The membrane protease γ-secretase, which processes the amyloid precursor protein (APP) to produce Aβ42, the 42-residue isoform of the amyloid-β peptide, has been a popular therapeutic target for Alzheimer's disease. Many efforts have focused on developing direct protease inhibitors, but intriguingly, a subset of non-steroidal anti-inflammatory drugs (NSAIDs) have been found to act as γ-secretase modulators (GSMs) that can selectively lower Aβ42 levels. Now, writing in Nature, Kukar et al. shed new light on the underlying mechanism, showing that GSMs target APP, the enzyme's substrate, rather than the enzyme itself.

Using two different biotinylated photoactivatable GSMs — fenofibrate (Fen-B) and a tarenflurbil derivative (Flurbi-Bpb) — in human neuroglioma cells, the authors found that these agents did not label any of the core components of the γ-secretase complex, but instead bound to APP. Although the GSM photoprobes could also bind to Notch (another γ-secretase substrate) in vitro, the binding affinity was considerably lower than that for APP. Targeting APP is likely to be a common mechanism of action for GSMs, as both Aβ42-lowering and Aβ42-raising GSMs competed with Fen-B for APP labelling.

Interestingly, the binding site of GSMs resides in a region of APP that has previously been shown to be crucial for amyloid-β aggregation. Furthermore, the authors show that several structurally unrelated compounds that bind to this region can also act as GSMs (altering the levels of Aβ42 and decreasing amyloid-β aggregation in vivo), and that mutation of this region alters the sensitivity of APP to GSMs.

Disappointingly, it has recently been announced that tarenflurbil failed in a late-stage clinical trial in Alzheimer's disease (http://www.myriad.com/news/release/1170283). Nevertheless, the findings from Kukar et al. suggest that it might be possible to develop optimized small molecules that can decrease Aβ42 production and prevent amyloid-β aggregation by selectively binding to APP. Furthermore, the study also raises the exciting possibility that substrate targeting might represent a novel general therapeutic strategy for modulating proteolytic enzymes.