Selective cell death of hyperploid neurons in Alzheimer's disease

Abstract

Aneuploidy, an abnormal number of copies of a genomic region, might be a significant source for neuronal complexity, intercellular diversity, and evolution. Genomic instability associated with aneuploidy, however, can also lead to developmental abnormalities and decreased cellular fitness. Here we show that neurons with a more-than-diploid content of DNA are increased in preclinical stages of Alzheimer's disease (AD) and are selectively affected by cell death during progression of the disease. Present findings show that neuronal hyperploidy in AD is associated with a decreased viability. Hyperploidy of neurons thus represents a direct molecular signature of cells prone to death in AD and indicates that a failure of neuronal differentiation is a critical pathogenetic event in AD.

Figure 1

Hyperploid neurons in Alzheimer’s disease (AD). A: Examples of diploid (2n) and tetraploid (4n) neurons in AD entorhinal cortex, detected by CISH and FISH using chromosome-specific probe for chromosome 17 (ZytoDotCEN probes, ZytoVision, Germany, which target α-satellite-sequences of the centromere of the respective chromosome). B: Examples of hyperploid and diploid (arrow) pyramidal neurons in AD, identified by the abundance of Nissl substance, the typical shape of their soma and the apical dendrite approaching from the soma. Note the presence of hyperploid cells in the vicinity of a diploid pyramidal neuron in the far right panel (CISH with chromosome-specific probe for chromosome 17, counterstained for Nissl substance with cresyl echt violet). C: Mapping of neurons throughout the entorhinal cortex according to their DNA content by slide-based cytometry (SBC) in a case of mild AD (2n: diploid content of DNA; 4n: tetraploid content of DNA). D: To ensure the neuron-specificity of the SBC analysis, only cells immunoreactive for neurofilament (mab SMI 311: green) were considered (propidiumiodide signal: red). E: Intermethod reliability of SBC and quantitative CISH with a ZytoDotCEN 17 probe (ZytoVision, Bremerhaven, Germany). Linear regression analysis fits data with a correlation coefficient according to Bravais-Pearson of r = 0.93; P < 0.001.

Figure 2

Quantification of hyperploid neurons in the entorhinal cortex by SBC in healthy controls and cases with preclinical, mild, and severe AD. Mean values ± SEM; differences are significant at the following levels: *P < 0.05, ***P < 0.001, Student t test.

Figure 3

Relation between changes in the number of diploid and hyperploid neurons and total neuronal loss. Total loss of neurons during progression of AD highly correlates with changes in the number of hyperploid neurons (regression line 2: r = 0.86; P < 0.001) but not with diploid neurons (regression line 1: r = 0.18; n.s.). Hyperploid neurons (red) account for a loss of 60 of 67 neurons, which corresponds to 89% of the total neuronal loss.

Figure 4

Relation between changes in the number of hyperploid neurons and total neuronal loss in healthy controls and preclinical, mild, and severe AD. Hyperploid neurons are increased already in preclinical AD. They remain stable in number during transition from preclinical to mild AD and decrease with further progression of the disease. Linear regression analyses are indicated by dotted lines: 2, mild plus severe AD, r = 0.86, P < 0.001; 3, healthy controls, r = 0.71, P < 0.01; 4, preclinical AD, r = 0.50, n.s.; 5, preclinical plus mild AD, r = 0.36, n.s.