Faculty Bibliography

The pattern of neuronal loss in the rat hippocampus following 10-min-long cardiac arrest-induced global ischemia was analyzed using the unbiased, dissector morphometric technique and hierarchical sampling. On the third day after ischemia, the pyramidal layer of sector CA1 demonstrated significant (27%) neuronal loss (P<0.05). At this time, no neuronal loss was observed in other cornu Ammonis sectors or the granular layer of the dentate gyrus. On the 14th postischemic day, further neuronal loss in the sector CA1 pyramidal layer was noticed. At this time, this sector contained 31% fewer pyramidal neurons than on the third day (P<0.05) and 58% fewer than in the control group (P<0.01). On the 14th day, neuronal loss in other hippocampal subdivisions also was observed. The pyramidal layer of sector CA3 contained 36% fewer neurons than in the control group (P<0.05), whereas the granular layer of the dentate gyrus contained 40% fewer (P<0.05). The total number of pyramidal neurons in sector CA2 remained unchanged. After the 14th day, no significant alterations in the total number of neurons were observed in any subdivision of the hippocampus until the 12th month of observation. Unbiased morphometric analysis emphasizes the exceptional susceptibility of sector CA1 pyramidal neurons to hypoxia/ischemia but also demonstrates significant neuronal loss in sector CA3 and the dentate granular layer, previously considered 'relatively resistant'. The different timing of neuronal dropout in sectors CA1 and CA3 and the dentate gyrus may implicate the existence of region-related properties, which determine earlier or later reactions to ischemia. However, the hippocampus has a unique, unidirectional system of intrinsic connections, whereby the majority of dentate granular neuron projections target the sector CA3 pyramidal neurons, which in turn project mostly to sector CA1. As a result, the early neuronal dropout in sector CA1 may result in retrograde transynaptic degeneration of neurons in other areas. The lack of neuronal loss in sector CA2 can be explained by the resistance of this sector to ischemia/hypoxia and the fact that this sector is not included in the major chain of intrahippocampal connections and hence is not affected by retrograde changes

In Alzheimer's disease (AD), neurofibrillary degeneration of neurons starts in the transentorhinal cortex and spreads in a time-dependent manner to the entorhinal cortex, which provides a major input to the hippocampus--a key structure of the memory system. People with Down's syndrome (DS) develop neurofibrillary changes more than 30 years earlier than those with sporadic AD. To characterize AD-related pathology in the entorhinal cortex in DS, we examined seven subjects with DS of 60-74 years of age who died in the end stage of AD, and four age-matched control subjects. The volume of the entorhinal cortex in brains of subjects with DS was 42% less than that in control cases; however, the total number of neurons free of neurofibrillary changes was reduced in DS by 90%: from 9,619,000 +/- 914,000 (mean +/- standard deviation) to 932,000 +/- 504,000. The presence of 2,488,000 +/- 544,000 neurofibrillary tangles in the entorhinal cortex of people with DS, the prevalence of end-stage tangles, and the significant negative correlation between the total number of intact neurons and the percentage of neurons with neurofibrillary changes indicate that neurofibrillary degeneration is a major cause of neuronal loss in the entorhinal cortex of people with DS. The relatively low amyloid load (7 +/- 1%) and lack of correlation between the amyloid load and the volumetric or neuronal loss suggest that the contribution of beta-amyloid to neuronal loss in the entorhinal cortex is unsubstantial

A characteristic feature of the parvopyramidal layer of the presubiculum of 6 individuals with Alzheimer disease (AD) was the presence of large, evenly distributed amyloid-beta (A beta) deposits, which in the end stage of the disease occupy 80.9 +/- 12.2% of the parvopyramidal layer. The strong reaction of A beta deposits with antibodies 4G8 (17-24 amino acids, aa), 6E10 (1-17 aa), and R165 (32-42 aa), and their weak reaction with antibody R162 (32-40 aa) indicate that potentially highly fibrillogenic A beta1-42 is a major constituent of presubicular amyloid. However, A beta deposits in the presubiculum are thioflavin-S- and Congo red-negative--and thus, nonfibrillar--even after 11 to 19 years of AD. The unique properties of presubicular amyloid appear to be related to their origin; amyloid-associated proteins such as apolipoproteins E, and AI, alpha1-antichymotrypsin, and heparan sulfate proteoglycan, which are promoters of fibrillization or stabilizers of A beta in neuritic plaques, are absent; activated astrocytes, which are the source of these proteins, are also absent. The unchanged number and distribution and the resting appearance of microglial cells revealed with RCA-I histochemistry suggest that they do not respond to diffuse A beta deposits. The source of nonfibrillar presubicular A beta is probably local neurons or neuronal projections to the parvocellular layer of the presubiculum. Neuronal, lake-like A beta deposition appears to be characteristic of AD pathology. The presubiculum is most likely the model brain structure for the study of amyloid of exclusively neuronal origin. The parvopyramidal layer of the presubiculum reveals only a small population of the neurons (2.5 +/- 2%) affected by neurofibrillary pathology

The retrograde axonal transport method was used to compare the topography and organization of the visual zone of the claustrum in rat, guinea pig, rabbit and cat. First, massive Fluoro-Gold injections were placed into the primary visual cortex and the secondary areas. Experiments showed differences in the location of the visual zone among the animals under study. In rat, the visual zone occupied the posteroventral part of the claustrum and spread to its anterior pole. In guinea pig, neurons projecting to the visual cortex were located dorsally in the posterior half of the claustrum. In rabbit, similarly to the rat, they were localized in the posteroventral part; however, they did not reach the anterior pole. In cat, neurons that project to the visual cortex were concentrated dorsally in the posterior fourth of the claustrum. In double-injection experiments, Fast Blue and Diamidino Yellow were placed into the primary and secondary visual areas in various combinations. The experiments showed that in the rat and the rabbit claustral neurons project to primary visual cortex (area 17) as well as to both secondary visual areas (areas 18a and b). Populations of neurons sending axons to the primary and secondary areas showed full overlap. The presence of double-labeled neurons indicates that some claustral neurons project both to the primary and secondary fields. In cat, neurons that project to the primary visual cortex appear to be clearly separated from those connected with the secondary visual area, as no double-labeled neurons were found. In all studied species, the double injections placed into the visual and primary somatosensory cortex did not result in any double-labeling neurons. Our results indicate that the location of the visual zone in the posterior part of the claustrum is a phylogenetically stable feature, whereas its dorsoventral shift as well as the extent toward the anterior pole is related to the particular species. The overlap of neurons projecting to the primary and secondary visual areas in the rat and rabbit as well as the separation of both projections in cat appear to reflect the higher degree of complexity of the visual system in the latter

Previously, by using in vivo microdialysis, we demonstrated a huge release of 45Ca2+ from prelabeled tissues to dialysate that was evoked by application of N-methyl-D-aspartate (NMDA) to the rat dentate gyrus (DG) and sector 4 of the cornu ammonis. To establish the mechanism of this phenomenon, in the present study, we characterized its NMDA receptor dependence, investigated the mechanism of 45Ca2+ removal from the cells, and evaluated the possible involvement of calcium-binding protein calbindin D28k and of ryanodine receptors. Microdialysis experiments demonstrated a dose-response relation between NMDA and 45Ca2+ release and sensitivity of this phenomenon to inhibition by 10 microM MK-801 and 5 mM 5-(N,N-dimethyl)-amiloride, thus indicating the NMDA receptor dependence and a role of Na+/Ca2+ exchanger in mediating 45Ca2+ release from cells. Immunocytochemical experiments confirmed that DG granule cells in the investigated inbred rat strain are strongly calbindin D28k-immunopositive, indicating probable involvement of this protein. However, microdialysis studies demonstrated that NMDA-evoked 45Ca2+ release was suppressed by 100 microM dantrolene and 250 microM ryanodine, whereas 50 microM ryanodine stimulated this effect. This points to a key role in the investigated phenomenon of calcium-induced calcium release (CICR) via ryanodine receptors. To our knowledge, this is the first in vivo demonstration of NMDA-evoked CICR. We postulate the usefulness of microdialysis in such studies

Methods of retrograde axonal transport were employed to evaluate the topography and overlap of claustroneocortical connections in the rat. Fluorescent tracers Fast Blue (FB) and Diamidino Yellow (DY) were injected simultaneously in various combinations into the motor, somatosensory, auditory and visual cortical areas. Experiments showed that claustroneocortical projections are organized in two main cortico-related zones: sensorimotor and visuoauditory. The sensorimotor zone occupies the anterodorsal part whereas the visuoauditory occupies the posteroventral part of the claustrum. Between these two main zones only a scanty overlap was observed. In the sensorimotor zone a large overlap between neurons projecting to the motor and somatosensory cortical areas exists. The visuoauditory zone is characterized by a full overlap of neuronal populations projecting to the visual and auditory areas

The morphology of claustral neurons projecting to the motor, somatosensory, auditory and visual cortical areas in the rat was analyzed by means of combination of axonal retrograde transport and morphometric analysis. Fluoro-Gold (FG) injections placed into various cortical fields resulted in labeling in the claustrum four neuronal types: pyramidal with thick main dendrite, oval with a few thin dendrites spreading out in various directions, fusiform possessing two main dendrites arising from opposite poles of the cell body and polygonal. Pyramidal neurons prevailed in populations of neurons projecting to the motor cortex of the contralateral hemisphere. Oval neurons outnumbered other types in populations projecting to the somatosensory, auditory and visual cortical fields. The number of fusiform and polygonal neurons did not exceed at 12.5% together in any populations. Neurons projecting to the contralateral hemisphere were the largest claustral neurons (mean cross-section are 167.19 +/- 2.9 micron 2) whereas neurons projecting to the motor cortex where the largest claustral neurons projecting ipsilaterally (141.89 +/- 2.22 micron 2). There was no significant difference between neurons projecting to the somatosensory (113.46 +/- 1.9 micron 2) cortex and to the visual (111.8 +/- 1.4 micron 2) cortex whereas neurons related to the auditory are (95.98 +/- 1.75 micron 2) were the smallest claustral neurons. These observations pointed out that the morphology of claustral neurons is closely related to a cortical area to which they send axons

The topography and cytoarchitectonics of the claustrum as well as morphometric parameters of its neurons were studied in 10 human brains obtained from patients without any detectable neuropathological changes. We distinguished four parts of the claustrum: dorsal, orbital, temporal and paraamygdalar. The dorsal and orbital parts contain larger cells, than those of the temporal and paraamygdalar parts, although these differences were statistically non significant. The highest neuronal density was observed in the paraamygdalar part. The nucleus and nucleus@cell body area ratio was significantly smaller in the dorsal part than in other parts of the claustrum. We described three types of neurons in the claustrum: (1) medium-sized either fusiform or triangular cells with darkly stained cytoplasm; they predominate in the dorsal and temporal parts, (2) medium-sized as well as large cells, either multipolar or pyramidal-like with lightly stained cytoplasm; they are most numerous in the orbital and paraamygdalar parts, (3) small, multipolar or oval neurons with darkly stained ring of cytoplasm; these types of neurons are uniformly distributed throughout all parts of the claustrum. The subdivision of the human claustrum is in accordance with our observations that each of these parts possesses connections with different cortical regions.

The influence of fixation and histological procedure on morphometric parameters of human and rat brains were analyzed. We found negative correlation of brain weight with the fixation time both in human and rat material. Human brains fixed for shorter period of time than one year showed large differences in weight. The group of brains fixed for longer than one year was characterized by much smaller dispersion. Moreover, we found that during histological preparation the shrinkage of brain tissue of older human subjects is statistically significantly smaller than younger ones (below 65 years). In order to correct these artifacts easily we propose three simple formulas.