In an investigation of heat shock proteins (HSPs) in the brains of Alzheimer's disease (AD) patients and cognitively intact control subjects, we found that 2 HSPs, termed "HSP72" and "GRP78," underwent major changes in expression in AD. HSP72, which was present at very low levels in control brains, increased dramatically in AD patients, and was localized exclusively in neuritic plaques and neurofibrillary tangles. We hypothesize that HSP72 is induced as an early response to the formation of abnormal proteins, perhaps targeting them for proteolysis. In contrast, GRP78 increased in AD only in neurons that remained cytologically normal, especially in the CA3 subfield of the hippocampus and the deep layers of the entorhinal cortex. The increased expression of GRP78 within successfully surviving neurons suggests that this protein may protect such cells from AD-specific damage.

Whether all etiologic forms of Alzheimer's disease (AD) share a final common pathway is a major issue. We determined the severity and regional distribution of neuronal loss, amyloid plaques, neuritic plaques (NPs), and neurofibrillary tangles (NFTs), and calculated the ratio of neuronal loss to NPs and NFTs in brains of 19 familial AD (FAD) patients with linkage to chromosome 14, six AD patients with mutations of chromosome 21 (codon 717 of the beta-amyloid percursor protein gene), and 11 sporadic AD (SAD) patients. There was no difference in the pattern of distribution of the various pathologic features or in the ratio of neuronal loss to NPs or NFTs in any AD group. However, FAD groups could be distinguished from SAD by the greater severity and the lack of influence of apolipoprotein E genotype on pathology. These differences may reflect differences in age at onset rather than different etiopathologic mechanisms. The similarity of pathologic findings in the different AD groups provides evidence for a final common pathophysiologic pathway in AD.

No abstract

To determine whether apolipoprotein E epsilon 2 (APOE-epsilon 2) affects neuropathology in aging and Alzheimer's disease (AD), we compared beta-amyloid plaque (A beta P) and neurofibrillary tangle densities, neuropil thread formation, and amyloid angiopathy in five APOE-epsilon 2/3 AD patients, five APOE-epsilon 3/3 AD patients, five APOE-epsilon control patients, and five APOE-epsilon 3/3 control patients. We examined the (frontal and parietal) neocortex, hippocampus, entorhinal cortex, and cerebellum and found A beta P densities to be lower (t = 3.121, p = 0.011) in the cortex of APOE-epsilon 2/3 AD patients than in APOE-epsilon 3/3 AD patients. Amyloid angiopathy was also less in APOE-epsilon 2/3 patients than in APOE-3/3 patients (U = 4.500, p = 0.027). Control APOE-epsilon 2/3 brains had little AD-related pathology; even our 102-year-old control case showed few A beta Ps compared with the elderly APOE-epsilon 3/3 cases. The APOE-epsilon 2/3 genotype may influence pathologic phenotype in some aged normal and AD populations.

To determine whether apolipoprotein E (APOE) genotype affects neuropathology in Lewy body disease (LBD), we examined 18 cases of LBD that did not have concurrent Alzheimer's disease by the CERAD criteria. We obtained APOE genotypes, determined diffuse beta-amyloid plaque (A beta P) and Lewy body densities in multiple brain regions, and graded the intensity of CA2-3 ubiquitin-positive neurites, vacuolar change, nigral pathology, amyloid angiopathy, and subpial amyloid deposition. The APOE allele frequencies were as follows: epsilon 2, 0.14 +/- 0.07; epsilon 3, 0.64 +/- 0.08; and epsilon 4, 0.22 +/- 0.03. The mean A beta P density was lower in APOE epsilon 3/3 cases (14.5 A beta Ps per mm2) than in the groups with the APOE epsilon 2 (67.0) or APOE epsilon 4 (46.6) alleles. This difference was due largely to the difference between A beta P density in the APOE epsilon 2 group and the APOE epsilon 3/3 group (F = 5.525, p < 0.02). CA2-3 neuritic degeneration was greater in those with the APOE epsilon 4 allele than in those with the APOE epsilon 3/3 genotype (grade = 1.9 +/- 1.3 versus 0.938 +/- 0.9; Kruskal-Wallis test statistic = 6.962, p < 0.05). These data are consistent with the hypothesis that APOE genotype may affect neuropathology in LBD.