24 August 2004. "Where are you? Who are you with?" These questions are commonly asked of teenagers. γ-secretase, one of the key enzymes in production of amyloid-β peptides and now a teenager itself, has been reluctant to own up, though little by little researchers have probed and prodded until they have built a fairly decent picture of its social life. Now in press in the Journal of Biological Chemistry, three papers reveal a little more about where the enzyme hangs out and with whom it associates.
First, the hangout. One of the big questions has been whether the γ-secretase complex acts primarily in the cell membrane or before it gets to the cell surface (see ARF related news story). The August 16 paper from Mikhail Khvotchev and Thomas Sudhof at the University of Texas Southwest Medical Center in Dallas suggests the latter. To demonstrate this, the authors capitalized on an interesting discovery about syntaxin1A: When truncated at amino acid 243, the synaptic membrane protein prevents exocytosis without compromising molecular traffic through the Golgi apparatus. Full-length syntaxin, on the other hand, blocks both trans-Golgi protein traffic and exocytosis. The smaller of the two proteins should, therefore, allow normal Golgi-based maturation of proteins, but should stop them from reaching the cell surface, while the full-length syntaxin should prevent protein maturation. So, would they both inhibit γ-secretase activity?
Not unexpectedly, when Khvotchev coexpressed the full-length syntaxin with AβPP in cultured cells, he found that AβPP processing was completely abolished. C-terminal fragments, usually released following α-/β-secretase cleavage, could not be detected, for example. However, in the presence of the 243-amino acid fragment, processing proceeded as normal. C-terminal fragments were as abundant as in untreated cells, and they became even more abundant when the cells were treated with a γ-secretase inhibitor, indicating that this enzyme was also fully functional in the absence of exocytosis. In support of this the authors found that cells expressing the 243-amino acid syntaxin failed to secrete Aβ40, but instead accumulated vast amounts of it intracellularly (5-6 times the levels found in control cells). In addition, in experiments where the AβPP intracellular domain was engineered to activate a Gal4-dependent reporter system, Khvotchev found that the reporter was turned on in the presence of truncated syntaxin1A, but not in the presence of the full-length protein.
“Our data demonstrate that AβPP cleavage by α-, β-, and γ-secretases largely requires that AβPP traverse the Golgi complex, but that it can occur at normal rates without AβPP ever reaching the cell surface,” write the authors. The finding raises several fundamental questions about the role of AβPP in vivo, perhaps most notably, whether it behaves similarly to Notch, another transmembrane protein that is cleaved by γ-secretase.
Notch cleavage is activated by extracellular ligands, such as Jagged and Delta (see ARF related news story), but this new data “implies that cleavage cannot be activated by an exogenous ligand, although it could still be controlled by an endogenous ligand, as suggested by the finding that F-spondin binds to AβPP and inhibits AβPP cleavage,” Khvotchev and Sudhof write (see ARF related news story on F-spondin and the evidence that AβPP may bind to itself in an ARF related news story).
Support for the idea that γ-secretase may be hanging out below the cell surface came one day later from a group led by Gopal Thinakaran at the University of Chicago. Kulandaivelu Vetrivel and colleagues used subcellular fractionation techniques to try to locate the bulk of the proteolytic enzyme. They found all four known components of γ-secretase, presenilins, nicastrin, APH-1 and PEN-2, in detergent-insoluble, cholesterol-laden lipid rafts. Significantly, the fractions containing γ-secretase were poor in plasma membrane markers, such as Na+/K+ ATPase. The authors confirmed this using confocal microscopy, finding the secretase colocalized with markers of the trans-Golgi network and of endosomes (including syntaxin 6 and 13, and VAMP4), but not with plasma membrane markers. In addition, immuno-isolation of rafts using antibodies to syntaxin 6 recovered presenilins but not SNAP-23, a protein found in plasma membrane rafts. Based on the results, the authors concluded that “a significant fraction of mature γ-secretase resides in lipid raft microdomains of post-Golgi and endosome membranes.” The work supports previous experiments indicating that β-secretase action occurs in lipid rafts (see ARF related news story).
As for whom γ-secretase is hanging out with, the aforementioned nicastrin, APH-1, and PEN-2 are well-known partners, but how they function has not been fully explored (see ARF related news story). PEN-2 (presenilin-enhancer 2), for example, is essential for proper assembly of the γ-secretase complex, but how, and what parts of the enhancer contribute, are unclear. Now, Peter St. George-Hyslop and colleagues at the University of Toronto and New York University Medical Center reveal that the C-terminus is the business end of PEN-2.
First author Hiroshi Hasegawa and colleagues focused on this end of the molecule because there are sequences there that are conserved among PEN-2 molecules from species as diverse as humans and roundworms. When the authors deleted short stretches of the C-terminus, particularly amino acids 90-96, they found that PEN-2 fails to bind to the γ-secretase complex. But in addition, the researchers noted that mutants lacking the very end of the molecule (up to amino acid 101) failed to fully rescue γ-secretase activity when endogenous PEN-2 was “knocked down” by RNAi. This suggested that either the C-terminal sequence or the length of the C-terminus was critical for the molecule to live up to its full potential. To determine which, the authors tested a series of mutants, one with four extra amino acids, one with alanine in place of three conserved residues (isoleucine, proline and glycine), and one with the last two amino acids removed. Only the first mutant was fully active, suggesting that it is indeed the length of the C-terminus that is critical, a finding that explains why C-terminal-tagged PEN-2 had previously failed to rescue C. elegans PEN-2 null mutants (see Francis et al., 2002 and ARF related news story).—Tom Fagan.
Khvotchev M, Sudhof TC. Proteolytic processing of APP by secretases does not require cell-surface transport. J. Biol. Chem. 2004. Aug 16. Abstract
Vetrivel KS, Cheng H, Lin W, Sakurai T, Li T, Nukina N, Wong PC, Xu H, Thinakaran G. Association of gamma -secretase with lipid rafts in post-golgi and endosome membranes. J Biol Chem. 2004 Aug 17 [Epub ahead of print] Abstract
Hasegawa H, Sanjo N, Chen F, Gu Y-J, Shier C, Petit A, Kawarai T, Katayama T, Schmidt S, Mathews P, Schmitt-Ulms G, Fraser PE, St George-Hyslop P. Both the sequence and length of the C-terminus of PEN-2 are critical for intermolecular interactions and function of presenilin complexes. J. Biol. Chem. 2004. Aug 17. Abstract