One major pathophysiological hallmark of Alzheimer's disease (AD) is senile plaques composed of amyloid β (Aβ). In the amyloidogenic pathway, cleavage of the amyloid precursor protein (APP) is shifted towards Aβ production and soluble APPβ (sAPPβ) levels. Aβ is known to impair synaptic function; however, much less is known about the physiological functions of sAPPβ. The neurotrophic properties of sAPPα, derived from the non-amyloidogenic pathway of APP cleavage, are well-established, whereas only a few, conflicting studies on sAPPβ exist. The intracellular pathways of sAPPβ are largely unknown. Since sAPPβ is generated alongside Aβ by β-secretase (BACE1) cleavage, we tested the hypothesis that sAPPβ effects differ from sAPPα effects as a neurotrophic factor. We therefore performed a head-to-head comparison of both mammalian recombinant peptides in developing primary hippocampal neurons (PHN). We found that sAPPα significantly increases axon length (p = 0.0002) and that both sAPPα and sAPPβ increase neurite number (p < 0.0001) of PHN at 7 days in culture (DIV7) but not at DIV4. Moreover, both sAPPα- and sAPPβ-treated neurons showed a higher neuritic complexity in Sholl analysis. The number of glutamatergic synapses (p < 0.0001), as well as layer thickness of postsynaptic densities (PSDs), were significantly increased, and GABAergic synapses decreased upon sAPP overexpression in PHN. Furthermore, we showed that sAPPα enhances ERK and CREB1 phosphorylation upon glutamate stimulation at DIV7, but not DIV4 or DIV14. These neurotrophic effects are further associated with increased glutamate sensitivity and CREB1-signaling. Finally, we found that sAPPα levels are significantly reduced in brain homogenates of AD patients compared to control subjects. Taken together, our data indicate critical stage-dependent roles of sAPPs in the developing glutamatergic system in vitro, which might help to understand deleterious consequences of altered APP shedding in AD patients, beyond Aβ pathophysiology.

In Alzheimer's disease, substantial evidence indicates the causative role of soluble amyloid β (Aβ) aggregates. Although a variety of Aβ assemblies have been described, the debate about their individual relevance is still ongoing. One critical issue hampering this debate is the use of different methods for the characterization of endogenous and synthetic peptide and their intrinsic limitations for distinguishing Aβ aggregates. Here, we used different protocols for the establishment of prefibrillar Aβ assemblies with varying morphologies and sizes and compared them in a head-to-head fashion. Aggregation was characterized via the monomeric peptide over time until spheroidal, protofibrillar, or fibrillar Aβ aggregates were predominant. It could be shown that a change in the ionic environment induced a structural rearrangement, which consequently confounds the delineation of a measured neurotoxicity toward a distinct Aβ assembly. Here, neuronal binding and hippocampal neurotransmission were found to be suitable to account for the synaptotoxicity to different Aβ assemblies, based on the stability of the applied Aβ aggregates in these settings. In contrast to monomeric or fibrillar Aβ, different prefibrillar Aβ aggregates targeted neurons and impaired hippocampal neurotransmission with nanomolar potency, albeit by different modalities. Spheroidal Aβ aggregates inhibited NMDAR-dependent long-term potentiation, as opposed to protofibrillar Aβ aggregates, which inhibited AMPAR-dominated basal neurotransmission. In addition, a provoked structural conversion of spheroidal to protofibrillar Aβ assemblies resulted in a time-dependent suppression of basal neurotransmission, indicative of a mechanistic switch in synaptic impairment. Thus, we emphasize the importance of addressing the metastability of prefacto characterized Aβ aggregates in assigning a biological effect.

Humanin (HN) is a linear 24-aa peptide recently detected in human Alzheimer's disease (AD) brain. HN specifically inhibits neuronal cell death in vitro induced by ß-amyloid (Aß) peptides and by amyloid precursor protein and its gene mutations in familial AD, thereby representing a potential therapeutic lead structure for AD; however, its molecular mechanism of action is not well understood. We report here the identification of the binding epitopes between HN and Aß(1-40) and characterization of the interaction structure through a molecular modeling study. Wild-type HN and HN-sequence mutations were synthesized by SPPS and the HPLC-purified peptides characterized by MALDI-MS. The interaction epitopes between HN and Aß(1-40) were identified by affinity-MS using proteolytic epitope excision and extraction, followed by elution and mass spectrometric characterization of the affinity-bound peptides. The affinity-MS analyses revealed HN(5-15) as the epitope sequence of HN, whereas Aß(17-28) was identified as the Aß interaction epitope. The epitopes and binding sites were ascertained by ELISA of the complex of HN peptides with immobilized Aß(1-40) and by ELISA with Aß(1-40) and Aß-partial sequences as ligands to immobilized HN. The specificity and affinity of the HN-Aß interaction were characterized by direct ESI-MS of the HN-Aß(1-40) complex and by bioaffinity analysis using a surface acoustic wave biosensor, providing a K(D) of the complex of 610 nm. A molecular dynamics simulation of the HN-Aß(1-40) complex was consistent with the binding specificity and shielding effects of the HN and Aß interaction epitopes. These results indicate a specific strong association of HN and Aß(1-40) polypeptide and provide a molecular basis for understanding the neuroprotective function of HN.

Essential for the proteolytic processing of APP by BACE1 - and therewith for the generation of Abeta - is the transport and sorting of both proteins through endosomal and Golgi compartments. The family of Golgi-localized γ-ear-containing ARF-binding (GGA) proteins was shown to have striking functions in cargo sorting in these pathways. Recently, it was shown that GGA1 and GGA3 can interact with BACE1, that they are expressed in neurons and GGA3 is reduced in AD brain tissues. The GGA VHS-domain was found to be the major binding motif necessary for interaction with BACE1. We applied Immunoprecipitation and Fluorescence Lifetime imaging Microscopy (FLIM) with different GGA-domain deletions to analyze VHS-domain independent BACE1-GGA interaction and the binding capacity of the different deletion mutants. We further extended our approach by ELISA assays and Western Blot analysis to measure the influence on APP processing. Co-immunoprecipitation experiments showed that all GGA-VHS deletion-mutants can still be precipitated by BACE1, though to a lower extent than their wild type forms. Additionally, in FLIM experiments the lifetime of GGA-VHS deletion-mutants was only slightly increased compared to wild-type GGAs. Furthermore, cells co-transfected with APP, BACE and either wild-type GGAs or VHS deletion-mutants showed no significant differences in APP processing in Mesoscale Elisa assays and Western Blot analysis. Therefore, we extended our approach to identify other domains responsible for GGA-BACE1 interaction using several GGA domain-deletions and point-mutations to identify the major binding motif necessary for GGA-BACE interaction. In conclusion, our data suggest that the GGA VHS-domain has only limited influence upon the processing of APP and upon interaction with and trafficking of BACE1. As all other known GGA cargo proteins seem to be dependent on the VHS-domain differing BACE1 interactions are of high interest in the attempt to find BACE1 specific transport and trafficking modifiers.

Cleavage of APP by BACE1 is the first proteolytic step in the production of amyloid-beta (Aβ), which accumulates in senile plaques in Alzheimer's disease. Through its interaction with APP, the low-density receptor-related protein 1 (LRP1) enhances APP internalization. Recently, BACE1 has been shown to interact with and cleave the light chain (lc) of LRP1. Since LRP1 is known to compete with APP for cleavage by gamma-secretase, we tested the hypothesis that LRP1 also acts as a competitive substrate for β-secretase. We found that the increase in secreted APP (sAPP) mediated by over-expression of BACE1 in APP-transfected cells could be decreased by simultaneous LRP1 over-expression. Analysis by multi-spot ELISA revealed that this is due to a decrease in sAPPβ, but not sAPPα. Interaction between APP and BACE1, as measured by immunoprecipitation and fluorescence lifetime assays, was impaired by LRP1 over-expression. We also demonstrate that APP over-expression leads to decreased LRP1 association with and cleavage by BACE1. In conclusion, our data suggest that – in addition to its role in APP trafficking – LRP1 affects APP processing by competing for cleavage by BACE1.

Previous studies identified engulfment adapter phosphotyrosine binding (PTB) domain containing 1 (GULP1) as an NPXY-motif interactor of low-density lipoprotein receptor-related protein 1 (LRP1) and suggested a potential relevance in Alzheimer's disease (AD). Since AD associated proteins amyloid-β A4 precursor protein (APP) and LRP1 were shown to interact with the PTB domain of Fe65 and several other adapters via their intracellular NPXY-motifs, we examined a possible interaction of GULP1 PTB domain with the YENPTY-motif of APP. Here we demonstrate that GULP1 is present in human hippocampal and neocortical neurons. Confocal live cell imaging revealed that coexpressed and endogenous GULP1 colocalizes with APP in the Golgi and endoplasmic reticulum. Analysis of the interacting domains by co-immunoprecipitation of point and deletion mutants revealed that the interaction depends on the PTB domain of GULP1 and the YENPTY-motif of APP. Coexpression of GULP1 affected APP cell surface localization and suppressed generation of Aβ40/42 and sAPPα. Taken together, these data identify GULP1 as a novel neuronal APP interacting protein that alters trafficking and processing of APP.

APPs intracellular domain contains a highly conserved YENPTY-motif present in the cytodomains of several receptors that serves as a docking site for phosphotyrosine binding (PTB) domain of adaptor proteins like Fe65, X11/MINT, Numb, JIP1b, and Dab1. Such adaptors play critical roles in tyrosine kinase mediated signal transduction, protein trafficking, phagocytosis, and neuronal development. Most importantly in AD, each of those proteins can impact APP trafficking and processing. Previous studies identified the PTB domain containing engulfment adaptor protein GULP/hCED6 as an LRP/CD91 NPXY-motif interacting protein. Although there is some information on the function of CED-6 during development in C. elegans and drosophila, there is much less known about GULP in mammalian mature brain. Using immunohistochemical staining of human hippocampus slices, immunoprecipitation and confocal live cell imaging (CLSM) approaches we analyzed whether GULP and APP are physiologically co-expressed and whether they co-localize within same cellular compartments. The biophysical interaction was analyzed by immunoprecipitation of overexpressed or endogenous proteins. Interaction sites were determined by point mutation within the YENPTY-motif of APP and deletion of GULPs PTB domain. Input upon APP processing was investigated by ELISA (Aβ40/42) and cell surface protein biotinylation assay. Here we demonstrate that GULP is expressed in mammalian brain and APP and GULP co-localize in vitro and in vivo. Analysis of the interacting domains of both proteins by point and deletion mutants confirmed that biophysical interaction depends on intact PTB domain and YENPTY-motif. Moreover, presence of APP altered GULPs cellular distribution indicating an affinity of the adaptor to APP. Functional assays revealed that GULP overexpression in turn decreases surface levels of APP and generation of Aβ. Together these data identify GULP as a novel APP interacting protein and indicate that GULP alters APP trafficking and may therefore play a role in APP processing by secretases and subsequent Aβ generation.

Abeta generation is caused by sequential cleavage of the amyloid precursor protein (APP) by two proteases, beta-site of APP cleaving-enzyme (BACE) and Gamma-secretase. Complex transport including internalization and return to the surface is necessary for this process. GGA1, a transport adaptor of the Golgi-localized gamma-ear-containing ARF-binding (GGA) family, has been shown to interact with BACE via its VHS-domain. We analyzed the differential roles of GGA1, GGA2 and GGA3 upon co-localization and interaction with BACE and effects upon APP-processing. In addition we tested the hypothesis that serine-phosphorylation of GGA1 and GGA3 affects BACE-interaction. We applied confocal imaging and fluorescence lifetime imaging microscopy (FLIM) for colocalization and interaction studies, an electrochemiluminescence-based assay to measure the influence of all three GGA upon APP processing, westernblotting to measure intracellular levels of APP cleavage fragments and insitu-hybridization to visualize correlated expression of GGAs and BACE in rat brain. All three GGAs colocalize with BACE1 at perinuclear compartments. FLIM revealed a donor-lifetime decrease indicating interaction between all three GGAs and BACE. Control experiments with GGA and BACE mutants show that the VHS domain of the GGA proteins and the DXXLL-motive in the BACE protein are necessary for this interaction. Mutants of GGA1 and GGA3 which represent non-phosphorylated forms of these proteins showed decreased lifetime whereas autoinhibited pseudo-phosphorylated mutants reversed this. Elisa and Westernblotting revealed an increase of intracellular sAPP upon overexpression of any GGA. However a decrease in sAPP secretion was observed. This effect was neither reversible with Δ-VHS mutants nor with GGA1/3 phosphorylation-mutants.Insitu-hybridization revealed spatial and time correlated expression of GGAs and BACE in postnatal and adult rats. These results indicate that all GGAs have related functions on BACE interaction, controlled by concentration, location and phosphorylation. Beside the interaction with BACE, we suggest additional BACE-independent influence of GGAs on APP transport and processing, and therefore an essential role in APP cleavage and subsequent Abeta generation.

Abeta is produced by sequential cleavage of amyloid precursor protein (APP) by beta-site of APP-cleaving enzyme 1 (BACE1) and gamma-secretase. It has been shown that the low density lipoprotein receptor-related protein (LRP) plays an important role in APP trafficking by interacting with APP leading to enhanced APP internalization. Recently the light chain (lc) of LRP has been shown to interact with and to be cleaved by BACE1. We applied a SEAP-reporter assay to measure secreted APP (sAPP) as well as an electrochemiluminescence-based multi-spot ELISA to distinguish between sAPP alpha and sAPP-beta secretion. The intra- and extracelluar cleavage fragments were analyzed by Western Blot. The SEAP reporter-assay showed a moderate increase of sAPP in the medium upon APP overexpression in HEK293 and N2A cells and an additional increase when overexpressed together with BACE1, indicating a sAPPbeta increase. This effect was depleted by adding lcLRP. To verify the cleavage product as sAPPbeta, we applied multi-spot ELISA. This ELISA revealed the expected increase of sAPPalpha after transfection with APP. This increase changed to sAPPbeta after BACE1 coexpression. Additional overexpression of lcLRP led to a decrease of both sAPPalpha as well as sAPPbeta in cell culture supernatant. We found a decrease of sAPPalpha as well as sAPPbeta upon lcLRP coexpression supporting our notion of a competition between lcLRP and APP towards the BACE1 cleavage. Our results suggest a complex role of LRP in APP processing besides influencing APP internalization.

Abeta generation is caused by sequential cleavage of the amyloid precursor protein (APP) by two proteases, first beta-site of APP-cleaving enzyme (BACE) followed by Gamma-secretase. Abeta accumulates in senile plaques in Alzheimer's disease (AD). APP and BACE traffic together. GGA1, a member of the Golgi-localized gamma-ear-containing ARF binding (GGA) protein family interacts with BACE phosphorylation dependent via its VHS domain in the Golgi and influences its subcellular distribution.Therefore we were interested in the question whether GGA2 or GGA3 or both also colocalize and interact with BACE and alter its subcellular occurrence and distribution. We applied confocal imaging and a novel technique to show close protein-protein vicinity in living cells with fluorescence protein tags: spectral fluorescence lifetime imaging microscopy (FLIM). Spectral FLIM is a novel technique, which combines spectral resolved and time resolved detection. Additionally an electrochemoluminescence-based assay was established to measure the influence of GGA1, 2 and 3 upon APP processing. We found colocalization of GGA1, GGA2 and GGA3 with BACE1 at perinuclear compartments. The colocalization of GGA2 and BACE1 was enhanced in endosomal as well as Golgi structures whereas the colocalizations of GGA3 and BACE1 were smaller and could only be shown at the ER or Golgi. Regarding the GGAs and BACE FLIM we found a clear decrease in donor lifetime indicating interaction between all three GGAs and BACE. By performing control experiments with deletion mutants of GGAs and BACE we were able to show that the VHS domain of the GGA proteins and the DXXLL-motive in the BACE protein are responsible for the interaction. We observed reduced sAPP secretion upon GGA overexpression. This effect seems not to be reversible with VHS-deletion mutants of the GGAs. Therefore we suggest a BACE-independent influence of GGA1, 2 and 3 upon the APP transport and processing. These results indicate that all GGAs may have differential impact upon the transport of BACE1 and APP and may therefore play an essential role in APP cleavage and the subsequent Abeta generation.

Cleavage of APP by BACE is the first proteolytic step in the production of Amyloid β (Aβ, which accumulates in senile plaques in Alzheimer’s disease. BACE-cleavage of APP is thought to happen in endosomes. However, there are controversial data whether APP and BACE can already interact on the cell surface dependent on the cholesterol level. To examine whether APP and BACE come into close proximity on the cell surface in living cells, we employed a novel technique by combining time-resolved Förster resonance energy transfer (FRET) measurements with total internal reflection microscopy (TIRET microscopy). Our data indicate that BACE and APP come into close proximity within the cell, but probably not on the cell surface. To analyze the impact of alterations in cholesterol level upon BACE-cleavage, we measured sAPP secretion. Alteration of APP processing and BACE proximity by cholesterol might be explained by alterations in cell membrane fluidity.

Abeta generation is caused by sequential cleavage of the amyloid precursor protein (APP). Beta-site of APP-cleaving enzyme (BACE) is known to cleave APP followed by Gamma-secretase. The resulting Abeta accumulates in extracellular senile plaques and is associated with further neuron degeneration. APP and BACE traffic together. GGA1, a member of the Golgi-localized gamma-ear-containing ARF binding (GGA) protein family interacts with BACE phosphorylation dependent via its VHS domain in the Golgi and influences its subcellular distribution.

BACE-cleavage of APP is the first proteolytic step in the production of Amyloid beta, which accumulates in senile plaques in Alzheimer's disease. There is a large body of evidence that this happens in endosomes. However there is controversial data whether APP and BACE come into close proximity and interact on the cell surface. To examine whether APP and BACE come into close proximity on the cell surface and /or within the cell in living cells we developed a novel technique by combining FRET-measurements by lifetime imaging with total internal reflection microscopy (TIRET) offering excellent axial resolution. Membrane fluidity was observed by fluorescence microscopy. To assess functional data upon APP cleavage we employed a novel electrochemiluminescent based assay to detect sAPPα and sAPPβ simultaneously. Our data show that BACE and APP are both present on the cell surface but interact only within the cell and not on the cell surface. This interaction is dependent upon cholesterol level. Membrane fluidity is also dependent upon cholesterol level. To analyze the impact of alterations in cholesterol level upon BACE-cleavage we measured sAPP secretion. sAPP beta secretion increased upon cholesterol enrichment and decreased upon depletion. sAPPα increased upon cholesterol depletion and decreased upon cholesterol enrichment. Thus the FRET approach developed here utilizing FLIM under TIRF conditions (TIRET) may be broadly applicable to the issue of identifying the localization of specific protein-protein proximity on the cell surface with exquisite spatial resolution, since membrane selective optical excitation is combined with ultra-sensitive and time-resolving fluorescence detection. Alteration of APP processing and BACE proximity by cholesterol might be explained by alterations in cell membrane fluidity.

Accumulation of amyloid-beta-peptide (Abeta) in senile plaques is a characteristic feature for Alzheimer's disease. The amyloid precursor protein (APP) is cleaved sequentially by beta-site APP-cleaving enzyme (BACE) and gamma-secretase to release Abeta. Recent data indicate that production, clearance, and aggregation of Aß are highly dependent on the specific neuronal subcellular localization wherein Aß is generated. We tested the hypothesis that APP and BACE are co-localized within lipid domains (e.g. lipid rafts) on the cell surface.