northern blot analysis for preproSS mRNA revealed an increase in MAM-treated cortex. The cellular distribution of SS has been determined using in situ hybridization and immunocytochemistry. There was a 30% increase in the density of SS-immunoreactive neurons in the cortex of the MAM-treated animals. These data suggest that SS neurons in the cortex are spared following MAM treatment at GD 15.

When should we expect prenatally damaged human brains to have abnormal neuronal connections?

Abnormal function in prenatally damaged brains may occur as the result of loss of neuronal an/or glial elements, abnormal morphology of neurons, or abnormal and inappropriate connections. By abnormal connection I mean projections from a given nucleus to a nucleus or cortical region in the brain which does not normally receive such a projection. In the present paper it is proposed that focal brain damage, i.e. that damage involving circumscribed regions of the developing brain on one side of the body, is more likely to induce abnormal connections, than diffuse brain damage, affecting most of the brain bilaterally. This proposition is based on the available evidence from animal studies, albeit scanty, and is presented as a working hypothesis for future studies of human congenital brain damage.

Numbers of neurons and glia in mature rat somatosensory cortex: effects of prenatal exposure to ethanol

Stereological methods were used to examine the consequences of prenatal exposure to ethanol on the structure of area 3, primary somatosensory cortex, of the mature hooded rat. Pregnant rats were fed a liquid diet containing 6.7% (v/v) ethanol (Et), pair-fed an isocaloric liquid control diet (Ct), or fed a diet of chow and water (Ch). Cresyl violet-stained sections of 3-month-old pups were examined. The corrected mean size of the cell bodies of neurons in layers other than layer V was significantly smaller in the Et-treated rats; conversely, the mean somatic size of glia in each layer was significantly larger in the Et-treated rats. The laminar cell packing density for neurons and glia, however, was similar in rats from both treatment groups. The overall volume of area 3 and the volume of individual layers were about 33% smaller in Et-treated rats than in the pair-fed controls. Thus, the estimated total number of neurons in Et-treated rats (1.79 X 10(6] was significantly fewer than in Ch-treated rats (2.77 X 10(6] and in Ct-treated rats (2.66 X 10(6]. The total number of glia also was about 30% fewer in Et-treated rats than in the controls. Not all layers were affected equivalently. The space occupied by the neuropil was significantly greater in Et-treated rats, but only in layers II/III, IV, and VI; hence, the cell body/neuropil ratio in these layers was less in Et-treated rats than in the controls. Therefore, microcephaly caused by prenatal exposure to ethanol results not only from a miniaturization of the brain, but also from a permanent abnormal organization of cerebral cortex.

Patterns of transiently expressed acetylcholinesterase activity in cerebral cortex and dorsal thalamus of developing rats with cytotoxin-induced microencephaly

Previous studies have demonstrated that acetylcholinesterase (AChE) activity is expressed transiently by thalamocortical neurons of primary sensory systems in developing rat pups. In the present study, prenatal treatment with methylazoxymethanol acetate (MAM) on embryonic day 15, 16, or 17 resulted in rat pups with cerebral cortices markedly reduced in thickness and areal extent. Histochemical studies demonstrated that AChE staining occurs in fiber-like plexuses in primary visual, auditory, and somatosensory regions of developing cerebral cortex of MAM-treated animals, just as in normal developing rats, but that the transient patterns of AChE are found more superficially than normal and they occur in an abnormal patchy distribution. Neuronal somata in thalamic lateral geniculate, medial geniculate and ventral basal nuclei of MAM-treated animals show transient AChE staining indistinguishable from that seen in normal animals. These data indicate: (1) AChE is expressed transiently by thalamocortical neurons in MAM-treated animals, (2) intensity of the transiently expressed AChE is not affected by MAM-induced loss of cortical neurons, and (3) the abnormal AChE patterns in cortex likely reflect the abnormal distributions of thalamocortical terminal fields that are characteristic of MAM-treated animals.

Morphometrical and microdensitometrical studies on peptide- and tyrosine hydroxylase-like immunoreactivities in the forebrain of rats prenatally exposed to methylazoxymethanol acetate

Methylazoxymethanol acetate (MAM Ac) injected into pregnant rats at a dose of 25 mg/kg at gestational day 15 causes microcephaly due to an atrophy of various telencephalic areas, mainly neocortex, hippocampus and basal ganglia. Previous studies demonstrated alterations in various neurochemical markers of classical transmitter systems in these regions. The present paper deals with changes in peptide and tyrosine hydroxylase (TH)-containing neurons in MAM Ac-induced microcephaly using immunocytochemistry coupled with computer-assisted morphometry and microdensitometry. No change in the number of vasoactive intestinal polypeptide (VIP)-immunoreactive neurons in the neocortex and neuropeptide Y (NPY)-immunoreactive neurons in the nucleus caudatus-putamen was found whereas cholecystokinin (CCK)-and NPY-immunoreactive neurons in the neocortex and CCK- and VIP-immunoreactive neurons in the hippocampus were decreased. The reduction of the latter peptide containing neuronal populations led to a maintained density of cells in MAM Ac-exposed rats, due to the parallel reduction of the overall mass of these regions. TH immunoreactivity was found to be unchanged in the basal ganglia, and increased in the cerebral cortex in agreement with previous reports on noradrenaline cortical system after MAM Ac exposure. The present results show a heterogenous vulnerability of different peptide immunoreactive neuronal populations to MAM Ac exposure. The sparing of VIP- and NPY-immunoreactive neurons may be due to their late development in the neocortex and striatum, respectively. The hypothesis is introduced that cortical VIP interneurons can develop independent of marked alterations in the intrinsic circuitry of the cortical region.

Gross and Histological Criteria for Developmental Disorders in Brains of Schizophrenics

Autopsy findings and examination of histological sections from 12 schizophrenic patients and from age-matched, non-schizophrenic controls were enlarged and confirmed our previous study. Four subgroups were differentiated with gross abnormalities, in most cases with asymmetry of the temporal sulcal gyral pattern. Two types of macroscopical deviations are described. Different periods of developmental termination, concerning the sulcal gyral pattern of the temporal lobe, suggest a period in which the disturbance of the development originated which coincides with a probably genetically-induced disturbed migration in the entorhinal region toward the end of the fifth month. It is thought that the parahippocampal or entorhinal area develops differently, space- and time-wise. This may explain why migratory disturbances particularly occur in this precisely limited area. The findings in the ventral insular area are discussed, together with factors which also suggest a developmental disorder.

Regional- and Age-Specific Neurochemical Alterations in Rats Rendered Microencephalic by Differentially Timed Gestational Methylazoxymethanol Treatment

Rats with different degrees of microencephaly were obtained by injecting pregnant mothers with methylazoxymethanol acetate (MAM) at gestational days 13.5 or 16.5. Specific markers for cholinergic (choline acetyltransferase, ChAT), GABAergic (glutamate decarboxylase, GAD), and glutamatergic (D-3H aspartate high affinity uptake) neurons, were measured in several brain regions (cortex, hippocampus, anterior and posterior striatum, medial septum plus nucleus of the diagonal band, globus pallidus) in young and adult microencephalic rats. In adult rats born to mothers injected with MAM at gestational day 16.5 (G16.5) ChAT level was increased in the cortex, hippocampus and striatum but decreased in the septal complex; GAD was decreased in the globus pallidus and, to a little extent, in the hippocampus while D-3H aspartate uptake was decreased in the striatum. One month old rats belonging to the same group showed comparable differences with the exception of larger increase of ChAT in the cortex and striatum. In adult rats born from mothers injected with MAM at gestational day 13.5 (G13.5) differences in the cholinergic marker were in general less pronounced; GAD was not decreased in the globus pallidus and D-3H aspartate uptake was unchanged in the striatum but significantly decreased in the hippocampus. The results are correlated with morphological brain alterations caused by differentially timed MAM treatment and with available information on the generation time of various neuronal populations. They show that the balance between different neurotransmitter systems can be experimentally altered and suggest that MAM treatment may provide an experimental tool for studying the development of this balance.

Investigations of the origins of transient acetylcholinesterase activity in developing rat visual cortex

Transient acetylcholinesterase (AChE) activity is characteristic of cortical area 17 of the developing laboratory rat during the second and third postnatal weeks of life. This AChE activity is most intense in a band that corresponds to cortical layer IV and the deep part of layer III, but also is found in the outer half of cortical layer I and in layer VI. The morphology of the pattern of the histochemical reaction product indicates that the transient AChE is characteristic of an axonal terminal field. The present report describes results of 3 sets of experiments aimed at determining the source of transient AChE in cortical area 17. First, placement of lesions in portions of the basal forebrain or in the cingulate bundle results in a decrease in the general pattern of AChE throughout occipital cortex and especially in layer I, but the transient bands of AChE in layers III-IV of cortical area 17 are not eliminated. Second, kainic acid or cobalt chloride injections in cortical area 17 result in the loss of many AChE-positive neuronal somata but do not eliminate the transient pattern of AChE in thalamo-recipient layers of cortical area 17. Similarly, treatment of fetuses with mitotic inhibitors that eliminate many of the neurons destined for granular and supragranular layers does not eliminate transient patterns of AChE. Third, lesions that include the lateral geniculate nucleus of the thalamus or geniculocortical projections result in a marked loss of the pattern of AChE in thalamo-recipient layers of cortical area 17, without significant loss in other layers of area 17 or in other regions of occipital cortex. These data support the hypothesis that the transient AChE found in thalamo-recipient layers of cortical area 17 is contained within geniculocortical axon terminals.

Development and decision-making in the mammalian cerebral cortex

One of the fundamental tasks of neurobiology is to understand how the precision and specificity of the adult nervous system is achieved during development. This paper reviews the progress that has been made toward this end in studies of the developing mammalian cerebral cortex. Particular attention is focused on the problem of how cortical neurons make decisions during development: the correlation between a neuron's 'birthday' and its final laminar destination and projection patterns has raised the possibility that young neurons may be committed to their adult fates very early on in development, perhaps prior to migration. Indeed, several lines of evidence reviewed here suggest that at least some of the decisions made by cortical neurons are intrinsic properties of the cell itself. These studies include experiments on the reeler mouse mutant, and more recent attempts to manipulate developmental fates by pharmacological interventions and transplantation techniques. It is concluded that early commitment events in the cerebral cortex may specify a neuron's laminar position and restrict the range of potential axonal projections that the cell may form, but that local positional cues direct neurons to select (or maintain) only certain of the possible projections.

Vascularity and cytochrome oxidase distribution in the occipital cortex in MAM Ac-induced micrencephaly

The laminar distributions of cytochrome oxidase activity and vascularization of the occipital cortex of animals made micrencephalic by exposure to MAM Ac on E14 or E16 were examined and compared to control animals. Despite severe disruptions of the normal cytoarchitecture in MAM Ac exposed animals, the laminar distributions of vascular profiles, branch points and cytochrome oxidase activity in micrencephalic animals were similar to those in control animals.

Selective ablation of neurons by methylazoxymethanol during pre- and postnatal brain development

Neonatal or pregnant albino rats were injected with either single or double doses of methylazoxymethanol (20 mg/kg) to test the temporal specificity of its effect on clearly definable regions of the brain. A single dose, to dams from gestational day 11 to 21 (G11-G21) and to neonatal rats from birth to postnatal day 5 (P0-P5), produced differential weight reductions among various brain regions. Two prominent peaks of reduction were found: one occurring between G13 and G15 for the cerebrum and hippocampus and one occurring between P0 and P1 for the cerebellum and olfactory bulbs. Dual injections of the drug on G14 and G15 produced 60% weight reduction in the cerebrum, and slightly earlier injections on G13 and G14 reduced the weight of the cerebellum by about 23%. This weight reduction was accompanied by narrowing of the cerebellar width, which we believe was due to fewer Purkinje cells. Dual injections of methylazoxymethanol at P0 and P1 reduced the weight of the olfactory bulb by 65%, the cerebellum by 62%, and the hippocampus by 18%. These results show that its short action is within the window of cell division for various neurons and becomes additive on two successive days. This precise toxic effect on brain development can be used to disproportionally reduce the number of neuroblasts in specific regions of the brain. A differential ablation allows analysis of plasticity on pyramidal and nonpyramidal cells of the neocortex and hippocampus, Purkinje and granule cells of the cerebellum, and the granule and mitral cells of the olfactory bulbs.

A quantitative study of developing axons and glia following altered gliogenesis in rat optic nerve

Axonal and glial cell development within rat optic nerve in which gliogenesis was altered by systemic injection of 5-azacytidine (5-AZ) was examined by quantitative electron microscopy. In neonatal (0-2 days) rat optic nerves, all fibers are premyelinated, and they exhibit a fairly uniform diameter (approximately 0.22 micron). These fibers occupy approximately 55% of the optic nerve volume. At this early age, glia within the optic nerve consist only of cells of astrocytic lineage and progenitor cells. These glia occupy approximately 28% of the optic nerve volume, and there are approximately 80 glial cells/optic nerve cross section. In 14-day-old normal optic nerves, myelinated and ensheathed fibers comprise approximately 17% and 9%, respectively, of the total number of axons. Mean axonal diameter of myelinated fibers is approximately 0.75 micron, while mean diameter for ensheathed axons is approximately 0.50 micron. By volume, these fibers occupy approximately 25% of the nerve, which is similar to the volume occupied by premyelinated axons in these nerves. At 14 days of age, there are approximately 300 glial cells/optic nerve transverse section, and these glia occupy approximately 37% of the volume in normal optic nerve. Oligodendroglia represent approximately 40% of total glial cells present, while astroglia and progenitor cell each comprise approximately 30% of the cells. In optic nerves from 14-day-old rats treated with 5-AZ, few myelinated fibers are present and the number of oligodendroglia is markedly reduced. Axonal diameter of premyelinated fibers is similar to that of age-matched controls. Myelinated and ensheathed fibers comprise approximately 2% of the total fibers present in 5-AZ-treated optic nerves, with the remaining fibers being premyelinated. The few myelinated and ensheathed fibers present in 5-AZ-treated optic nerves display similar axonal diameters to corresponding fibers from age-matched control tissue. Glial cells occupy approximately 40% of the nerve volume, and there are approximately 200 glia/nerve cross section in 5-AZ-treated rats. Astroglia comprise approximately 63% of the total glial cells, while approximately 12% of the cells are oligodendroglia. These results demonstrate that 5-AZ is a potent inhibitor of oligodendrogliogenesis, with a concomitant marked reduction in the number of myelinated fibers.

Molecular structure of the axolemma of developing axons following altered gliogenesis in rat optic nerve

Axonal and axolemmal development of fibers from rat optic nerves in which gliogenesis was severely delayed by systemic injection of 5-azacytidine (5-AZ) was examined by freeze-fracture electron microscopy. In neonatal (0-2 days) rat optic nerves, all fibers lack myelin, whereas in the adult, virtually all axons are myelinated. The axolemma of neonatal premyelinated fibers is relatively undifferentiated. The P-fracture face (P-face) displays a moderate (approximately 550/micron 2) density of intramembranous particles (IMPs), whereas the E-fracture face (E-face) has few IMPs (approximately 125/micron 2) present. By 14 days of age, approximately 25% of the axons within control optic nerves are ensheathed or myelinated, with the remaining axons premyelinated. The ensheathed and myelinated fibers display increased axonal diameter compared to premyelinated axons, and these larger caliber fibers exhibit marked axonal membrane differentiation. Notably, the P-face IMP density of ensheathed and myelinated fibers is substantially increased compared to premyelinated axolemma, and, at nodes of Ranvier, the density of E-face particles is moderately high (approximately 1300/micron 2), in comparison to internodal or premyelinated E-face axolemma. In optic nerves from 14-day-old 5-AZ-treated rats, few oligodendrocytes are present, and the percentage of myelinated fibers is markedly reduced. Despite delayed gliogenesis, some unensheathed axons within 5-AZ-treated optic nerves display an increased axonal diameter compared to premyelinated fibers. Most of these large caliber fibers also exhibit a substantial increase in P-face IMP density. Small (less than 0.4 micron) diameter unensheathed axons within treated optic nerves maintain a P-face IMP density similar to that of control premyelinated fibers. Regions of increased E-face particle density were not observed. The results demonstrate that some aspects of axolemma differentiation continue despite delayed gliogenesis and the absence of glial ensheathment, and suggest that axolemmal ultrastructure is, at least in part, independent of glial cell association.

Transplantation of fetal postmitotic neurons to rat cortex: survival, early pathway choices and long-term projections of outgrowing axons

A system for studying growth and development of transplanted subpopulations of postmitotic cerebral cortical neurons is described. The cytotoxic drug methylazoxymethanol (MAM) was given to pregnant rats on the fourteenth day of gestation to destroy precursor cells of layers II-IV of the fetal cerebral cortex. Layer V and VI precursor cells which had completed their final division before MAM treatment and were unaffected by it, were labeled by a prior injection of [3H]thymidine. This strategy provides a donor cerebral cortex containing mainly neurons destined to form layers V and VI of the adult cerebral cortex; these cells are postmitotic. Pieces of donor cerebral cortex were transplanted to the cerebral hemispheres of normal newborn hosts at one day, two days, or 6 days after MAM treatment; survival was assessed 1-12 weeks after transplantation by autoradiography of histological sections. Radiolabeled graft cells survived in 89% of recipients and many of these grew axons into the host, as indicated by retrograde labeling with horseradish peroxidase. Significant numbers of graft cells could also be stained immunocytochemically for glutamic acid decarboxylase or for the peptides, somatostatin, vasoactive intestinal polypeptide or cholecystokinin. A second group of experiments examined the routes of early axon outgrowth from normal and postmitotic fetal grafts. When the donor cortex had been incubated in a mixture of [3H]proline and [3H]leucine for 20 min prior to transplantation, the earliest axons growing out of the graft into normal newborn hosts could be assessed by autoradiography of axoplasmic transport after survivals in the host of 7 days. Normal and postmitotic grafts taken at E15 or E20 were capable of outgrowth, though the axons of E20 postmitotic cells did not grow far. The location of the transplant was the major determinant of where graft cells' axons grew and growth was mainly into existing axonal pathways of the host. In a third group of experiments, long term axonal projections from normal and postmitotic fetal transplants to 4 regions of the host brain--thalamus, contralateral cortex, striatum, and hippocampus--were examined with retrograde tracers 2-4 months after transplantation. Projections from grafts to the 4 host sites were highly dependent on the presence of nearby host axons connecting with those sites. Neurons in all types of graft projected to one or other of the 4 sites, but generally in small numbers. Higher proportions of cells in grafts from E15 MAM-treated donors projected to the host thalamus.(ABSTRACT TRUNCATED AT 400 WORDS)

Effects of altered gliogenesis on activity-dependent K+ accumulation in the developing rat optic nerve

Gliogenesis in the rat optic nerve is disrupted by neonatal treatment with the mitotic inhibitor 5-azacytidine (5-AZ). The rate of myelination and number of glial cells are markedly reduced in treated animals. We analyzed the physiological consequences of these chemically induced morphological alterations in terms of activity-dependent K+ accumulation in brain extracellular space and characteristics of the compound action potential (CAP). Nerves from 5-AZ-treated animals older than 5 days of age showed significantly higher activity-dependent 'ceiling levels' of extracellular K+ concentration ( [K+]o) than controls. This result is consistent with the hypothesis that glial cells are involved in K+ homeostasis at a cellular level and play a role in helping to set the ceiling level of activity-dependent K+ accumulation. The CAPs of 5-AZ-treated nerves older than 5 days of age were larger and generally of simpler configuration than those observed in control animals, perhaps due, among other factors, to the retained uniformity of axonal conduction velocity caused by inhibition of myelination.

Rat optic nerve: disruption of gliogenesis with 5-azacytidine during early postnatal development

The normal sequence of gliogenesis in the rat optic nerve was disrupted by neonatal treatment with the mitotic inhibitor 5-azacytidine (5-AZ). This protocol caused a marked reduction in the number of glial cells, especially oligodendrocytes, seen in 11- and 14-day-old animals. Myelin formation was also greatly reduced in animals of this age compared to controls, but optic nerve axons appeared to be well preserved. Electrophysiological studies demonstrated similar excitability properties and activity-evoked [K+] changes in normal and 5-AZ-treated nerves prior to 5 days of age. However, in older nerves there were striking changes in the compound action potential and activity-dependent K+ accumulation in 5-AZ-treated nerves compared to controls. This simple model of disrupted central nervous system gliogenesis should prove useful in analyzing a variety of questions regarding neuroglial interactions including the role of glial cells in ionic homeostasis of brain extracellular space.

The angiogenesis of micrencephalic rat brains caused by methylazoxymethanol acetate. I. Superficial venous system. A quantitative analysis

The angiogenesis of the rat cerebrum was studied under pathologic conditions caused by the administration of the neurotoxin methylazoxymethanol acetate (MAMAc) in the time (E14) of neuroblast migration. The sinovenous junction of the main superficial cerebral veins and the morphological changes of the veins were examined by a quantitative analytic method. The hypoplastic areas of the brains showed extremely malformed venous systems with pathologic changes of the sinovenous junctions depending on the degree of disturbance of the neuroblast migration. These findings suggest the primary role of the neuronal maturation in the angioarchitectonic development and the direct dependency of the vascular differentiation on the neuroblast migration of the drained territory.

The angiogenesis of micrencephalic rat brains caused by methylazoxymethanol acetate. II. Superficial and basal arterial system

The angioarchitecture of the superficial and basal arterial system of the hypoplastic rat brain caused by the administration of the neurotoxin, methylazoxymethanol acetate (MAMAc), at the time of neuroblast migration was studied. Increased variation of the arterial branching of the basal main stem of arteries and local vessel changes were observed. The findings suggest a close relationship between vascular and neuronal development, showing a generalized disturbance and local adaptation of vasculature to the altered neuronal architecture of the corresponding hypoplastic areas.

A Golgi analysis of unlayered polymicrogyria

The cytoarchitectonics of the cerebral unlayered polymicrogyria located at the borders of a bilateral porencephalic defect is characterized by minute convolutions not exteriorized by sulci, in which blood vessels and increased numbers of fibrillary astrocytes are present in the fused molecular layers. The cellular organization, based on the analysis of Golgi sections, differs among gyral, intermediate, and sulcal regions and represents variable degree of cellular damage and structural organization of the cerebral mantle injured approximately in gestational month 5. Polymicrogyria may be produced by incomplete ischemia of radial territories vascularized by cortical blood vessels penetrating at right angles from the surface which is the result of the imbalance between the impaired cerebral blood flow of occluded large prerolandic arteries, responsible for the porencephalic defect, and the arterial meningeal anastomoses. Abnormal folding in polymicrogyria may be generated by lateral differences in the cortical thickness of adjoining areas, and by the imbalance in growth rates of laterally contiguous cortical regions.

Neuroanatomical study of somatomotor cortex in microcephalic mice induced by cytosine arabinoside

Somatomotor cortex of mice with microcephaly induced by DNA polymerase inhibitor cytosine arabinoside (Ara-C), has been studied with a modified Golgi-Cox staining and a HRP retrograde tracing method. Microcephalic mice were prepared by prenatal injections of cytosine arabinoside on days 13.5 and 14.5 of pregnancy. Cytoarchitectonically, the cerebral cortices of adult microcephalic mice are characterized by atypical pyramidal cells with abnormal dendrites and irregular patterns of cellular lamination. Semiquantitative analyses of the abnormality of dendrites in Golgi-Cox preparation indicate that both the degree and direction of ramification are severely affected in Ara-C treated mice. In adult control cerebrum, original neurons of corticospinal tract labeled after HRP injection into the lumbar cord were situated in layer V. In the microcephalic brains, however, HRP labeled neurons, some of which had abnormal polarity, were scattered throughout all layers. This HRP study for corticospinal tract neurons also confirms the irregular pattern of the cortex in which only three layers are recognized.

Effects of cytotoxic deletions of somatic sensory cortex in fetal rats

Pregnant rats were injected on the 14th day of gestation with the cytotoxic drug methylazoxymethanol acetate. This compound causes the death of neural precursor cells that were synthesizing DNA at the time of injection. After birth, the progeny of treated mothers grew to maturity with a neocortex that was greatly reduced in area by the death of all cells, particularly at the frontal and occipital poles but at medial and lateral margins of neocortex as well. In the remaining cortex layers II through IV failed to develop. The experiment deprived growing thalamocortical axons, which innervate the somatic sensory cortex late in development, of part of their normal target area and of a substantial number of their definitive target cells. It also deprived them of any cues they might have received from these target cells migrating through them as the axons accumulate beneath the cortical plate. Anatomical experiments indicated that, despite these defects, thalamocortical axons could still colonize the sensorimotor areas and form synapses in their typically bilaminar pattern, though the outer, denser lamina of terminations occurred abnormally at the level of the apices of layer V pyramidal cell bodies. Receptive field mapping of single and multiunit responses in the somatic sensory region showed brisk responses and receptive fields of normal size. It also indicated the formation of a body map that was topographically intact except for deletions at its periphery; that is, a total map was not compressed into a smaller area. This suggests that somatic sensory thalamocortical fibers recognize only remaining cortical target cells in appropriate fields. Moreover, successful ones among them seem to recognize neighborhood relations and conserve synaptic space at the expense of those that would have innervated the deleted peripheral parts of the area. Pyramidal neurons in the remaining cortical layers and in ectopic islands of cells that had incompletely migrated from the neuroepithelium, probably on account of destruction of radial glial cell precursors, were shown by retrograde labeling to send their axons only to appropriate subcortical targets. Pyramidal neurons, though recognized as such, also adopted a variety of abnormal orientations, such as inversion, apparently in an attempt to grow apical dendrites toward major zones of synaptic terminations.

Fetally-induced noradrenergic hyperinnervation of cerebral cortex results in persistent down-regulation of beta-receptors

Methylazoxymethanol acetate (MAM)-induced cortical hypoplasia resulted in a 2-fold increase in the concentration of norepinephrine and its metabolite, 3-methoxyl-4-hydroxyphenylglycol sulfate, and a 26% decrease in the Bmax for the beta-receptor ligand, [3H]dihydroalprenolol (DHA) in the rat cortex. Chronic treatment with desmethylimipramine did not further decrease DHA-labeled sites although prior lesion of the noradrenergic terminals with 6-hydroxydopamine markedly increased the number in the MAM-lesioned cortex.

The pathogenesis of abnormal cytoarchitecture in the cerebral cortex and hippocampus of the mouse treated transplacentally with cytosine arabinoside

Pregnant mice were treated with cytosine arabinoside on days 13.5 and 14.5 of pregnancy. Brains of the offspring were studied histologically. The matrix layer of the embryonic brains was extensively destroyed 12 h after the injection of cytosine arabinoside, but regenerated partially on day 17 of gestation. In the cerebral cortex of 1-, 3-, and 5-day-old treated mice, abnormal clusters of young neurons were found on the surface of the developing cerebral cortex. Some clusters still had a supply of immature neurons from the remnants of the regenerated matrix layer. After 20 days, the clusters became gradually indistinct, although some vestigial groups of neurons were observed even after 120 days. In the hippocampus of young mice, the pyramidal cells decreased in number and were disarranged. Heterotopic pyramidal cell masses were found in the stratum radiatum and in the molecular layer of the dentate gyrus. Apical dendrites of pyramidal cells exhibited abnormal arborization. It was demonstrated by 3H-thymidine autoradiography that young neurons in the abnormal clusters in the cerebral cortex were those produced in the matrix layer regenerated after the destructive change by cytosine arabinoside.