The major component of amyloid plaque cores and cerebrovascular amylold deposits found in Alzheimer disease Is the P/A4 peptide, which is derived from the AIzheimer amylold protein precursor (APP). Recent evidence sugests that abnormalities in f/A4 peptide production or ,B/A4 peptide agregation may underlie cerebral amyloidosis.In the present study, treatment of cells with phorbol dibutyrate, which activates protein kinase C, and/or okadaic acid, which inhibits protein phosphatases 1 and 2A, reduced P/A4 peptide production by 50-80%. These effects were observed with APP6. and APP75s expressed in stablY btrasfected CHO cells, as well as with endogenous APP in human glioma (Hs 683) cells. Phorbol dibutyrate also decreased P/A4 peptide production in cells expressing various mutant forms of APP associated with familial Alzheimer disease, one of which was reported to m est greatly incrased P/A4 peptide production in cultured cells. Mastoparan and mastoparan X, compounds which can activate phospholipase C and hence protein khin C, also decreased f/A4 peptide production in CHO cells stably transfected with APP695. A model is presented in which decreases in 3/A4 peptide production can be achieved by accelerating the metabolism of APP through a nonamyloidgenic secretory pathway. Pulse-chase labeling of cells was carried out on confluent cell monolayers in six-well culture dishes (Corning) with 1 ml of methionine-free Dulbecco's modified Eagle's medium (DMEM) supplemented with 1 mCi (37 MBq) of [35S]methionine (EXPRE35S35S, NEN). Metabolic labeling was carried out for 2 hr, followed by a chase period of2 hr. The chase was initiated by replacing the labeling medium with DMEM containing excess unlabeled methionine. Two minutes after the start of the chase, the indicated test compounds were added. This protocol maximized the probability that any observed effects were attributable to changes in APP metabolism rather than APP transcription. After incubation, conditioned medium was centrifuged for 5 min at 10,000 x g to remove cellular debris. 3/A4 peptide and p3 were immunoprecipitated with an anti-p/A4 antibody (4G8 except where indicated), and secreted APP (APPs) was immunoprecipitated with an anti-APP NH2-terminal antibody (22C11). Immunoprecipitated APP fragments were subjected to SDS/ PAGE in either 10-20% gels with Tris/tricine buffer (for f/A4 peptide and p3) or 6% gels with Tris/glycine buffer (for APPs), autoradiographed, and quantified with a PhosphorImager (Molecular Dynamics). Analysis of variance followed by Fisher's post-hoc analysis was used to determine the significance of observed differences.
RESULTSWhen the medium of metabolically labeled CHO cells stably transfected with cDNA encoding the 695-amino-acid isoform of APP (APP695) was subjected to immunoprecipitation with an antibody raised against a synthetic peptide corresponding to ,/A4 peptide, two small peptide bands of apparent molecular mass 3 and 2 kDa were observed (Fig. 1). These peptides had migration rates similar to those of two peptides...