EnglishCookie löschen - von nun an wird die Spracheinstellung Ihres Browsers verwendet.
Münch, Anna (2012): Zelluläre Modulatoren der Ektodomänenspaltung von APP. Dissertation, LMU München: Medizinische Fakultät



The processing of APP occurs in two alternative ways: upon release of the ectodomain by α-secretase, the neuroprotective APPsα-fragment is produced. But if APP is cleaved by the β-secretase the Aβ-peptide can be produced. To be able to influence the production of Aβ-peptides, it is essential to understand how it is decided if cleavage occurs by α- or β-secretase. At present little is known about the control of the alternate processing. Until now, the molecular mechanisms and especially the responsible cellular modulators are not understood in detail or not yet identified. To get a better understanding of cellular regulatory processes and to identify novel cellular modulators of APP ectodomain shedding, the present work chose two approaches: on the one hand cellular mechanisms of TMEM59-mediated inhibition ectodomain shedding of APP were investigated. On the other hand a genome-wide RNAi screening in Drosophila cells was performed in order to identify novel cellular modulators of APP ectodomain shedding in human cells. TMEM59 was identified as a novel modulator of APP ectodomain shedding in a cDNA expression screening in the lab (Neumann et al., 2006; Schobel et al., 2008; Schobel et al., 2006). TMEM59 is a Golgi protein that inhibits on the one hand processing and maturation of APP and on the other hand Golgi glycosylation reactions (Fischer, 2008). My own work could verify these effects of TMEM59 and its homolog TMEM59L on processing and maturation of APP. In particular, it was shown that these effects are not only true for transiently expressed APP but also for endogenous levels of APP. In detailed immunofluorescence studies it was shown that TMEM59 colocalizes with different markers of the Golgi subcompartments and that therefore TMEM59 is present throughout the whole Golgi apparatus. This finding points to a more general modulation of Golgi glycosylation reactions by TMEM59. To test if TMEM59-dependet modulation of Golgi glycosylation reactions also affects APP secretases ADAM10 and BACE1, which are also glycosylated proteins, the activities of these proteases were investigated. It was shown that proteolytic activities were not changed, ruling out that impairment of secretase activities by TMEM59 could cause the observed inhibition of APP processing. But interestingly, studies of intracellular APP transport could show that TMEM59 caused retention of APP in the Golgi apparatus and blockage of transport towards the cell surface and into endosomal compartments. Since APP is cleaved by α-secretase at the plasma membrane and by β-secretase in endosomes it is likely that a TMEM59-dependent APP transport block causes the observed inhibition of APP ectodomain shedding. For further validation of TMEM59 and its homolog TMEM59L as modulators of APP ectodomain shedding, a double knockdown study was performed. In this approach effects on APP ectodomain shedding could also be established, affirming TMEM59 and its homolog TMEM59L as modulators of APP ectodomain shedding with novel cellular mechanisms. In order to identify novel cellular modulators of APP ectodomain shedding a genome wide RNAi screening in Drosophila cells was performed and candidate genes were investigated in human cells in present work. Initially a suitable Drosophila reporter cell line expressing a reporter construct of APP ectodomain shedding (HRP-APP) was established. Other constructs were used to monitor general secretion (GLuc) and transfection efficiency (FLuc). Using Kuzbanian, the α-secretase in Drosophila (Sapir et al., 2005), as a positive control guaranteed that transfection of cDNAs into Drosophila cells did not interfere with uptake of dsRNAs or efficiency of RNAi and that the reporter construct HRP-APP is normally produced and processed in reporter cells. After successful establishment of the reporter cell line the genome wide RNAi was performed in two steps: a primary screening revealed approx. 300 candidate genes out of which 43 could be confirmed in a secondary screening to be modulators of APP ectodomain shedding. The RNAi screening was verified by the several-fold appearance of Kuzbanian among the top modulators. For further investigation of the top candidates human ortholog genes were identified. The 30 human candidate genes were investigated in RNAi studies in human SH-SY5Y cells. In these cells, APP is processed by α-secretase ADAM10 as well as by β-secretase BACE1. Therefore effects on both shedding products (APPsα and APPsβ) were investigated upon depletion of candidate genes using siRNAs. It is known that siRNAs produce a high rate of off target effects, to this end a robust validation strategy was developed. Candidate genes were first depleted with two different siRNA pools and their effects on APP shedding were compared. Afterwards the remaining 12 candidate genes were depleted using single siRNA sequences and the effects were compared to those of the siRNA pool. Only when a reproduction of effects was obtained in a next step correlation of knockdown and phenotype were assessed. Using these steps of validation 5 candidate genes could be verified as modulators of APP shedding in human cells: next to genes coding for a histone protein (HIST1H4C), a ribosomal protein (RPL36AL), a protein of the minor spliceosom (ZMAT5), an unknown gene (METTL16) and the gene VPS24 („vacuolar protein sorting-associated protein 24“), coding for a protein of intracellular protein transport, were identified. VPS24 was chosen for further validation by a pathway analysis. VPS24 belongs to the ESCRT machinery („endosomal sorting complex required for transport“) and therefore participates in endosomal-lysosomal protein transport. In further RNAi studies other members of the ESCRT machinery were depleted in human cells and effects on APP shedding were compared to VPS24 depletion. For most of the ESCRT members a consistent reduction in APPsβ production could be observed. To engross these results VPS24 was depleted by using an alternative RNAi system. With this stable knockdown approach, the knockdown phenotype could be confirmed. This stepwise validation strategy for candidate genes of the initial Drosophila RNAi screening verified VPS24 as a modulator of APP ectodomain shedding in human cells.