To improve γ-secretase modulators, which tweak rather than block the enzyme, scientists need to understand where these compounds bind and how that interaction modifies protease activity. Reporting July 8 in the Proceedings of the National Academy of Sciences, scientists led by Taisuke Tomita of the University of Tokyo identified an extracellular pocket on γ-secretase that binds one class of modulator. This interaction adjusts the intramembrane catalytic center of the enzyme and makes it more likely to cleave amyloid precursor protein into Aβ38 and smaller peptides that are more innocuous than Aβ42 peptides.
The paper gives the most detailed look yet at a γ-secretase modulator (GSM), said Steve Wagner, University of California, San Diego, who was not involved in the project. “The study provides further insight into how these molecules work, where they bind, and shows that they are truly allosteric modulators.”
After β-secretase, γ-secretase delivers a second cut that removes Aβ42 from the amyloid precursor protein (APP). By coaxing the enzyme to pump out slightly shorter forms of Aβ, modulators may avoid toxic side effects associated with γ-secretase inhibitors, which completely block the processing of APP and other substrates, including the important signaling receptor Notch. Such side effects likely played into the failure of semagacestat, Eli Lilly and Company’s γ-secretase inhibitor, which made patients worse (see Aug 2010 news story).
GSMs fall into two broad categories—acidic compounds derived from nonsteroidal anti-inflammatory drugs, and more potent non-acidic phenylimidazole derivatives. A previous study found that this latter variety bound to the N-terminal fragment of presenilin (PS), the catalytic subunit of γ-secretase (see Sep 2011 news story). However, the exact binding site and mode of change still eluded scientists.
To confirm that the non-acidic GSMs bound PS, first author Koji Takeo and colleagues used photoaffinity labeling. They modified a phenylimidazole-type GSM, called ST1120, and found that it bound directly to both PS1 and PS2.
To pinpoint the specific amino acids that bound the GSM, the authors substituted conserved residues in PS2 one at a time with alanine. ST1120 failed to bind when either of two specific tyrosines (Y112 or Y246), or an asparagine (N141) were replaced. When Takeo substituted these amino acids with cysteine they bound a probe that cannot penetrate the cell membrane, indicating that they sit on the outside of the cell. Comparing the mammalian amino acid sequence with that of a homologous archaeal enzyme for which a crystal structure had already been derived confirmed that these amino acids occur in three extracellular regions of the enzyme that together form a pocket—hydrophilic loop 1, transmembrane domain (TM) 2, and TM5.
How, then, does ST1120 alter enzyme activity, given that the catalytic center is buried in the cell membrane? The researchers previously reported that two leucine residues—381 and 383—face the hydrophilic catalytic core inside the membrane. When they replaced either of these with cysteine and then measured how well each residue bound a probe, L381C bound less in the presence of ST1120, but L383C bound more. This suggested that ST1120 changes the conformation of the catalytic site. In a cell-based assay, ST1120 reduced production of Aβ42 and Aβ40, while it boosted levels of Aβ37 and Aβ39. ST1120 also accelerated proteolysis. Put together, the results suggest that ST1120 allosterically alters the catalytic site, speeds up the enzyme’s reactions, and tips Aβ production toward shorter products.
The new understanding of GSM interactions may lead to better therapeutics down the road, most researchers noted. “This integrated approach helps us get a better handle on GSMs’ mechanism of action,” said Yueming Li, Memorial Sloan Kettering Cancer Center, New York. “Any knowledge of how they work will help us design better modulators.” A recently published article that details the three-dimensional structure of human γ-secretase could help solve the mechanism further (see Lu et al., 2014).
Wagner cautioned against generalizing these results to all non-acidic GSMs, as they differ in structure, potency, and length of Aβ peptides released. Other non-acidic GSMs may bind different regions on similar subunits, he said.
“This and more work along these lines will ultimately provide the level of structure–function insights needed to develop better and safer drugs that target γ-secretases,” Bart De Strooper and Lucia Chávez-Gutiérrez, KU Leuven, Belgium, wrote to Alzforum in an email (see full comment below).—Gwyneth Dickey Zakaib
- Lilly Halts IDENTITY Trials as Patients Worsen on Secretase Inhibitor
- Evidence Mounts That Some γ-Secretase Modulators Bind Presenilin
- Lu P, Bai XC, Ma D, Xie T, Yan C, Sun L, Yang G, Zhao Y, Zhou R, Scheres SHW, Shi Y. Three-dimensional structure of human γ-secretase. Nature (2014) doi:10.1038/nature13567
- Takeo K, Tanimura S, Shinoda T, Osawa S, Zahariev IK, Takegami N, Ishizuka-Katsura Y, Shinya N, Takagi-Niidome S, Tominaga A, Ohsawa N, Kimura-Someya T, Shirouzu M, Yokoshima S, Yokoyama S, Fukuyama T, Tomita T, Iwatsubo T. Allosteric regulation of γ-secretase activity by a phenylimidazole-type γ-secretase modulator. Proc Natl Acad Sci U S A. 2014 Jul 22;111(29):10544-9. Epub 2014 Jul 9 PubMed.