O2‐12‐06: GLYCOSYLATION REGULATES DEGRADATION OF BACE1 IN LYSOSOME

contrast, in AD patients a substantial amount of deposited Abeta is N-terminally truncated and pyroglutamate-modified (Abeta pGlu3-x). In order to study the origin of such truncated peptides, we analyzed proteolytic pathways leading to their generation in mammalian cells. Methods: In a cellbased study, vectors containing APP (wt) or APP bearing the KM595/ 596NL "Swedish" mutation (APP (sw)) were transiently expressed in different permanent cell lines including immortalized fibroblasts generated from BACE I/II knockout mice (courtesy of Prof. B. de Strooper, KU Leuven, Belgium). Furthermore, Abeta formation in primary neurons from APP (wt) and APP(sw)-overexpressing mice (lines: hAPP-wt I5 and Tg2576) was investigated. Analytical methods comprised highly specific ELISAs discriminating between different N-terminal Abeta variants, siRNA andWestern Blot experiments.Results:As expected, full-length Abeta peptides were generated via BACE I-dependent cleavage and inhibition of BACE I led to a dramatic reduction of full-length Abeta. However, these full-length Abeta peptides represent only a fraction of the total Abeta pool produced by the applied cell lines and primary neurons. Expression of an APP vector containing a monitoring mutation (E599Q) indicative for the formation of Abeta pGlu3-x revealed that the fraction of alternative Abeta N-termini contains substantial amounts of Abeta pGlu3-x. Interestingly, the generation of these truncated pGlu3-x peptides was found to originate from BACE I-independent beta-secretase cleavage of APP. Conclusions: Due to their high amyloidogeneity and toxicity, N-terminally truncated and pGlu-3modified Abeta peptides represent potential targets for AD drug development. However, BACE I inhibition is not sufficient to block the generation of pGlu-3 Abeta peptides. Our data qualify alternative proteases as novel targets for drug development in AD.