Epidemiological studies have indicated that cholesterol-lowering drugs, such as the statins, can reduce the risk of developing Alzheimer's disease (AD). Paradoxically, however, reducing cholesterol levels in the rodent brain seems to promote neurodegeneration. To try to resolve this apparent contradiction, Abad-Rodriguez and colleagues carried out a more direct investigation of the effects of neuronal membrane cholesterol levels on the production of the amyloid-β (Aβ) peptide — the main constituent of the amyloid plaques that form in the brains of patients with AD.

Cleavage of amyloid precursor protein (APP) to generate Aβ was previously thought to occur predominantly in cholesterol-rich microdomains of the neuronal membrane, known as rafts or detergent-resistant membrane (DRM) microdomains. However, the evidence for this model came largely from experiments that involved overexpression of amyloid precursor protein (APP) and the APP-cleaving secretase enzymes.

Abad-Rodriguez et al. showed that when APP and the β-secretase BACE1 were expressed at physiological levels in the membranes of human and rodent hippocampal neurons, APP was almost entirely excluded from the DRMs, whereas BACE1 was present in both DRM and non-DRM fractions. If the membrane fluidity was increased by reducing the cholesterol level, APP and BACE1 were co-localized more frequently, and this led to a rise in Aβ production.

These findings contradict the idea that Aβ synthesis takes place in the DRM domains — in fact, the cholesterol in these domains seems to function as a barrier to the interaction between APP and BACE1. Loss of cholesterol from the membrane probably releases BACE1 from the DRMs, making it more likely to encounter APP in the more fluid regions of the membrane.

So, how can these observations be reconciled with the effects of statins in humans? As Abad-Rodriguez et al. point out, most of the commonly used statins are poor penetrators of the blood–brain barrier, so their benefits might derive from anti-inflammatory or antioxidant properties rather than direct effects on brain neurons. Their new findings indicate that, rather than trying to reduce brain cholesterol, care should actually be taken to preserve the cholesterol balance in the membranes of brain neurons.