Laurence J. Garey

Synaptogenesis in human visual cortex — evidence for synapse elimination during normal development ☆

Age-related changes in synaptic density in human visual cortical area 17 (striate cortex) were determined, based on counts in material prepared for electron microscopy with the phosphotungstic acid method. The results were correlated with measurements of the volume of striate cortex from celloidin sections. Two periods were defined, one of rapid synapse production which ends at about postnatal age 8 months, and a subsequent longer period of synapse elimination which extends past age 3 years. Exuberant synaptic connections during early childhood may impart to the immature cerebral cortex plasticity which is lost in the adult.

Structural development of the lateral geniculate nucleus and visual cortex in monkey and man

This study concerns the development of the primary visual pathway of the primate. The lateral geniculate nucleus (LGN) is the principal thalamic relay to the visual cortex (area 17), and its neurons have similar morphological characteristics in both monkey and man, as identified by Golgi impregnation. The commonest neuron is the multipolar with a radiate or tufted dendritic tree; next is the bipolar neuron with two or three diametrically opposed dendritic trunks. Less frequent are neurons with beaded dendrites and others with fine, axon-like dendritic processes, possibly interneurons. The dendritic tree of all neurons remains generally within a lamina, but some dendrites cross interlaminar zones. LGN neurons are identifiable before birth and differ from their adult form by the presence of immature features, especially numerous dendritic and somatic spines, most frequent at birth in monkeys and at about 4 months postnatally in man. They disappear almost completely by 3 months in monkeys and 9 months in man. The human LGN has reached its adult volume by this age. Two stages in the development of the human area 17 can be defined. The first is marked by a rapid growth to its adult volume by about 4 months, and by intense synaptogenesis beginning in the foetus and reaching a maximum around 8 months. The second stage is one of stabilization in the volume of area 17 and loss of synapses to reach adult synaptic density around 11 years, at about 60% of the maximum values.

Monocular and binocular deprivation in the monkey: Morphological effects and reversibility

Measurements were made of the cross-sectional area of neurons in the lateral geniculate nucleus (LGN) of normal monkeys, and of monkeys subjected to monocular and binocular eyelid suture. Early monocular deprivation caused a failure of normal cell growth in the LGN such that the neurons in the laminae innervated by the deprived eye were, on average, 15% smaller than those innervated by the normal eye. In the first 4 days of life monocular deprivation caused significant retardation of growth. The effect was marked in the first 6 weeks, but was not found in a monkey monocularly deprived from 11-16 months of age nor in an adult deprived for more than 6 months. Binocular deprivation from birth appeared to arrest neuronal growth at the neonatal size. The effect of monocular deprivation could be cancelled by reverse-suture (opening the closed eye and closing the other) during the first 2 months of life. Changes in the size of LGN neurons following monocular deprivation and reverse-suture correlated closely with changes in the relative width of ocular dominance columns in layer IVc of area 17 of the visual cortex, measured physiologically in the same animals.

Comparison of neuronal and glial numerical density in primary and secondary visual cortex of man

SummaryThe numerical density of neurons and glial cells was estimated in visual area 18 of the adult human cerebral cortex and compared with that of area 17. Blocks of areas 17 and 18 came from the same brains and this allowed the comparison of 1) neuronal and glial numerical densities through the whole cortical depth with calculation of the neuron/glia ratio, 2) neuronal and glial numbers under one square millimeter of cortical surface, and 3) neuronal numerical densities in three groups of identified layers. The mean neuronal density is approximately 40000 neurons/mm3 in area 17 and 31500/m3 in area 18. The mean glial density is around 27000/mm3 in area 17 and 32000/mm3 in area 18. This gives a neuron/glia ratio of approximately 1.5 in area 17 and of 1.0 in area 18, but the total cellular density is similar in both areas. There are about 90000 neurons and 64000 glial cells under one square millimeter of cortical surface in area 17, and some 73000 neurons and 74000 glial cells in area 18. The higher neuronal density in area 17 is found through the whole depth of cortex and does not seem to be more pronounced in layer IVc of area 17 compared to layer IV in area 18 than in the groups of layers II–III and V–VI.

Morphology of the neurons in the human lateral geniculate nucleus and their normal development

SummaryNeurons in the adult human lateral geniculate nucleus have been classified using Golgi preparations. The neuronal classes correspond to those previously described in monkey (Saini and Garey 1981). The commonest are multipolar neurons with either “radiate” or “tufted” dendritic trees. Also seen frequently are bipolar neurons with two or three diametrically opposed dendrites. Rarer classes include neurons with beaded dendrites and those with “axon-like” dendritic processes, perhaps inter-neurons. Neurons are also found in the circumgenicule capsule. Most neurons have dendrites restricted to the laminae, but some dendrites cross the borders of both magno- and parvocellular laminae. Somata are also seen in interlaminar zones with dendrites reaching the adjacent laminae. No significant difference, apart from size, is found between neurons in magno- and parvocellular laminae.Most neuronal types are found at birth. They are, however, strikingly different from their adult forms in having growth cones and filopodia and an abundance of dendritic and somatic spines and “hair-like” processes. Morphological maturity is reached by about nine months postnatally. Similar maturational changes occur in monkey in the first two months of life (Garey and Saini 1981) and in both monkey and man this period of maturation of the lateral geniculate nucleus corresponds to increasing visual acuity and a time when each species is most likely to be affected by visual deprivation.

Heme oxygenase-1 activity after excitotoxic injury: Immunohistochemical localization of bilirubin in neurons and astrocytes and deleterious effects of heme oxygenase inhibition on neuronal survival after kainate treatment

An increased expression of the inducible form of heme oxygenase (HO), HO‐1, is found in the hippocampus after kainate injection, but thus far it is unclear whether the HO‐1 is enzymatically active. The present study was carried out, using monoclonal antibodies to bilirubin and HO‐1 and histochemical staining for iron, to compare the products of HO enzymatic activity, bilirubin and iron, with HO‐1 expression in the kainate‐lesioned hippocampus. There was a close correlation between bilirubin and HO‐1 expression, and both bilirubin and HO‐1 were observed in damaged neurons at early times, and astrocytes at later times (weeks), after kainate injection. These results indicate that the increased HO‐1 in the hippocampus is enzymatically active. Too determine whether HO‐1 activity after kainate could have a protective or, perhaps, destructive effect, kainate‐injected rats were injected intraperitoneally with a blood–brain barrier‐permeable inhibitor of HO, tin protoporphyrin (SnPP), and the effects of such treatment were compared with effects in rats that received kainate and saline injection. It was found that SnPP treatment did not improve neuronal survival. Instead, increased mortality was observed in rats treated with SnPP. Four SnPP‐injected rats vs. one saline‐injected rats died after kainate treatment. The surviving SnPP‐treated rats showed significantly less hippocampal field that containing Nissl or MAP2 staining (an indicator of surviving neurons) compared with the saline‐injected rats. These results indicate that HO‐1 induction had a net protective effect on neurons in the kainate model of excitotoxic injury.

Distribution of the cells of origin of the corpus callosum and anterior commissure in the marmoset monkey

SummaryThe neurons of origin of the great cerebral commissures of the marmoset monkey were identified by horseradish peroxidase histochemistry and their distribution was studied. Six adult marmosets were used. Three were normal: the others were subjected to section of the corpus callosum (CC), sparing the anterior commissure (AC). All six were injected with horseradish peroxidase throughout one cerebral hemisphere. The three normals provide information on the origins of both the CC and AC, whereas the three callosotomized monkeys allow study of the origins of the AC alone. All CC and AC neurons in the marmoset are pyramidal cells. Except for layer I, all cortical layers possess commissural cells; their laminar organization varies according to cortical area. There exists a progression in predominance from supra- to infragranular commissural neurons proceeding from temporal through occipital to parietal and finally to frontal cortex. Major acallosal zones are found in the primary visual cortex and the fore- and hindlimb representations of the somatosensory cortex. Correlations between commissural neuron distribution and cytoarchitectonic areas are not always obvious. Commissural neurons were not organized in columnar fashion.

The use of ibotenic acid lesions for light and electron microscopic study of anterograde degeneration in the visual pathway of the cat.

Ibotenic acid was injected in the lateral geniculate nucleus of cats to investigate the suitability of the technique in studies involving anterograde degeneration in the thalamo-cortical pathway when it is important not to damage fibres of passage at the lesion site. Small injections cause localized degeneration of neuronal somata in the lateral geniculate nucleus, but axons passing through the injected zone remain intact. Degeneration can be localized in visual cortex by light microscopic silver techniques and by electron microscopy. The appearance, density and distribution of degenerating cortical axon terminals is similar to what is found after thalamic electrolytic lesions.

Primary and secondary subcortical projections of the monkey visual system. An autoradiographic study

Recent studies on the primary visual pathways carried out by means of autoradiographic tracer techniques15,1~,20, 21, have considerably improved our knowledge on the terminations of retinal fibers in diencephalic and mesencephalic areas. In addition, transsynaptic transport to secondary retinal target sites has also been observed in a number of speciesS,9, 22. Since pertinent data in the monkey are still scarce, we decided to study transsynaptic transport along the visual projection pathways in this animal and to pay special attention to pretectal and mesencephalic areas. The results are briefly summarized in this report. An equal mixture of [3H]fucose and [aH]proline, 2-6 mCi in 50-100/~I was injected in the vitreous of one eye in 5 monkeys (2 adult Macaca fascieularis and 3 young Erythroeebus patas). One macaque (A79-188) received a second injection 7 days after the first injection, 12 mCi in total. The animals were allowed to survive from 11 to 21 days after the first injection and then were perfused with buffered 4 ~ paraformaldehyde. The brains were cut in either frozen or paraffin sections, and processed for autoradiography, exposed for several weeks at 4 °C, then developed and counterstained with cresyl violet. The density of silver grains was determined at a magnification of 600 × with an eyepiece reticule consisting of 100 squares (each square ---625 sq./~m) which was placed at random over the nucleus to be analyzed; silver grains in 60-100 squares were counted and the mean grain density per unit area (625 sq.#m) in each nuclear region was determined. The mean grain density over subependymal layers in the lateral wall of the aqueduct was regarded as background level in each case. Statistical differences were examined by the Students t-test.

Distribution of the neurons of origin of the great cerebral commissures in the cat

SummaryLarge injections of horseradish peroxidase throughout major portions of the right cerebral hemispheres of four cats revealed extensive distributions of the neurons of origin of the corpus callosum, the anterior commissure and the hippocampal commissures in the uninjected left hemispheres. The distributions of labelled neurons were mapped by semiautomatic computer microscope. The radial and tangential neuron distributions presented here are of a higher density and greater extent than those in previously published studies based on injections of transportable label to more circumcribed areas of the cerebral cortex of the cat. Generally, commissural neurons in the cat were distributed in a bilaminar fashion with supragranular cells more numerous than infragranular cells.