D. M. Bowen

Atrophy with ageing of human whole brain, entire temporal lobe, and caudate nucleus was assessed in autopsy specimens, by biochemical techniques. Only the caudate nucleus showed changes. Markers for several neurotransmitter systems were also examined for changes with age. In neocortex and temporal lobe of human brain, small decreases were detected in markers of cholinergic nerve terminals, whereas a large decrease (79%) occurred in the caudate nucleus. Findings were similar in striatum from 3--33-month-old rats. No change occurred in binding of [3H]quinuclidinyl benzilate by human samples. Markers of serotonergic terminals were also unchanged in human and rat brain. By contrast, binding of [3H]lysergic acid diethylamide and [3H]serotonin was decreased (32-81%) in human neocortex and temporal lobe, but not in caudate nucleus. A 43% loss of a marker of gamma-aminobutyrate terminals occurred in human neocortex, while [3H]muscimol binding increased (179%). No changes were detected in markers of catecholamine synapses in temporal lobe or rat striatum. Hence, with human ageing there appears to be a loss of markers of gamma-aminobutyrate neurones intrinsic to neocortex and acetylcholine cells intrinsic to the caudate nucleus, as well as a change in postsynaptic serotonin receptors in neocortex. These losses are accompanied by relative preservation of markers of ascending projections from basal forebrain and brain stem.

A quantitative study has been made of the number of neurofibrillary tangles and of the choline acetyltransferase activity in several sites in the cerebral hemispheres of eight patients who had had Alzheimer's disease. The neurofibrillary tangles were maximal in structures in the medial temporal lobe (uncus, amygdala, hippocampus and parahippocampal gyrus), severe in the neocortex on the lateral surface of the temporal lobe, moderate in the "association cortex" of the parietal and frontal lobes and minimal in primary somatic and visual sensory areas. There was a significant decrease in choline acetyltransferase activity in almost all areas, and the means of the percentage decreases for the different groups of areas correlate well with the counts of the neurofibrillary tangles. These results support the hypothesis that the pathological process in Alzheimer's disease may spread along a sequence of corticocortical connections between the main sensory areas and the hippocampal formation. The disease process may also spread along the reciprocal connections between the amygdala and the neocortex because the numbers of tangles in different areas of the neocortex closely parallel the density of their connections and the amygdala.

Cell counts have been performed on cholinergic subcortical nuclei, dorsal raphe nucleus, and locus caeruleus from up to 18 cases of Alzheimer's disease and 10 age-matched control subjects. In general, the extent of cell loss in these structures was similar. In the basal nucleus the anteromedial subdivision was the least, and the posterior subdivision the most affected. A subgroup of demented subjects with Alzheimer's disease had a relatively preserved basal nucleus, and frontal lobe (CAT) choline acetyltransferase activities similar to those in control subjects, but significantly more neuronal loss in the locus caeruleus.

Neocortical tissue prisms prepared from rat and human brain were frozen to -196 degrees C by a two-step freezing procedure and 10% dimethyl sulphoxide as cryoprotectant. Frozen and thawed rat neocortical prisms incorporated glucose into acetylcholine and carbon dioxide at 89% and 86% of control values, respectively, and noradrenaline uptake into frozen and thawed rat prisms was 94% of the control value. Frozen and thawed prisms from three human neocortical specimens showed a similar degree of protection from freeze-thaw injury.