Transmembrane Substrate Determinants for Γ-Secretase Processing of APP CTFβ

Marty A. Fernandez,Kelly M. Biette,Georgia Dolios,Divya Seth,Rong Wang,Michael S. Wolfe
DOI: https://doi.org/10.1021/acs.biochem.6b00718
IF: 3.321
2016-01-01
Biochemistry
Abstract:The amyloid β-peptide (Aβ) of Alzheimer's disease (AD) is generated by proteolysis within the transmembrane domain (TMD) of a C-terminal fragment of the amyloid β protein-precursor (APP CTFβ) by the γ-secretase complex. This processing produces Aβ ranging from 38 to 49 residues in length. Evidence suggests that this spectrum of Aβ peptides is the result of successive γ-secretase cleavages, with endoproteolysis first occurring at the ε sites to generate Aβ48 or Aβ49, followed by C-terminal trimming mostly every three residues along two product lines to generate shorter, secreted forms of Aβ: the primary Aβ49-46-43-40 line and a minor Aβ48-45-42-38 line. The major secreted Aβ species are Aβ40 and Aβ42, and an increased proportion of the longer, aggregation-prone Aβ42 compared to Aβ40 is widely thought to be important in AD pathogenesis. We examined TMD substrate determinants of the specificity and efficiency of ε site endoproteolysis and carboxypeptidase trimming of CTFβ by γ-secretase. We determined that the C-terminal negative charge of the intermediate Aβ49 does not play a role in its trimming by γ-secretase. Peptidomimetic probes suggest that γ-secretase has S1', S2', and S3' pockets, through which trimming by tripeptides may be determined. However, deletion of residues around the ε sites demonstrates that a depth of three residues within the TMD is not a determinant of the location of endoproteolytic ε cleavage of CTFβ. We also show that instability of the CTFβ TMD helix near the ε site significantly increases endoproteolysis, and that helical instability near the carboxypeptidase cleavage sites facilitates C-terminal trimming by γ-secretase. In addition, we found that CTFβ dimers are not endoproteolyzed by γ-secretase. These results support a model in which initial interaction of the array of residues along the undimerized single helical TMD of substrates dictates the site of initial ε cleavage and that helix unwinding is essential for both endoproteolysis and carboxypeptidase trimming.
What problem does this paper attempt to address?