Brains of patients with Alzheimer disease/senile dementia of Alzheimer type (AD/SDAT) develop a progressive accumulation of amyloid, which deposits primarily in the form of characteristic parenchymal 'plaques' (senile or neuritic plaques/SP's) and as mural deposits in the walls of capillaries and arterioles (cerebral amyloid angiopathy /CAA). A major component of this amyloid is a small and unique peptide composed of 39-43 amino acids, beta/A4, which is cleaved from a much larger precursor protein (APP) that has several isoforms. Brain amyloid can be detected in autopsy or biopsy brain tissue by classical, immunohistochemical and ultrastructural (including immuno-electron microscopic) methods of varying sensitivity and specificity. Beta/A4 amyloid deposition is remarkably variable (e.g. predominantly parenchymal or vascular, or a mixture of parenchymal and vascular) among patients with AD/SDAT. Despite its abundance in the brains of AD/SDAT patients, the precise role of beta/A4 in the pathogenesis of the neurological deficit, neocortical atrophy and progressive synapse loss associated with AD/SDAT has yet to be determined. However, mutations in the gene that encodes APP are clearly associated with familial AD syndromes in which there is significant brain amyloid deposition. CAA, in addition to its association with AD/SDAT, can result in hemorrhagic and (possibly) ischemic forms of stroke. Work with recently developed transgenic mice which express large amounts of beta/A4 in the central nervous system is likely to elucidate mechanisms by which the protein is selectively or deposited in the brain in a parenchymal or microvascular form, and how it contributes to the pathogenesis of neurodegeneration.

Brain biopsy specimens from five patients with Alzheimer's disease obtained in the course of a trial of intracerebroventricular bethanechol were studied by immunohistochemical (antibody to A4 peptide) and ultrastructural techniques, with particular emphasis on the microvessels. In some cases, numbers of A4-immunoreactive lesions (senile plaques) correlated well with numbers of plaques demonstrable by silver stains. Prominent A4-immunoreactive amyloid angiopathy was seen in one patient. The patient with severe cerebral amyloid angiopathy (CAA) showed extensive arteriolar deposition of amyloid filaments with apparent destruction of the media but remarkably intact endothelium. A cell of origin for amyloid filaments was not apparent, although close proximity to smooth muscle cell remnants in the arteriolar media suggested this as one possible cell of origin. Frequent vessels showed medial or adventitial collagen deposition, even when the amount of amyloid was minimal or negligible. Thus relatively severe CAA can exist in the absence of overt endothelial injury, although related studies on this tissue indicate definite abnormalities of the blood-brain barrier. Conversely, destruction of smooth muscle cells and collagen deposition in vessel walls may be the cellular correlates of arteriolar weakening that can lead to CAA-related brain hemorrhage.

Extensive surgical resections of neocortical cerebral tissue (including hemispherectomies) from 13 infants and children with infantile spasms showed that 12 of 13 specimens contained either malformative and dysplastic lesions of the cortex and white matter (sometimes with associated hamartomatous proliferation of globular cells), or destructive lesions possibly acquired as a result of anoxic-ischemic injury, or a combination of the two. In brain tissue from 4 patients, coarse neuronal cytoplasmic fibrils resembling neurofibrillary tangles were seen in areas of dysplastic brain on silver-stained (Bielschowsky technique) sections. Immunohistochemical (immunoperoxidase) study of cortical lesions containing globular cells employing primary antibodies to glial fibrillary acidic protein and synaptophysin as markers of astrocytic and neuronal differentiation, respectively, revealed that many cells showed astrocytic and/or neuronal features, suggesting the local proliferation of primitive or multipotential neuroectodermal cells as one substrate for this seizure disorder. Morphological abnormalities of a severe degree and wide extent in the resected tissue (e.g., in one patient with hemimegalencephaly) often showed features to suggest that they may represent variants of tuberous sclerosis. These most likely result from abnormal movement and/or local proliferation of neuroectodermal precursors that have migrated from the germinal matrix to the cortical mantle. Cellular, molecular and neurophysiological study of these abnormalities is likely to yield information about basic molecular mechanisms of brain malformation and injury important in the pathogenesis of infantile spasms and other forms of focal or generalized epilepsy.

No abstract