Ab could be a marker, and neurotoxicity could be directly driven by a loss of APP function:

- No relationship between Ab deposits and degenerating neurons (see Braak or Delacourte)

- Ab deposits are produced simultaneously to tau pathology and not years before

- Ab deposits could be neurotoxic from the inside of the neurons, or the outside. 10 different ways used by this killer have been described. Which one is the good one, if any.

- Working on secretases to reduce Ab production is risky, since there is no demonstrated overproduction in sporadic AD (99% of all AD cases). Ab is a physiological peptide, and reducing its production could be deleterious.

Gamma secretase is a difficult target:

- it could activate Notch and be deleterious

- 10 different proteins are cut by a gamma-like activity, increasing by a log the possibility of adverse events

- gamma-secretase activity results from a complex with nicastrin, APH-1 and pen-2, showing the complexity of the target

BACE is a dangerous target:

- Inhibition of BACE prevents the formation of Ab, but simultaneously, it prevents the formation of Beta stubs, that are normal and major products in the brain tissue. If beta stubs have a physiological role, and this is likely, inhibiting BACE could accelerate the pace of Alzheimer's disease.

- Inhibition of BACE for a reduction of Ab which is not obviously increased in sporadic AD has a rational of its own.

APP and Ab are two different targets. All data demonstrate that APP is a central cause of AD. Ab could be the reflect of APP dysfunction, but not the cause.

- I