30 September 2004. γ-secretase, the multiprotein complex that unleashes amyloid-β1-42 (Aβ1-42) from the amyloidβ precursor protein (AβPP), seems an obvious place to start if you want to develop a drug to prevent or slow Alzheimer disease. There’s only one problem; γ-secretase is also needed to process other proteins, including the cell surface receptors Notch and ErbB-4, so axing γ-secretase may not be the best approach. But don’t fret; all may not be lost. Recent reports have suggested that rather than cripple the protease, some NSAIDs, or nonsteroidal antiinflammatory drugs, jolt it sufficiently so that it cuts AβPP at a different spot, yielding a shorter, less toxic amyloid peptide, Aβ1-38 (see ARF related news story and ARF news story). Now, reporting in this month’s Nature Medicine, researchers from the Alzheimer Research Unit at Massachusetts General Hospital offer a mechanistic explanation for this selective inhibition.
Brad Hyman, Brian Bacskai and colleagues did FRET. They used a nifty technique called fluorescence resonance energy transfer (FRET) to measure the conformation of presenilin 1 (PS1), the core protease of γ-secretase, in the presence or absence of different NSAIDs. FRET is capable of measuring the proximity of one fluorescent molecule to another, but in a variation on this theme, first author Alberto Lleo and colleagues labeled the N- and C-terminal ends of PS1 with fluorophores. The ends of the protease are sufficiently close, it turns out, so that photons emitted from the donor fluorophore at the N-terminus are absorbed by, and excite, the acceptor fluorophore at the C-terminus, allowing FRET to occur intermolecularly.
FRET depends on proximity, but Lleo found that resonance transfer between the N- and C-termini of PS1 was less efficient in the presence of ibuprofen. The fluorescence lifetime of the donor jumped from 1,800 picoseconds to 2,200 picoseconds when the NSAID was added to cells expressing PS1 and AβPP, indicating that the N- and C-terminal ends of presenilin had been pushed apart. Significantly, aspirin and naproxen—NSAIDs that have not been found to alter AβPP cleavage (see ARF related news story on the failure of a naproxen clinical trial)—had no effect on the process.
The authors also found that ibuprofen (both optical isomers) reduced proteolysis at the 42 position by about 50 percent, and that this was compensated by cleavage, yielding peptides shorter than 40 amino acids. They also found that Notch processing was unaffected. To determine if NSAIDs have similar effects in vivo, Lleo and colleagues fed mice an ibuprofen diet for 15 days, fixed the brain tissue, and used fluorescent labeled antibodies to the N- and C-termini of PS1 to measure their proximity. Again, FRET showed that in PS1 exposed to ibuprofen, a conformational change pushed the termini apart.
“Our results suggest that NSAIDs have an allosteric effect on γ-secretase,” write the authors, and they offer a model whereby the distance between the N- and C-termini of PS1 is crucial in determining the proteolytic cleavage site of AβPP. This is welcome news for those who are trying to find drugs that selectively reduce Aβ1-42 but do not affect processing of other γ-secretase targets. It also helps explain why NSAIDs that do not affect AβPP processing have so far proven ineffective in halting or preventing AD (see ARF related news story and ARF news story), while ibuprofen may prove better suited (see ARF related news story).—Tom Fagan.
Lleo A, Berezovska O, Herl L, Raju S, Deng A, Bacskai BJ, Frosch MP, Irizarry M, Hyman BT. Nonsteroidal anti-inflammatory drugs lower Aβ42 and change presenilin 1 conformation. Nature Medicine. 2004 September 26. Advanced online publication. Abstract