Quantitative analysis of a vulnerable subset of pyramidal neurons in Alzheimer's disease: II. Primary and secondary visual cortex.

In this study we investigated the primary and secondary visual areas of normal and Alzheimer's disease brains by using the SMI32 antibody. It is known that in Alzheimer's disease primary sensory areas are usually less devastated than association cortices, although visual symptomatology has been documented early in the course of the disease. In area 17, the SMI32 antibody primarily labeled the perikarya and dentritic tree of the large Meynert cells and cells in layer IVB. Smaller neurons in layers III, V, and VI were also immunoreactive (ir). In area 18, very large SMI32-ir pyramidal neurons in layers III and V were observed. In both areas, staining intensity was correlated with cell size, the largest neurons being the most intensely stained. Only a few changes were observed in the Alzheimer's disease cases. The only statistically significant differences in SMI32-ir neuron counts between control and Alzheimer's disease brains occurred in layer IVB cells and Meynert cells in area 17, and in layer III cells in area 18. In contrast with association cortices, there were no changes in staining intensity in the visual areas. There were fewer neurofibrillary tangles and neuritic plaques in these areas than in prefrontal and inferior temporal cortex, and a correlation between neurofibrillary tangle counts and SMI32-ir neuron loss was only observed in layer III of area 18. These observations show that in the primary and secondary visual cortex, SMI32 also labeled a distinct subset of pyramidal cells that are known from data obtained in the monkey brain to furnish long corticocortical as well as subcortical projections. Interestingly, although there is much less cell and/or neurofibrillary tangle formation in these occipital regions than in prefrontal and temporal association areas, there is significant loss within key subsets of pyramidal cells. The selective loss of this particular subpopulation of pyramidal neurons will disrupt association pathways linking primary visual cortex with areas involved in higher level visual processing. The partial disconnection of such pathways may be relevant to the visual symptomatology frequently observed in Alzheimer's disease patients. These data further support the hypothesis that subtypes of pyramidal neurons with specific anatomical and molecular profiles may display a differential vulnerability in Alzheimer's disease.