Α-Secretase Processing of the Alzheimer Amyloid-Β Precursor Protein and Its Homolog APLP2

Home , Alpha secretase, APLP2, Metalloendopeptidase, Sheddase

α-Secretase processing of the Alzheimer amyloid-β precursor protein and its homolog APLP2

Kristin Jacobsen

1 Doctoral dissertation, 2013 Department of Neurochemistry Arrhenius Laboratories for Natural Sciences Stockholm University

Cover: Schematic illustration of APP and APLP2 processing during stimulated conditions

©Kristin T. Jacobsen, Stockholm University 2013

ISBN 978-91-7447-732-0

Printed in Sweden by Universitetsservice US-AB, Stockholm 2013 Distributor: Department of Neurochemistry, Stockholm University

2 To my family

4 List of Publications

I. Jacobsen KT, Adlerz L, Multhaup G and Iverfeldt K. IGF-1-induced processing of amyloid-β precursor protein and APP-like protein 2 is mediated by different enzymes Journal of Biological Chemistry. 2010; 285 (14): 10223-10231

II. Holback S, Adlerz L, Gatsinzi T, Jacobsen KT and Iverfeldt K PI3-K- and PKC-dependent up-regulation of APP processing enzymes by retinoic acid Biochemical and Biophysical Research Communications. 2008; 365(2): 298-303

III. Jacobsen KT and Iverfeldt K. O-GlcNAcylation increases non-amyloidogenic processing of the amyloid-β precursor protein (APP) Biochemical and Biophysical Research Communications. 2011; 404(3): 882-886

IV. Jacobsen KT, Strååt Y, Koistinen N and Iverfeldt K. O-GlcNAcylation of the α-secretase ADAM10 selectively affects APP processing in neuron-like cells Manuscript

V. Jacobsen KT and Iverfeldt K The E1 domain of APP and APLP2 determines α-secretase specificity Manuscript

5 Additional publications

. Jacobsen KT and Iverfeldt K Amyloid precursor protein and its homologues: a family of proteolysis-dependent receptors Cellular and Molecular Life Science. 2009; 66: 2299-2318 (Review)

. Tracy, LM, Bergqvist F, Ivanova EV, Jacobsen KT and Iverfeldt K Exposure to the saturated free fatty acid palmitate alters BV-2 microglia inflammatory response Journal of Molecular Neuroscience. 2013; 51(3): 805-812

6 Abstract

The amyloid-β precursor protein (APP) has been widely studied due to its role in Alzheimer´s disease (AD). When APP is sequentially cleaved by β- and γ-secretase, amyloid-β (Aβ) is formed. Aβ is prone to aggregate and is toxic to neurons. However, the main processing pathway for APP involves initial cleavage at the α-site, within the Aβ region, instead generating a neuroprotective soluble fragment, sAPPα. APP is a member of a protein family, also including the proteins APLP1 and APLP2, which are processed in a similar way as APP. In addition, knock-out studies in mice have shown that the three proteins have overlapping functions where APLP2 play a key physiological role. The aim of this thesis was to study mechanisms regulating the α-secretase processing of APP and APLP2. We have used the human neuroblastoma cell line SH-SY5Y as a model system and have stimulated α- secretase processing with insulin-like growth factor-1 (IGF-1) or retinoic acid (RA). Our results show that the stimulated α-site cleavage of APP and APLP2 is regulated by different signaling pathways and that the cleavage is mediated by different enzymes. APP was shown to be cleaved by ADAM10 in a PI3K-dependent manner, whereas APLP2 was cleaved by TACE in a PKC-dependent manner. We further show that protein levels and maturation of ADAM10 and TACE is increased in response to RA, mediated by a PI3K- or PKC-dependent signaling pathway, respectively. Another focus of our research has been O-GlcNAcylation, a dynamic post-translational modification regulated by the enzymes O-GlcNAc transferase and O-GlcNAcase (OGA). We show that decreased OGA activity stimulates sAPPα secretion, without affecting APLP2 processing. We further show that ADAM10 is O- GlcNAcylated. Lastly, we show that APP can be manipulated to be cleaved in a similar way as APLP2 during IGF-1 stimulation by substituting the E1 domain in APP with the E1 domain from APLP2. Together our results show distinct α-site processing mechanisms of APP and APLP2.

7 Contents

1. Introduction ...... 13 1.1 The role of APP in Alzheimer’s disease ...... 13 1.2 The APP family ...... 14 1.2.1 Structure ...... 14 1.2.2 Proteolytic processing ...... 17 1.2.3 Biological functions ...... 19 1.3 Processing enzymes ...... 21 1.3.1 α-Secretase ...... 21 1.3.1.1 Substrate selectivity of TACE and ADAM10 ...... 23 1.3.2 β-Secretase ...... 24 1.3.3 γ-Secretase...... 25 1.4 Signaling pathways affecting APP family processing ...... 26 1.4.1 IGF-1 ...... 26 1.4.2 Retinoic acid ...... 27 1.5 Regulation of APP family α-site processing ...... 28 1.5.1 Trafficking of enzyme and substrate ...... 28 1.5.2 Post-translational modifications of enzyme and substrate ...... 30 1.5.2.1 Post-translational modifications of ADAM10 and TACE...... 31 1.5.2.2 Post-translational modifications of the APP family ...... 32 2. Methodological considerations ...... 34 2.1 Cell lines ...... 34 2.2 Cell treatments ...... 35 2.2.1 Retinoic acid and IGF-1 ...... 35 2.2.2 Pharmacological inhibitors ...... 35 2.3 siRNA gene silencing ...... 36 2.4 BCA assay ...... 36 2.5 Western blot ...... 37 2.6 ELISA ...... 37 2.7 32P-labeling and immunoprecipitation ...... 38 2.8 sWGA precipitation ...... 39 2.9 Biotinylation assay ...... 39 2.10 Design and cloning of the APP/APLP2 chimer ...... 39 3. Aims ...... 41 4. Results and discussion ...... 42

8 4.1 Stimulated α-secretase processing of APP is mediated by ADAM10 in a PI3K-dependent manner (Paper I and II) ...... 42 4.2 Stimulated α-secretase processing of APLP2 is mediated by TACE in a PKC-dependent manner (Paper I and II) ...... 44 4.3 O-GlcNAcylation induces α-secretase processing of APP but not of APLP2 (Paper III and IV) ...... 45 4.3.1 O-GlcNAcylation selectively enhances α-secretase processing of APP in neuron-like cells (Paper IV) ...... 46 4.3.2 ADAM10 is O-GlcNAcylated (Paper IV) ...... 47 4.4 The difference in regulation of α-secretase processing between APP and APLP2 is determined by their E1 domain (Paper III) ...... 48 4.5 What about APLP1? (Paper I and IV) ...... 50 5. Conclusions ...... 51 6. Populärvetenskaplig sammanfattning på svenska ...... 53 7. Acknowledgements ...... 55 8. References ...... 57

9 Abbreviations

Aβ Amyloid-β AD Alzheimer’s Disease ADAM A disintegrin and metalloprotease AICD APP intracellular domain ALID APP-like intracellular domain ANOVA Analysis of variance APH-1 Anterior pharynx defective-1 APLP APP-like protein APP Amyloid-β precursor protein BACE β-site cleaving enzyme BCA Bicinchoninic acid BDNF Brain-derived neurotrophic factor C83 APP C-terminal stub of 83 amino acids C99 APP C-terminal stub of 99 amino acids CNS Central nervous system CRABs Cellular retinoic acid binding proteins CSF Cerebrospinal fluid CS GAG Chondroitin sulphate glycosaminoglycan CTF C-terminal fragment CuBD Copper binding domain DAG Diacylglycerol DTT Dithiothreitol ECL Enhanced chemiluminescence ELISA Enzyme-linked immunosorbent assay ER Endoplasmatic reticulum ERK Extracellular signal-regulated kinase GFLD Growth factor-like domain HBD Heparin binding domain HRP Horseradish peroxidase IGF-1 Insulin-like growth factor IGF-1R IGF-1 receptor IP Immunoprecipitation IP3 Inositol 1,4,5-triphosphate IRS Insulin receptor substrate KI Knock-in K/O Knock-out

10 KPI Kunitz protease inhibitor LRP Low-density lipoprotein receptor-related protein LTP Long term potentiation MAPK Mitogen activated protein kinase NMDA N-methyl D-aspartate NMR Nuclear magnetic resonance OGA O-GlcNAcase O-GlcNAc O-linked β-N-acetylglucosamine OGT O-GlcNAc transferase PC Pro-hormone convertase PEN2 Presenilin enhancer 2 PI3K Phosphatidyl inositol-3 kinase PIP2 Phosphatidylinositol 4,5-bisphosphate PKC Protein kinase C PLC Phospholipase C PNS Peripheral nervous system PS Presenilin RA Retinoic acid RAR Retinoic acid receptor RARE Retinoic acid responsive element RBP Retinol binding protein RIP Regulated intramembrane proteolysis RISC RNA-induced silencing complex RNA Ribonucleic acid RT-PCR Reverse-transcriptase polymerase chain reaction RXR Retinoid X receptor sAPP Secreted APP SAP97 Synapse associated protein 97 SDS Sodium dodecyl sulphate Shc Src homology collagen siRNA Short interfering RNA SorLA Sorting-related receptor with A-type repeats TGF Transforming growth factor TGN Trans Golgi network TIMPs Tissue inhibitors of metalloproteases TM Transmembrane TMB Tetramethylbenzidine TNF Tumor necrosis factor Tspan Tetraspanin Wt Wild type ZnBD Zinc binding domain

12 1. Introduction

Proteolytic processing of the amyloid-β precursor protein (APP) has been extensively studied due to the link between the APP cleavage product, amyloid-β (Aβ) and Alzheimer’s disease (AD). However, this thesis is focused on mechanisms involved in APP α-secretase processing, i.e., the processing pathway that excludes Aβ formation. There are clear evidences that APP together with its homolog APP-like protein-2 (APLP2) have important functions in the brain. In our studies we aimed to elucidate how α-secretase processing of APP differs from the processing of its homolog APP-like protein- 2 (APLP2). α-Secretases are also involved in regulated intramembrane proteolysis (RIP), an important novel form of signal transduction. In addition, it has lately been shown that dysregulation of the α-secretase pathway may cause AD, and that enhancing this cleavage of APP constitutes a potential therapeutic strategy.

1.1 The role of APP in Alzheimer’s disease APP was, as the name suggests, identified as being the precursor of the Aβ peptide (Kang et al., 1987). Aβ is a secreted peptide generated through sequential cleavage of APP and its accumulation in the brain is today considered the main cause of the widespread neurodegeneration seen in AD (Hardy, 2009, Hardy and Higgins, 1992), although other mechanisms contribute. The neurodegeneration in turn leads to clinical manifestations such as problem with speech, disorientation and progressive loss of memory and cognitive function (Querfurth and LaFerla, 2010).

Aβ is highly hydrophobic and aggregates into several different confor- mations, including low molecular weight species such as dimers and trimers as well as oligomers and larger aggregates such as fibrils and plaques (reviewed in Finder and Glockshuber, 2007). The presence of plaques in AD post-mortem brain was described already in 1907 by the German physician Alois Alzheimer and was initially believed to be the inducer of cell death. However, later studies have demonstrated that it rather is Aβ oligomers that mediate the main neurotoxicity (Klyubin et al., 2005, Shankar et al., 2007, Walsh et al., 2002). The exact mechanism of Aβ neurotoxicity is not yet

13 established, but several studies indicate that Aβ may 1) bind to various cell surface receptors causing aberrant signaling 2) interact with the membrane and be inserted as pores causing calcium influx or 3) internalize and cause many cellular dysfunctions through direct interactions, including mitochon- drial dysfunction (reviewed in Kayed and Lasagna-Reeves, 2013). In addition to its direct harmful effects, Aβ may also induce hyperphosphorylation of the microtubule-associated protein tau, causing it to aggregate intracellu- larly into neurotoxic tangles (Chabrier et al., 2012). Both Aβ and tau aggregates may cause chronic activation of microglia, the immune cells of the brain. This can result in increased production of inflammatory agents in the brain amplifying the degree of neurodegeneration. This line of events trig- gered by Aβ is called the amyloid cascade hypothesis (Hardy and Higgins, 1992). It is supported by the fact that several mutations in APP, and in an enzyme component that cleaves APP to release Aβ, causes familial forms of AD (reviewed in Duyckaerts et al., 2009).

1.2 The APP family Soon after its discovery, APP was found to be evolutionary highly conserved. In addition, two mammalian genes encoding the homologous proteins, APP-like protein-1 and -2 (APLP1 and APLP2) were identified (Wasco et al., 1992, Wasco et al., 1993). APP shares 42% amino acid sequence homology with APLP1 (64% similar) and 52% homology with APLP2 (71% similar). However, only APP generates Aβ. Homologues in other species also include the APP-like proteins APPL (Rosen et al., 1989) and APL-1 (Daigle and Li, 1993) in fruit fly (Drosophila melanogaster) and nematode (Caenorhabditis elegans), respectively. Here, the structure, function and the proteolytically processing of the mammalian APP family proteins will be discussed.

1.2.1 Structure There are three major isoforms of mammalian APP; APP695, APP751 and APP770. Alternative splicing of APLP2 also produces multiple protein isoforms, the major ones being 763 and 707 amino acids long, while only one form of APLP1 has been detected. The main difference between the APP isoforms is the presence or absence of the Kunitz protease inhibitor (KPI) domain and a chondroitin sulfate glycosaminoglycan (CS GAG) attachment site. This is also the case for the APLP2 isoforms. APLP1 consists of 650 amino acids and lacks both the KPI domain and the CS GAG attachment site (Paliga et al., 1997, Wasco, et al., 1992). APP and APLP2 are ubiquitously expressed. However, the different isoforms can be preferentially expressed in different cell types, such as the 695 amino acid long isoform

14 of APP, which is mainly found in cells of neuronal origin. APLP1 expression has been reported to be restricted to the nervous system (Lorent et al., 1995, Slunt et al., 1994).

All APP family members are type 1 integral membrane proteins, with a single membrane-spanning domain, a large ectoplasmic N-terminal region and a shorter cytoplasmic C-terminal region (c.f., (Kang, et al., 1987, Dyrks et al., 1988). The APP sequence can be divided into multiple distinct domains (Fig. 1). The ectoplasmic region of APP, which constitutes the major part of the protein, can be divided into the E1 and E2 domains (reviewed in Gralle and Ferreira, 2007). The E1 domain can be further divided into a number of subdomains, including a heparin-binding/growth-factor-like domain (HBD/GFLD) and a copper binding domain (CuBD). The E1 domain is followed by an acidic region rich in aspartic acid and glutamic acid and a KPI domain (not present in APP695). The cytoplasmic region of APP contains a protein interaction motif, namely the YENPTY sequence (including the NPXY internalization signal), which is conserved in all APP homologues. The sequences of APLP1 and APLP2 can be divided into similar domain structures as APP.

Figure 1. Schematic illustration of the domain structure of APP. The structure can be divided into distinct domains, including the ectoplasmic E1 and E2 domains. The E1 domain can be further divided into a heparin- binding/growth factor-like domain (HBD/GFLD) and a cupper-binding domain (CuBD). The E1 domain is linked to the E2 domain via an acidic region, and in some APP isoforms also a Kunitz protease inhibitor domain (KPI). APP also has a cytoplasmic/intracellular domain (ICD), which contains the YENPTY protein interaction motif.

15 There is no complete crystal structure of APP, but the structure of several domains has been studied separately. The major part of the protein is as previously mentioned comprised of the ectoplasmic domain, which has several independent folding structures. The most N-terminal part of APP, the HBD/GFLD has been described to 1.8Å (Rossjohn et al., 1999), showing that this part forms one α-helix and nine β-sheets that are mostly comprised of very short strands of amino acids. Three disulfide bridges bring the domain into a compact, globular shape creating a positively charged surface and a hydrophobic region. The basic surface was suggested to interact with glycosaminoglycans and the hydrophobic surface to be important for ligand binding. The disulfide bridge between Cys98 and Cys105 (in the APP695 isoform) creates a loop which contains several positively charged amino acids. The corresponding loop section in APLP2 also contains positively charged residues, whereas in APLP1 these mainly are uncharged.

The CuBD has been shown to contain one α-helix and three β-sheets using nuclear magnetic resonance (NMR) (Barnham et al., 2003). The CuBD also contains three disulfide bridges dividing the domain into one positive and one negative region. There is also a small hydrophobic region within the CuBD. Recently, the entire E1 structure was solved at 2.7Å resolution, showing that the GFLD and CuBD interacted with each other forming one functional unit stabilized by a salt bridge and several hydrogen bond net- works (Dahms et al., 2010). The KPI domain, which is not present in all APP isoforms, contains approximately 60 residues with six cysteines ar- ranged in three disulfide bonds with little secondary structure. The E2 domain has been described to 2.0Å (Lee et al., 2011, Wang and Ha, 2004). This part of the protein has the most complex secondary structure including six α-helices, which form two coiled coil antiparallel substructures (e.g. parts are folded over each other with the N-terminal of one helix aligned with the C-terminal of the other helix). Recently, a metal binding site was discovered within the E2 domain of APP containing 4 highly conserved histidines (His314, His382, His432 and His436). The site was shown to bind copper and zinc thereby inducing large conformational changes of the E2 domain (Lee, et al., 2011).

Several studies indicate that the APP family proteins are able to form homo- and heterodimers in both cis (within the same cell) and trans (between two adjacent cells), thereby affecting their function and processing (Eggert et al., 2009, Kaden et al., 2008, Soba et al., 2005). Three regions have been implicated in dimerization, namely the E1, the E2 and the transmembrane domains. Co-immunoprecipitation studies with overexpressed APP/APLPs constructs where ectoplasmic domains were deleted showed that the E1 domain was the main interaction interface for dimerization (Soba, et al., 2005). In another study, deletion of the E1 domain in APP or APLP2 was shown to

16 diminish homo-dimerization completely whereas the deletion of the E1 domain in APLP1 only reduced its homo-dimerization by 50% (Kaden et al., 2009). It has further been shown that heparin is required for APP E1 homo- dimerization and that this complex formation was initiated by about 10 sugar rings in heparin that span the positively charged surface area made up by two GFLDs bridging them together (Dahms, et al., 2010). Other studies indicate that also the E2 domain can dimerize. X-ray spectroscopy revealed an antiparallel orientation of the two dimeric E2 domains (Wang and Ha, 2004). The same mode of E2 dimerization has also been shown for APLP1 (Lee, et al., 2011, Xue et al., 2011). As for E1 dimerization, E2 dimerization has also been shown to be induced by heparin (Lee, et al., 2011). The transmembrane (TM) domain was shown by NMR to be a flexible curved α-helix (Barrett et al., 2012). A GxxxG motif is found within the TM domain of APP. This motif is known to mediate TM helix-helix interactions within the membrane (reviewed in Russ and Engelman, 2000), which has been shown to occur also for APP (Munter et al., 2007). Notably, there is no GxxxG motif in APLP-1 or -2.

1.2.2 Proteolytic processing The proteolytic processing of APP can be divided into two different pathways, the amyloidogenic pathway, which leads to generation of Aβ, and the non-amyloidogenic pathway (Fig. 2). Both pathways include several cleavage events. The amyloidogenic processing of APP is initiated through cleavage by β-secretase, which leads to secretion of the large N-terminal ectodomain, sAPPβ. The remaining C-terminal fragment (CTF) of 99 amino acids (C99) can then be further processed by γ-secretase, generating Aβ and a soluble APP intracellular domain (AICD). The γ-secretase processing of C99 occurs at multiple sites. APP is first cleaved at the ε-site at the Thr645- Leu646 bond or at the Leu646-Val647 bond, generating soluble AICD50 or AICD49, respectively (Takami et al., 2009, Weidemann et al., 2002, Yu et al., 2001). Next, the remaining membrane-bound fragment is cleaved stepwise at every third residue from the ε-site towards the γ-site (Qi-Takahara et al., 2005, Takami, et al., 2009), generating the Aβ49Aβ46Aβ43Aβ40 or Aβ48Aβ45Aβ42Aβ38 product lines.

In the non-amyloidogenic pathway, generation of Aβ is precluded since APP is initially cleaved by α-secretase within the Aβ sequence near the ectoplasmic side of the plasma membrane (Esch et al., 1990, Sisodia et al., 1990). The α-secretase cleavage of APP releases the N-terminal ectodomain, sAPPα from the cell surface leaving an 83 amino acid long C-terminal membrane- bound fragment (C83). The remaining C83 fragment can be further processed by γ-secretase in a similar way as in the amyloidogenic pathway, giving rise to the small peptide p3 and AICD.

Figure 2. Schematic illustration of APP processing. In the non-amyloidogenic pathway, APP is initially cleaved by α-secretase, generating the secreted sAPPα fragment. In the amyloidogenic pathway APP is instead cleaved by β-secretase, generating sAPPβ. The remaining C-terminal membrane-bound stubs, C83 and C99, are subsequently cleaved by γ-secretase, releasing the APP intracellular domain (AICD) and p3 or Aβ. γ-Secretase cleaves APP first at the ε-site and then stepwise at every third residue (indicated by arrowheads) towards the γ-site.

The amyloidogenic and the non-amyloidogenic pathways have been proposed to compete with each other, since enhanced α-secretase activity in animal models of AD or in cultured cells can significantly lower Aβ generation (Nitsch et al., 1992, Postina et al., 2004). In addition to the two major processing pathways, APP has been shown to undergo N-terminal processing, generating several soluble fragments varying in size between 12 to 30 kDa (Vella and Cappai, 2012). This processing was regulated by protein kinase C (PKC) in an α- and β-secretase-independent manner. APP may also be cleaved by caspases (6 or 8) in the C-terminal at Asp664, generating a cytotoxic C31 fragment (Gervais et al., 1999, Lu et al., 2000).

18 Although the amino acid sequences in APP where the secretases cleave are very different in the two mammalian homologs, both APLP1 and APLP2 are believed to be proteolytically processed in a similar way as APP. The first indication that APLPs were cleaved by γ-secretase was the detection of ele- vated levels of an APLP1 membrane-bound C-terminal fragment in brain tissues from animals deficient in the γ-secretase component, presenilin-1 (Naruse et al., 1998). Cleavage of APLPs by γ-secretase was further shown to generate fragments corresponding to AICDs, denoted ALID1 and 2 (APP- like intracellular domain 1 and 2) (Gu et al., 2001). Furthermore, β-secretase processing of the APLPs was demonstrated in a study by Li et al. where co- expression of the identified β-secretase BACE1 (β-site APP cleaving enzyme 1) with APLP-1 or -2 induced production of CTFs slightly larger in size than those seen without BACE1 overexpression (Li and Südhof, 2004). β- secretase processing of APLP2 has also been demonstrated in vivo (Pastorino et al., 2004), whereas this processing event of APLP1 has been questioned (Eggert et al., 2004). Additionally, the production of secreted fragments and CTFs in APLP1 or APLP2 transfected cells were decreased by the metalloproteinase inhibitor batimastat (known to inhibit α-site processing of APP) (Eggert, et al., 2004), indicating α-secretase processing of the two homologs. Recently, the α-, β- and γ-secretase cleavage sites in- APLP2 were determined by mass spectrometry (Hogl et al., 2011). The α- secretase cleavage site was mapped to the Arg670-Val671 bond (in the APLP2 753 isoform), the β-cleavage was shown to occur at the Leu659- Asp660 bond and the γ-cleavage between Ala694 and Val700. Notably, the α-site in APP is at a distance of 12 amino acids from the membrane, whereas the corresponding site in APLP2 is located 22 amino acids from the membrane. The exact cleavage sites in APLP1 have not yet been determined.

1.2.3 Biological functions APP has mainly been studied in relation to AD, but several physiological functions in both the adult and developing nervous system have been suggested, including roles in neuronal migration, cell adhesion, axonal transport, dendritic outgrowth, synapse formation and learning and memory. Even though Aβ is considered the main cause of AD pathogenesis, APP loss of function might also play a role, though this is less investigated. The functions of APP may be mediated by the full-length protein as well as secreted and intracellular fragments.

Full-length APP has been shown to participate in cell-adhesion through direct contacts with extracellular matrix molecules, such as laminin and heparin sulphate proteoglycans and also through trans-dimerization (Narindrasorasak et al., 1992, Soba, et al., 2005). Several studies have also suggested a role of APP, and especially sAPPα in neurite outgrowth

19 (Milward et al., 1992, Perez et al., 1997, Small et al., 1994). The mechanism was further elucidated in a study by Young-Pearce et al., showing that sAPPα interfered with the interaction between APP and integrin β1, thereby inducing integrin β1-mediated neurite elongation (Young-Pearse et al., 2008). The same phenomenon was observed for APLP1 and APLP2 as well. sAPPα has also been shown to have several neuroprotective properties. For example, pre-treatment of sAPPα prevented neuronal death in human cortical cell cultures deprived of glucose or exposed to excitotoxins (Mattson et al., 1993). Further, sAPPα has even been shown to protect against Aβ toxicity (Goodman and Mattson, 1994). Survival of rat hippocampal neurons was reduced to 40% after a four-day exposure to Aβ, whereas 75% of the neurons survived the same exposure when co-treated with sAPPα.

APP knock-out (K/O) mice were generated in order to further study the function of APP in vivo. These mice were viable and fertile, but with minor defects including reduced grip strength, increased copper levels in the cerebral cortex and liver and impaired long term potentiation (LTP) in aged mice paralleled by impaired performance in memory-related behavioral tasks (Dawson et al., 1999, Muller et al., 1994, Seabrook et al., 1999, White et al., 1999). Interestingly, these deficits in LTP in aged APP K/O mice could be rescued by sAPPα knock-in (KI), indicating that sAPPα plays a pivotal role in synaptic plasticity (Ring et al., 2007). About 60% of APP K/O mice die within 48 h after induced cerebral ischemic injury and display reduced cerebral blood flow in response to hypoxia. The same response is seen in BACE1 K/O mice, suggesting a role for APP as a stress response protein mediated by the action of sAPPβ and/or Aβ (Koike et al., 2012).

APLP1 K/O mice were viable and fertile but with reduced body weight, and APLP2 K/O mice had no apparent phenotype neither in young nor in aged mice (Heber et al., 2000, von Koch et al., 1997). However, triple K/O leads to perinatal lethality with a 100% penetrance, clearly demonstrating essential functions for the APP family members during development and also illus- trating a functional redundancy amongst the three homologs (Herms et al., 2004). Furthermore, APP/APLP1 K/O mice were viable and fertile without additional defects compared to single disruptions, whereas APP/APLP2 K/O as well as APLP1/APLP2 K/O mice died within 24 h after birth (Heber, et al., 2000), probably due to pre- and postsynaptic defects at the neuromuscular junction and at central synapses and widening of the endplate (Wang et al., 2005, Wang et al., 2009). These studies indicate that APLP2 has a key physiological role amongst the APP family, being the only member capable of compensating for simultaneous loss of both other family members.

Interestingly, sAPPα-KI, rescued the perinatal lethality of APP/APLP2 (DM) deletion in about 50% of the animals, however, the surviving sAPPα-DM

20 mice displayed a complex phenotype with severe deficits both in synaptic transmission in the adult PNS and CNS and in learning and memory (Weyer et al., 2011). This indicates that although sAPPα is essential for development, other APP or APLP2 derived fragments and/or full-length APP or APLP2 have important functions in the adult nervous system. In contrast to sAPPα, sAPPβ-KI was not able to rescue the perinatal lethality of APP/APLP2 K/O in mice (Li et al., 2010). In addition, sAPPβ has been shown to be further processed and thereby inducing apoptosis through interaction with the death receptor 6 (Nikolaev et al., 2009).

As discussed previously, APP is sequentially processed. This type of proteolytic processing is also called regulated intramembrane proteolysis (RIP). Several other transmembrane proteins, including Notch, a key player in developmental processes, are cleaved in the same way, generating a soluble intracellular domain (ICD) which can translocate to the nucleus and induce gene transcription (reviewed in Lichtenthaler et al., 2011). AICD has also been shown to regulate gene transcription, though this effect is dependent on its interaction with the adaptor protein Fe65 and the histone acetyltransferase Tip60 (Cao and Südhof, 2001). AICD has been shown to regulate transcription of neprilysin, p53, APP, Fe65 and Tip60 (reviewed in Müller et al., 2008). Furthermore, AICD overexpression has also been shown to induce neuron-specific apoptosis (Kinoshita et al., 2002, Nakayama et al., 2008). Recently, several studies have shown that a functional AICD is only generated from the β-pathway (Belyaev et al., 2010, Flammang et al., 2012), since α-pathway generated AICD is quickly degraded in the cytosol.

1.3 Processing enzymes The identity of the APP processing enzymes was initially unknown and they were therefore referred to as α-, β-, and γ-secretases. Today, much more is known about these enzymes, which like the APP family have been shown to be highly conserved. The focus of most research has been on γ-secretase in the hope to develop an inhibitor of Aβ generation. In this thesis, the focus is on regulation of α-secretase processing, and therefore a more detailed description of α-secretase will be presented as compared to β-, and γ-secretase.

1.3.1 α-Secretase α-Secretase cleavage of APP was early determined to occur at the cell surface, suggesting that α-secretase was a plasma membrane-bound protease (Sisodia, 1992). It was soon characterized as a zinc metalloproteases that cleaved APP at the Lys613-Leu614 bond (Roberts et al., 1994). Several

21 members of the ADAM (a disintegrin and metalloprotease) family were suggested as α-secretase candidates, but today the consensus is that ADAM10 and ADAM17, also known as TACE (tumor necrosis factor-α (TNFα) converting enzyme) cleaves APP at the α-site. It should be noted that ADAM9 was previously considered to have α-secretase activity (Koike et al., 1999), though it has since then been shown that ADAM9 regulates APP processing through modulation of ADAM10 activity (Cisse et al., 2005, Moss et al., 2011).

ADAMs are type 1 integral membrane glycoproteins with a multi-domain structure, including signal peptide, pro-domain, catalytic metalloprotease domain, disintegrin/cysteine-rich domain, transmembrane domain and a short cytoplasmic domain (Howard et al., 1996). The pro-domain is cleaved off by a cysteine switch mechanism in order to activate the protease. A major function of ADAM proteases is the ectoplasmic cleavage of cell membrane proteins, also called ectodomain shedding. Many functionally diverse proteins, such as cytokines, growth factors and their receptors, and cell- adhesion proteins are initially synthesized as membrane-anchored proteins, which are released upon proteolysis mediated by ADAMs (reviewed in Edwards et al., 2008).

TACE was identified as an α-secretase candidate since it was shown that it could cleave an α-site spanning peptide (Buxbaum et al., 1998). Several other studies has suggested a role for TACE in α-site cleavage of APP based on the fact that overexpression of TACE or that certain stimuli induces sAPPα secretion in a TACE-dependent manner (Buxbaum, et al., 1998, Mer- los-Suarez et al., 2001, Slack et al., 2001). The relevance of TACE overexpression is discussed further in section 1.3.1.1. Interestingly, no study has shown that TACE cleaves endogenous APP in response to a physiological stimulus. Therefore, the physiological importance of TACE processing of APP during endogenous conditions still remains to be investigated. Lately, ADAM10 was identified as the constitutive α-secretase for APP since knock-down of ADAM10, but not TACE completely blocked sAPPα generation in primary murine neurons (Kuhn et al., 2010), which was in agreement with our previous findings (see section 4.1).

Overexpression of either ADAM10 or TACE has been shown to increase the secretion of sAPLP2 in cultured cells (Endres et al., 2005). Lately, ADAM10 was shown to be the constitutive α-secretase for APLP2 in primary murine neurons, as previously shown for APP (Hogl, et al., 2011, Kuhn, et al., 2010). However, due to the redundant nature of ADAM10 and TACE (discussed below), these studies do not exclude the possibility that APP and APLP2 may be cleaved by TACE in response to certain stimuli.

22 1.3.1.1 Substrate selectivity of TACE and ADAM10 TACE and ADAM10 have a higher sequence homology to each other (39% identity), than to any other mammalian ADAMs (Black et al., 1997). Like all members of the ADAM family, they both have broad substrate specificity that seems to somewhat overlap. However, some substrates are considered to be preferentially cleaved by ADAM10 or TACE, e.g., TGFα (transforming growth factor α) is cleaved by TACE (Peschon et al., 1998) and E-cadherin by ADAM10 (Maretzky et al., 2005). The partial overlap of substrates makes it difficult to understand the factors important for their selectivity. The amino acid sequence of the cleavage site is usually not an important factor for substrate recognition (Deuss et al., 2008, Tsakadze et al., 2006). Instead, it seems like the membrane proximity and the length of the stalk (i.e., the susceptible membrane-proximal region between the ectoplasmic globular domain and the transmembrane domain) is more important. The substrate recognition of TACE and ADAM10 was further studied by Caescu and colleagues (Caescu et al., 2009). They demonstrated by peptide library screening and analysis of individual substrates that TACE and ADAM10 have distinct amino acid preferences at multiple positions surrounding the substrate cleavage site. Furthermore, they performed mutation studies of the active sites of the enzymes and found that certain amino acids in the active site could partially explain why some substrates are selective for either ADAM10 or TACE. Stawikowska et al. have recently shown that the secondary structure of the substrate regulates its interaction with non-catalytic domains of ADAM10 and TACE, thereby differently modifying the enzymes’ activity. For example, inducing α-helical conformation of a TNFα based sequence dramatically increased TACE activity, whereas the effect on ADAM10 activity was not as significant (Stawikowska et al., 2013).

Interestingly, even though certain substrates are thought to be selective for either ADAM10 or TACE, Le Gall and colleagues demonstrated that their substrate selectivity can be manipulated (Le Gall et al., 2009). The group had previously demonstrated that TACE was required for constitutive and phorbol ester-induced TGFα shedding (Peschon, et al., 1998), and were surprised to find shedding activity in TACE-/- cells upon ionomycin (a Ca2+ ionophore) stimulation. Acute treatment of wild type (wt) cells with a selective TACE inhibitor showed that TACE was the main sheddase for TNFα and TGFα. However, upon chronic inhibition of TACE, ADAM10 rescued the shedding of these substrates. This indicates that many studies in which APP and/or ADAM10/TACE have been overexpressed or knocked-out may not be of physiological significance since ADAM10 and TACE clearly can be trig- gered to cleave substrates that they under physiological conditions do not cleave. As further discussed in section 4.1 the latent promiscuity of the enzymes makes it difficult to get a complete understanding of APP α-secretase

23 processing. Clearly, there is a sensitive balance in the selectivity between ADAM10 and TACE. In agreement with LeGall et al. another study showed that even though TACE was the major sheddase for TNFα in most cell types investigated (Hikita et al., 2009), ADAM10 was shown to be the major sheddase in some cell types (e.g. the mouse embryo fibroblast NIH3T3 cell line). The significance of ADAM10 and TACE as sheddases for TNFα could be correlated to the expression levels, but also to the abundance of different TIMPs (tissue inhibitors of metalloproteases) which specifically inhibit either ADAM10 or TACE.

In some cases the overlapping substrate specificity may reflect a selective use of ADAM10 and TACE as sheddases, depending on the stimulus. Both TACE and ADAM10 have been shown to cleave Notch1. However, in a study by Bozulak et al., the role of the two proteases was further investigated. It was found that ADAM10 was required for Notch1 signaling induced by ligand binding (delta-like-1), whereas ligand-independent signaling required TACE (Bozkulak and Weinmaster, 2009). TACE was not able to rescue ADAM10 processing of Notch1 in ADAM10-/- mouse embryonic fibroblasts.

1.3.2 β-Secretase β-Secretase is an aspartyl protease that cleaves APP at the Met597-Asp598 bond, which constitutes the first and rate-limiting step towards Aβ production. Two β-secretases have been identified; BACE1 (β-site APP-cleaving enzyme) and BACE2 (Hussain et al., 1999, Sinha et al., 1999, Vassar et al., 1999, Yan et al., 1999). BACE1 is more widely expressed in the brain and BACE1 K/O animals do not produce any detectable levels of Aβ (Luo et al., 2001, Roberds et al., 2001). BACE1 is a single domain integral protein with its active site located on the ectoplasmic side of the membrane (Hussain, et al., 1999, Vassar, et al., 1999). It has an N-terminal prodomain that is cleaved by a furin-like protease or through autoproteolysis generating a mature enzyme. The optimal pH of BACE1 activity is approximately 4.5, indicating that the -site cleavage of APP takes place in more acidic cellular compartments, such as the endosomes (Vassar, et al., 1999). BACE1 cleavage has also been suggested to occur in lipid rafts (Ehehalt et al., 2003), but this has been challenged since inhibition of BACE1 palmitoylation, a pre- requisite for raft localization has no effect on Aβ production (Vetrivel et al., 2009). A recent study by Das et al. demonstrated that BACE1 was mainly localized to recycling endosomes in cultured hippocampal neurons express- ing low levels of fluorescent-labeled BACE1 and APP. The study further showed that increased neuronal activity induced clathrin-dependent endocytosis of APP, in turn leading to co-localization of APP and BACE1 in recycling endosomes and subsequent β-site cleavage (Das et al., 2013). BACE1

24 was initially thought to be rather specific for the APP family, but today it is known to cleave many substrates. Importantly, it regulates the function of neuregulin-1, which is involved in myelination of Schwann cells in the peripheral nervous system (Willem et al., 2006).

1.3.3 γ-Secretase γ-Secretase is an aspartyl protease that cleaves APP stepwise within the transmembrane domain. The enzyme is a protein complex that consists of anterior pharynx defective 1 (APH-1), nicastrin, presenilin-1 or -2 (PS1 or PS2), and presenilin enhancer-2 (PEN-2) (Edbauer et al., 2003, Francis et al., 2002, Goutte et al., 2000, Yu et al., 2000). Although other proteins are known to interact with the complex, these four components are sufficient for its activity. PS1 is a 9 TM protein and has been shown to harbor the active site of the enzyme (Yu et al., 1998). Two highly conserved aspartate residues (Asp257 and Asp385 in human PS1) within TM6 and TM7 constitute the core of the catalytic site (Wolfe et al., 1999). Both presenilins undergoes endoproteolysis within the cytosolic loop between TM6 and TM7, generating a C- and an N-terminal fragment. The two fragments form a stable heter- odimer, which constitutes the active conformation of presenilin. Interesting- ly, most mutations causing familial AD are found in the PS1 gene.

Nicastrin has one TM and a large ectodomain, proposed to function as a gatekeeper to the PS active site (Shah et al., 2005). Nicastrin also partici- pates in binding of the substrate and is important for the assembly process of the γ-secretase complex. The functions of APH-1 and PEN-2 are not yet fully understood. However, PEN-2 has been shown to stabilize the final complex and APH1 plays an important part during assembly of the complex and has further been implicated in substrate recognition (Chen et al., 2010, Wolfe, 2006). There are two APH-1 homologs that each can undergo alternative splicing generating either a short or a long isoform. The presence of different isoforms in the γ-secretase complex seems to affect its substrate specificity (reviewed in De Strooper et al., 2012). Mature, proteolytically active γ-secretase is mainly localized at the plasma membrane and in the endosomal/lysosomal system (Dries and Yu, 2008). The subcellular localization of the enzyme complex has been shown to affect the ε-cleavage of APP, generating longer AICDs at the plasma membrane than in endosomes (Fukumori et al., 2006). In addition to APP, γ-secretase has more than 90 identified substrates, including Notch, N-cadherin and ephrin B, and the list is continuously growing. Most of the substrates are type I transmembrane proteins that initially is cleaved in the ectodomain by another protease (reviewed in McCarthy et al., 2009).

25 1.4 Signaling pathways affecting APP family processing Direct activation of PKC by phorbol ester stimulation was one of the earliest α-secretase activating stimuli identified (Buxbaum et al., 1993, Caporaso et al., 1992). Like APP, APLP1 and APLP2 processing have been demonstrated to be induced by phorbol esters (Eggert, et al., 2004, Xu et al., 2001). Since then, processing of APP by the amyloidogenic and the non- amyloidogenic pathway has been shown to be differentially modified by activation of certain cell surface receptors such as the 5-hydroxytryptamin (5-HT4) receptor, metabotropic glutamate receptors, muscarinic acetylcholine receptors, synaptic NMDA receptors and many growth factor receptors. Activation of these receptors induces downstream signaling events, including activation of several kinases, small GTPases, cAMP production and regulation of cytosolic calcium levels (reviewed in Allinson et al., 2003). In our studies we have mainly used the neurotrophic factors retinoic acid (RA) and insulin-like growth factor-1 (IGF-1) to induce α-secretase processing. There- fore, a more detailed description of IGF-1 and RA signaling is presented below. It should be noted that both RA and IGF-1 are physiological relevant stimuli that in addition are decreased in AD brains (Corcoran et al., 2004, Steen et al., 2005). Furthermore, previous studies in our group have shown that RA and IGF-1 stimulates α-secretase processing of all three APP family proteins. The mechanisms and signaling involved during this stimulated α- secretase processing will be discussed in section 4.

1.4.1 IGF-1 Insulin-like growth factor-1 (IGF-1) is a hormone that regulates cell survival and differentiation. In the brain, IGF-1 plays important roles in neuroprotec- tion, regulation of energy and has also been implicated in modulation of LTP (reviewed in Carro and Torres-Aleman, 2004). IGF-1 mainly signals through the IGF-1 receptor (IGF-1R), which belongs to the receptor tyrosine kinase family. However, since IGF-1 is structurally related to insulin, it can also bind to the insulin receptor; although with 100-fold lower affinity than to IGF-1R (Navarro et al., 1999). The IGF-1R has a tetrameric structure composed of two extracellular subunits and two transmembrane subunits cova- lently linked by disulfide bridges. Upon IGF-1 binding, the receptor undergoes conformational changes causing autophosphorylation of cytoplasmic tyrosine residues. This in turn leads to recruitment of the adaptor proteins insulin receptor substrate (IRS) and Shc (Src homology/collagen), which then become phosphorylated enabling them to activate the mitogen activated kinase (MAPK) pathway and the phosphatidyl inositol-3 kinase (PI3K)/Akt pathway. Some studies also indicate that the IGF-1R can signal through phospholipase C-γ (PLCγ) (Chattopadhyay and Carpenter, 2002), which hydrolyses phosphatidyl-inositol 4,5-bisphosphate (PIP2) to produce inositol

26 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 promotes release of calcium from the endoplasmatic reticulum (ER) and DAG activates PKC.

Impaired insulin and IGF-1 signaling has been suggested to be a part of AD pathology, since AD patients have been shown to have lower insulin and IGF-1 levels in the cerebrospinal fluid (CSF) and higher plasma insulin levels than non-AD individuals (Steen, et al., 2005). Furthermore, increased serum IGF-1 levels, through chronic subcutaneous infusion for ten weeks have been demonstrated to decrease Aβ levels in the brain of aging rats (Carro et al., 2002), and in AD transgenic mice (Carro et al., 2006). In addition, IGF-1 has been shown to induce a shift in the processing of APP, increasing the levels of sAPPα (Adlerz et al., 2007), which will be discussed further in the results section.

1.4.2 Retinoic acid Retinoic acid (RA) is a vitamin A derivate and a physiological signaling molecule that is involved in neuronal differentiation, axon outgrowth and neuronal plasticity (reviewed in Maden, 2007). Retinol is taken up through the diet and is transported into the cells by retinol binding proteins (RBPs), which interact with membrane receptors. In the cytoplasm, retinol is metabo- lized to RA, which can leave the cell and signal in an autocrine or paracrine fashion. RA exerts its effect by translocating to the nucleus, assisted by cellular RA-binding proteins (CRABs). RA then binds to the RA receptor (RAR) or the retinoic X receptor (RXR), which both are ligand-activated nuclear receptors. RAR and RXR function as heterodimers, and bind to RA response elements (RAREs) in the target genes. Binding of RA to the receptor complex induces gene transcription of the target genes, e.g., various transcription factors, cell signaling molecules, structural proteins etc. RA in complex with RAR can also signal through rapid non-transcriptional pathways, through direct interaction and subsequent activation of PKC (Hoyos et al., 2000), PI3K and MAPK (Masia et al., 2007).

Interestingly, a connection between RA signaling and AD has been observed, since vitamin A deprived mice displayed Aβ accumulation and RAR down-regulation (Corcoran, et al., 2004). Furthermore, RA has been shown to regulate the expression of APP, ADAM10, TACE and BACE1 (see section 4.2), as well as to induce a shift in the APP processing toward the α- secretase pathway (Adlerz et al., 2003, Holback et al., 2007, Holback et al., 2005).

27 1.5 Regulation of APP family α-site processing As mentioned above, activation through signaling pathways can alter processing of APP. However, the mechanism behind this is not well characterized. One important determinant is co-localization of enzyme and substrate in subcellular compartments. Moreover, post-translational modifications, such as phosphorylation or O-GlcNAcylation of substrate, enzyme or inter- acting protein could affect localization, enzyme activity or substrate conformation resulting in effects on α-secretase processing.

1.5.1 Trafficking of enzyme and substrate Since α- and β-secretase processing of APP is spatially segregated; ADAMs are mainly active at the cell surface whereas BACE1 is active in endosomes, it is clear that APP trafficking plays a pivotal role in regulating its processing. APP is synthesized in the ER and is then targeted into the secretory pathway and transported to the cell surface. During the transit through the Golgi, APP is glycosylated and sulfated. Only 10% of APP is located at the plasma membrane, based on overexpression of APP in cultured cells, whereas the majority of APP is localized to the Golgi and trans-Golgi network (TGN). In neurons, APP is rapidly transported to axons and dendrites in post-Golgi transport vesicles mediated by kinesin-1 (Kins et al., 2006).

APP is not cleaved at the cell surface is rapidly internalized, due to the presence of the YENTPY internalization motif in the C-terminal (Lai et al., 1995), followed by delivery to endosomes. Blockage of APP endocytosis through deletion of the internalization motif or by overexpression of a dominant-negative dynamin mutant impairs Aβ production (Carey et al., 2005, Chyung and Selkoe, 2003). A fraction of endocytosed APP is then recycled to the cell surface and some is degraded in lysosomes (Haass et al., 1992). APP may also be shuttled (in both directions) between the TGN and lysosomes. APLP2 has been shown to undergo similar trafficking, whereas APLP1 is present at the cell surface to a much higher degree (Kaden, et al., 2009). Although the main site for APP α-secretase cleavage is the cell surface, some studies also implicate the TGN as a site for this cleavage event (Skovronsky et al., 2000).

There are many proteins known to regulate the trafficking of APP. The YENPTY motif in APP serves as a binding site for several cytosolic proteins, including Fe65, Fe65-like proteins-1 and -2, Mint/X11 1-3, Dab1 and Shc (reviewed in Jacobsen and Iverfeldt, 2009). Fe65 and Mint/Xll are the most well studied APP adaptor proteins and have been shown to interact with both APLPs. Overexpression of Fe65 or X11 in AD transgenic mice leads to decreased Aβ generation, possibly through regulation of APP traf-

28 ficking to the cell surface (Ando et al., 2001, Borg et al., 1998, Lee et al., 2003a, McLoughlin and Miller, 2008). In addition to APP and APLPs, Fe65 can also interact with the endocytotic receptor LRP1 (low-density lipoprotein (LDL) receptor-related protein 1), forming a trimeric complex with APP (Pietrzik et al., 2004) and thereby accelerating internalization of APP (Cam et al., 2005). APP trafficking has also been shown to be regulated by the neuronal trafficking receptor SorLA (sorting-related receptor with A-type repeats) (Andersen et al., 2005). SorLA has been suggested to interact with the N-terminal of APP in early endosomes, where retromer, an adaptor complex in the endosome-to-Golgi retrieval pathway, induces retrograde sorting of the SorLA-APP complex to the TGN (Fjorback et al., 2012). This results in decreased levels of APP at the plasma membrane and subsequently less α- and β-secretase cleavage. Another receptor in the same family (the Vps10- domain receptor family), namely Sortilin has also been shown to affect APP processing (Gustafsen et al., 2013). Sortilin interacts with the E2 domain of APP in neurites of hippocampal neurons and promotes α-secretase processing of APP without affecting subcellular localization of APP.

There is limited information concerning how TACE and ADAM10 trafficking is regulated and how it affects APP processing. Similar to APP, both TACE and ADAM10 are synthesized in the ER and are then targeted into the secretory pathway and transported to the cell surface. Immunohistochemical studies suggest that most active TACE is localized in the cellular perinuclear region, with a small amount present at the plasma membrane (Schlondorff et al., 2000). However, as it was pointed out by a group who analyzed TACE trafficking, overexpressed TACE is processed differently than its endogenous counterpart, which should be considered when analyzing trafficking of overexpressed TACE (Borroto et al., 2003). TACE trafficking has been reported to be regulated by iRhom2, a proteolytically inactive member of the rhomboid family through direct interaction with TACE, thereby promoting exit from the ER. It was further shown by siRNA down-regulation of iR- hom2, that the protein was essential for lipopolysaccharide (LPS)-induced TNFα shedding (Adrain et al., 2012). TACE has further been shown to interact with several binding partners, but nothing is known about how they affect TACE trafficking (reviewed in Gooz, 2010). TACE trafficking is affected by phorbol ester stimulation, which will be further discussed in section 1.5.2.

ADAM10 trafficking was shown in a study by Dornier and colleagues to be regulated by tetraspanins of the TspanC8 subfamily (Dornier et al., 2012). Tetraspanins are integral membrane proteins implicated in several cellular processes, including membrane compartmentalization (Berditchevski and Odintsova, 2007). Confocal microscopy analysis demonstrated that TspanC8 expression led to redistribution of ADAM10 from the ER to the cell surface

29 in HeLa cells, which have low TspanC8 endogenous expression and also low endogenous cell surface ADAM10 levels. ADAM10 retention in the ER had previously been shown to be dependent on a mechanism involving an arginine-based motif in the C-terminal of ADAM10 (Marcello et al., 2010), that probably is masked by complex formation with tetraspanins, thereby allow- ing ER exit. Similar to APP, ADAM10 endocytosis has been shown to be inhibited by expression of a dominant negative form of dynamin, resulting in increased cell surface levels of both mature and immature ADAM10 and subsequent increase of sAPPα (Carey et al., 2011). Neither phorbol esters nor muscarinic receptor M3 activation, known to increase sAPPα secretion was shown to affect ADAM10 cell surface localization. SAP97 (synapse associated protein 97), which traffics proteins to the excitatory synapse has been shown to interact with ADAM10 (Marcello et al., 2007). NMDA receptor activation increased the interaction between SAP97 and ADAM10, leading to ADAM10 trafficking from dendritic shaft to spine-like structures. In vivo experiments have further demonstrated that disruption of the interaction between ADAM10 and SAP97 in adult mice, leads to decreased sAPPα secretion and increased Aβ production as well as tau hyperphosphorylation after only two weeks (Epis et al., 2010). Furthermore, ADAM10/SAP97 interaction was shown to be reduced in the brain of AD patients.

1.5.2 Post-translational modifications of enzyme and substrate The post-translational modification we have focused on in our studies is O- GlcNAcylation. The reason for our interest in O-GlcNAcylation is firstly that already in 1995 APP was reported to be O-GlcNAc modified (Griffith et al., 1995) and secondly that during recent years O-GlcNAcylation has emerged as an important type of post-translational modification playing roles in synaptic function and plasticity (Hart et al., 2011, Tallent et al., 2009). O- GlcNAcylation has also been directly linked to AD since the level of protein O-GlcNAcylation has been shown to be decreased by approximately 20% in AD brains (Liu et al., 2004b). Furthermore, it has been demonstrated that decreased O-GlcNAcylation of tau is correlated to tau hyperphosphorylation and thereby increased formation of neurofibrillary tangles.

O-GlcNAcylation is the attachment of the monosaccharide β-N- acetylglucosamine (GlcNAc) to serine and threonine residues of target proteins. The modification, unlike other glycosylations is dynamic and can cycle rapidly on and off the target protein similar to phosphorylation. In contrast to phosphorylation, which is regulated by several different enzymes, O- GlcNAcylation is regulated by only two enzymes, O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA), which catalyzes the attachment and removal of O-GlcNAc, respectively (Dong and Hart, 1994, Haltiwanger et al., 1992). OGT utilizes UDP-GlcNAc produced in the hexosamine biosynthetic

30 pathway. OGA K/O in mice results in developmental delay (Yang et al., 2012), and OGT K/O is lethal (Shafi et al., 2000). The levels of O-GlcNAc seem to be regulated by the expression of OGT and OGA, the concentration of UDP-GlcNAc and availability of protein substrates (Groves et al., 2013). The cycling of O-GlcNAc is influenced by both intracellular and extracellular stimuli, including insulin, nutrient levels, cellular stress, cell cycle changes and development. NMR studies have shown that O-GlcNAcylation of proteins may directly influence the conformation of the target protein (Simanek et al., 1998), which may affect interactions with other proteins. O- GlcNAcylation has also been shown to affect cellular localization. In this thesis, we have investigated O-GlcNAcylation in relation to effects on α- secretase processing of the APP family.

In addition to O-GlcNAcylation, other post-translational modification of both substrate (the members of the APP family) and the enzyme (ADAM10 and/or TACE) might be important in regulating α-site processing, possibly through the effect on enzyme activity or interactions with other proteins thereby inducing changes in subcellular localization. Here, some of these modifications, mostly glycosylation and phosphorylation are discussed in relation to APP family processing.

1.5.2.1 Post-translational modifications of ADAM10 and TACE The most important post-translational modification of ADAM10 and TACE is the pro-domain removal, which occurs mainly in the TGN. Furin, as well as several other pro-hormone convertases (PCs) catalyzes the pro-domain removal of ADAMs. PC7 has been specifically implicated in maturation of ADAM10, whereas PC1-6 seems to be important for TACE (Anders et al., 2001, Srour et al., 2003). In addition to inhibiting ADAM activity during biosynthesis, the pro-domain has also been suggested to possess chaperone functions to ensure correct folding of the enzyme (Anders, et al., 2001). In- terestingly, mutations in the pro-domain (Glu170His and Arg181Gly) have been shown to impair this chaperone function of the pro-domain, without affecting ADAM10 trafficking, thereby decreasing sAPPα production and increasing Aβ formation (Suh et al., 2013). These mutations have been linked to late onset AD.

TACE has been shown to be phosphorylated, and stimulation with phorbol esters, which enhances the shedding activity of TACE, was reported by Soond et al. to increase phosphorylation of pro-TACE at Thr735 in the intracellular domain (Soond et al., 2005). The phosphorylation was extracellular signal-regulated kinase (ERK)-dependent, and increased both maturation of TACE and its translocation from ER to the cell surface. TACE has further been shown to co-localize to the plasma membrane with one of its substrates,

31 CD44, within 10 min after phorbol ester stimulation (Nagano et al., 2004). These studies suggest that phosphorylation of TACE affects the subcellular localization, leading to increased substrate shedding. In a study by Alfa Cis- sé and colleagues it was demonstrated that mutations at Thr735 abolished the carbachol (an acetylcholine receptor agonist)-induced increase of TACE phosphorylation as well as the induced processing of the cellular prion protein (Alfa Cisse et al., 2007). This study demonstrates a clear and direct correlation between TACE phosphorylation and activity induced by a physiological stimulus. However, in one study the cytoplasmic domain of TACE was shown not to be required for phorbol ester stimulated shedding of TNFα (Reddy et al., 2000), which have made the significance of TACE phosphorylation a debated area. TACE has also been shown to be N-glycosylated (Peiretti et al., 2003), but there are no data how this affects its activity or trafficking.

ADAM10 is N-glycosylated at four sites, three located in the metalloprotease domain and one in the disintegrin domain. Mutation of the N278 site in the metalloprotease domain of ADAM10 resulted in accumulation in ER as the unprocessed immature pro-form. Mutations of the three other N- glycosylation sites (N257, N439 and N551) all lead to cell surface localization similar to wt ADAM10. However, the mutated constructs showed strongly reduced ability to cleave the natural substrate L1 cell adhesion molecule, as compared to wt ADAM10. The authors speculate that the lack of glycosylation does not directly interfere with substrate binding, but instead affects the conformation of the protease, causing a decreased activity (Escrevente et al., 2008).

1.5.2.2 Post-translational modifications of the APP family In addition to proteolytic processing, APP is known to undergo several different kinds of post-translational modifications, including sulfation, sumoy- lation, siaylation and addition of CS GAG (chondroitin sulphate glycosaminoglycan). The two major modifications of APP, glycosylation and phosphorylation, and their relation to APP processing will be discussed here.

There are two consensus sites for N-glycosylation in APP, N467 and N496, but only N467, which also is highly conserved within the family, has been shown to be modified by complex glycans (Pahlsson et al., 1992). Three O- glycosylation sites have been identified by mass spectrometry, namely Thr291, Thr292 and Thr576 (Perdivara et al., 2009). There are contradictory results concerning whether or not O-glycosylation is required for ectodomain shedding of APP (Pahlsson and Spitalnik, 1996, Tomita et al., 1998). Several reports show that inhibition of N-glycosylation diminishes sAPP secretion (McFarlane et al., 1999, Pahlsson and Spitalnik, 1996). However, when mu-

32 tating the N-glycosylation sites in APP, secreted fragments were still detected in the cell medium of transfected cells, indicating that N-glycosylation of APP is not essential for secretion. The rate of the secretion was however reduced (Pahlsson and Spitalnik, 1996, Yazaki et al., 1996). Instead, N- glycosylation of another, yet unidentified protein seems to be necessary for APP shedding. Another study by Eggert et al. show that only APLP1, and not APP or APLP2 processing is affected by tunicamycin (N-glycosylation inhibitor), leading to production of α-, and γ-generated fragments that were not seen in absence of tunicamycin (Eggert, et al., 2004).

APP has several sites in its intracellular domain which can be phosphorylated in vitro (Gandy et al., 1988, Knops et al., 1993, Suzuki et al., 1994), and many of these sites have also been found to be phosphorylated in post- mortem brain tissue of patients with AD (Lee et al., 2003b). Several studies have shown that Thr668 phosphorylation affects APP processing, mostly indicating increased Aβ production, but there are contradictory reports (reviewed in Jacobsen and Iverfeldt, 2009). It was later shown that the peptidyl- prolyl cis/trans isomerase Pin1 binds to Thr668 phosphorylated APP both in vitro and in vivo, thereby promoting large conformational changes in the C- terminal by accelerating the production of the trans conformation of Pro669 (Pastorino et al., 2006). It was further demonstrated that overexpression or knockout of Pin1, reduced or increased the secretion of Aβ, respectively. Pin1 K/O mice were further shown to display age-dependent neuropathy (Liou et al., 2003). In post-mitotic neurons both cyclin-dependent kinase 5 (cdk5) and glycogen synthase kinase-3β (GSK-3β) have been implicated to phosphorylate APP at Thr668 (Aplin et al., 1996, Iijima et al., 2000). Other phosphorylation sites besides Thr668 have also been implicated in the regulation of APP processing, e.g., Tyr687, since overexpression of the mutant Tyr687Ala decreased the level of the C83 fragment (Takahashi et al., 2008). The mechanism of altered APP processing upon phosphorylation may include altered interactions with intracellular adaptor proteins, such as Fe65 and X11/Mint.

Synthetic peptides corresponding to the cytoplasmic domain of APLP1 and APLP2 have been shown to be phosphorylated by PKC (Suzuki et al., 1997), as previously shown for APP (Gandy, et al., 1988). Alignment of the APP family members shows that the APP Thr668 phosphorylation site is conserved in all APP homologues; APLP1 (Thr624) and APLP2 (Thr636). APLP2, as well as APP was observed in a study by Taru and Suzuki to be phosphorylated at this site in response to cellular stress (Taru and Suzuki, 2004). Furthermore, Pro669 in APP, that undergoes isomerization following Thr668 phosphorylation is conserved in APLP2 (Pro737) but not in APLP1. Four other phosphorylation sites in APP, Tyr653, Tyr682, Thr686 and Tyr687, are also conserved in all APP homologues.

33 2. Methodological considerations

2.1 Cell lines We have used human neuroblastoma SH-SY5Y cells as a model to study APP family processing. This cell line is used in all papers of this thesis and is a commonly used cell line serving as an in vitro model for many types of studies especially concerning neuronal function and differentiation. The SH- SY5Y cell line is a third generation clone originating from SK-N-SH, which was isolated from a metastatic tumor in the bone marrow of a four year old girl in 1971 (Biedler et al., 1978). SH-SY5Y cells can be differentiated into a neuron-like phenotype by treatment with retinoic acid (RA) (Pahlman et al., 1984). During RA-differentiation, SH-SY5Y cells express TrkB receptors (Encinas et al., 2000), which can be activated by BDNF and thereby inducing further differentiation of SH-SY5Y cells into an even more neuron-like phenotype. SH-SY5Y cells have been shown to endogenously express all members of the APP protein family (Adlerz, et al., 2003, Beckman and Iver- feldt, 1997), and are commonly used to study APP processing. SH-SY5Y cells also express insulin and IGF-1 receptors. It should be noted that in contrast to what is seen in primary cortical neurons, α- and β-secretase processing of APP was recently found not to be coupled in SH-SY5Y cells (Colombo et al., 2012). This is however not in agreement with our studies using SH-SY5Y cells, were we consistently have seen an inverse coupling of sAPPα and Aβ secretion into the cell medium.

In paper V, we also used human neuroblastoma SK-N-AS cells and human embryonic kidney HEK293 cells to compare effects on APP processing in different cell lines. SK-N-AS is derived from the bone marrow of an eight year old boy. The cancer cells had metastasized from adrenal glands to the bone marrow. The HEK293 cell line was generated by adenovirus transfor- mation of cultured human kidney cells derived from a single healthy fetus legally aborted under Dutch law (Graham et al., 1977).

34 2.2 Cell treatments During this work, we have used RA and IGF-1 in combination with several different kinase inhibitors to analyze signaling pathways involved in APP family processing. Both RA and IGF-1 had previously been shown by our group to induce a shift in APP processing towards the α-secretase pathway (Adlerz, et al., 2007, Holback, et al., 2005). In addition, inhibitors targeting OGT and OGA have been used to analyze effects of O-GlcNAcylation.

2.2.1 Retinoic acid and IGF-1 RA was used in paper I and II. SH-SY5Y cells were treated for 6 days with 10 µM RA in serum-containing medium or 1 µM RA in serum-free medium. In paper II, we also treated the cells with RA concomitantly with 50 ng/ml BDNF to analyze additional effects on expression levels of ADAM10, TACE and BACE1, since it previously was shown in the group that BDNF further increase RA-induced α-secretase processing (Holback, et al., 2005). We used IGF-1 to stimulate α-secretase processing in paper I and V. We treated the cells for 18 h with 10 nM IGF-1 in serum-free medium supple- mented with N2 (transferrin, insulin, progesterone, selenite and putrescine). Longer treatments of IGF-1 were avoided to minimize the effects of IGF-1 on the expression levels of the APP family members.

2.2.2 Pharmacological inhibitors The effect of various kinases during IGF-1 and RA treatment was investigated in paper I, II and V by co-treating the cells with pharmacological inhibitors for the last 18 h of the treatment. To inhibit PKC we used bisindolylma- leimide XI (BIM11), which is a staurosporine derivate that is cell-permeable and blocks PKC activity by competitively inhibiting the binding of ATP by the kinase. BIM11 inhibits all PKC isoforms, but with varying potency with a 3- to 10-fold preference for the PKCα isoform. It should be noted that an independent screen of various kinase inhibitors revealed that the related BIM10 also inhibited Msk-1 and MAPKAP-K2 to a similar potency as PKC (Davies et al., 2000). These kinases are further “downstream” in signaling pathways than PKC, and are mainly involved in phosphorylation of the transcription factor CREB at Ser133. In paper I, we also used the natural compound curcumin as an unselective PKC inhibitor. However, curcumin is known to also inhibit c-Jun kinase (JNK) as well as the transcription factors NFκB and AP-1. LY29002 (LY) is a reversible, cell-permeable inhibitor of PI3K that acts through competitive binding to the ATP-binding site of the catalytic subunit of the kinase. According to the manufacturer (Merck) LY does not affect the activity of MAPK, PKC or PI4K at low concentrations. However, in the previously mentioned independent screen, LY was shown to

35 also inhibit casein kinase 2, which is involved in cell cycle control and DNA repair. PD098059 (PD) was used to inhibit MEK1/2. It interacts with the dephosphorylated form of MEK1, thereby preventing its activation. PD also works as a weak inhibitor of MEK2. PD was one of the most selective kinase inhibitor in the independent screen.

To raise the level of cellular O-GlcNAcylation in paper III, IV and V we used the OGA inhibitor PUGNAc. PUGNAc is a 1,5-hydroximolactone, proposed to be a transition-state analogue. Lately, it has been shown that PUGNAc, in addition to its effect on OGA, acts as an inhibitor of a variety of N-acetylhexosaminidases. In paper III, we also used siRNA down- regulation of OGA to confirm that the effects on APP processing stimulated by PUGNAc indeed were mediated through decreased O-GlcNAcylation and not by off-target effects.

2.3 siRNA gene silencing In paper I we wanted to investigate the effect of TACE on APP and APLP2 processing, and since no specific pharmacological inhibitor was available on the market at the time, we chose to knock-down the expression of TACE with short interfering RNA (siRNA). We also used siRNA targeted against OGA in paper III to further strengthen the effects seen with the inhibitor PUGNAc. siRNA, like the name suggests, are small RNA sequences which interfere with the expression of a specific gene. siRNA incorporates into the RNA-induced silencing complex (RISC), which cleaves the target mRNA strand complementary to the bound siRNA, leading to post-transcriptional gene silencing. Two strategies were applied to knock-down the expression of the target protein. SH-SY5Y cells were transfected with either 100 nM of a mix of 4 different siRNA sequences, or with 5-50 nM of one single sequence (two different sequences were used separately). The use of different sequences makes it possible to determine the specificity of the knock-down, and to exclude off-target effects. In addition, a negative control composed of non-targeting siRNA was used.

2.4 BCA assay The protein content in the cell lysate and cell medium was determined using bicinchoninic acid (BCA) assay. The principle of the BCA assay is similar to the Lowry procedure. However, in the BCA assay the mechanism can be divided into two reactions. First, cysteines, tryptophans, tyrosines, and the peptide bond will reduce Cu2+ from the cupric sulfate present in the BCA stock solution, to Cu1+. Then, Cu1+ will form a purple-blue complex, by che-

36 lating with two BCA molecules. This complex has a strong absorbance at 562 nm. The amount of protein present in a solution can be quantified by measuring the absorption spectra and comparing with protein solutions, e.g., bovine serum albumin (BSA) with known concentrations. The BCA assay is more sensitive than Bradford and Lowry, and is also less susceptible to de- tergents.

2.5 Western blot Western blot is a frequently used method to analyze the expression levels of a specific protein. The protein sample is first boiled together with sodium dodecyl sulphate (SDS), which denaturates the proteins and gives them a negative net charge in proportion to their size. This difference in charge can then be used to separate proteins according to size on a polyacrylamide gel that is run in an electric field. The separated proteins in the gel are then transferred to a membrane with the help of an electric current. The protein of interest can now be probed for by using a specific antibody directed against the target protein. A secondary antibody, coupled to horseradish peroxidase (HRP) that binds to the primary antibody is added in order to visualize the protein by enhanced chemiluminescence (ECL). When adding a solution containing hydrogen peroxide (H2O2) and luminol, HRP will catalyze the oxidation of luminol by H2O2, generating acridium ester intermediates, which in turn will react with peroxide and generate an excited state that emits light as it decays to a lower energy level. This signal is captured on a film or by a CCD camera. It is important to make sure that the exposure time is not too long, i.e., the signal is saturated. The relative abundance of the protein is quantified by densiometric analysis. Both the specificity and sensi- tivity of the method mainly relies on the nature of the antibody. If there is cross reactivity of the antibody, the results may be difficult to interpret and the specificity of the band should be confirmed using either overexpression or knock-down of the target protein. It is also important to make sure that a correct concentration of the antibody is used to ensure quantitative meas- urements.

2.6 ELISA Enzyme-linked immunosorbant assay (ELISA) was carried out in paper I and III to analyze the levels of secreted Aβ40, since this is a more sensitive method compared to western blot. The method, like western blot is based on the binding of an antigen (in this case Aβ40) to a specific antibody. We used sandwich ELISA, in which a capture antibody directed against the N- terminal of Aβ40 is immobilized on the bottom surface of the wells in a 96

37 well plate. The cell medium sample was then added to the wells, resulting in Aβ40 binding to the antibody. Without the first layer of capture antibody, any proteins in the sample may competitively adsorb to the plate surface. A detection antibody directed against the C-terminal Aβ40 and a secondary, HRP- linked antibody was added before colorimetric detection. For colorimetric detection, the antibody-antigen-antibody-antibody/HRP complex (or sandwich) is incubated with a chromogen, in this case tetramethylbenzidine (TMB). TMB is oxidized during the enzymatic degradation of H2O2 by HRP. The oxidized product of TMB has a deep blue color, but after addition of an acidic stop solution, a clear yellow color is formed. The optical density of the yellow color is measured at 450 nm, and the intensity reflects the amount of Aβ40. In order to translate intensity to concentration, standard samples containing known concentrations of Aβ40 are run alongside the cell medium samples. This is a highly sensitive method, detecting Aβ40 concentrations as low as 6 pg/ml. However, when treating SH-SY5Y cells with either IGF-1 or PUGNAc, the Aβ levels decrease close to the detection limit. Previous studies in the group have shown that the Aβ42 levels follow the same trend as Aβ40 during similar conditions as used in the studies in this thesis.

2.7 32P-labeling and immunoprecipitation In paper II we wanted to study if TACE was phosphorylated in response to IGF-1. A common method to study phosphorylation of proteins is 32P- labeling. Cells are incubated with 32P-orthophosphate and during this incubation the 32P will be taken up by the cells and incorporated into ATP. In this study, we labeled phospho-proteins in cells in the presence or absence of a protein kinase C inhibitor. After the labeling, the cells were treated with IGF-1 before harvesting. To isolate TACE from the cell lysate we performed immunoprecipitation (IP) with an antibody directed against the C-terminal of TACE. IP is based on specific binding of an antibody against the target protein and immobilization on sepharose or agarose beads. The complex con- sisting of sepharose/agarose beads, antibody and target protein can then be precipitated by centrifugation. The precipitated protein is finally released from the beads by boiling in SDS-sample buffer. The immunoprecipitated TACE sample was resolved by SDS-PAGE and the levels of phosphorylated TACE was then visualized by exposure to a phosphoimager screen. The signal to noise ratio is often low in these kinds of experiments due to the difficulty to enrich large amounts of phosphoproteins. Immunoprecipitation was also used in paper III and IV. High background is a common problem when only using immunoprecipitation followed by western blot, since secondary antibodies also will detect the denatured heavy and light chain of the immunoglobulin.

38 2.8 sWGA precipitation Succinylated weat germ agglutin (sWGA) is a lectin that specifically binds GlcNAc. Before succinylation, WGA will recognize both silaic acid and GlcNAc. In paper IV we used sWGA coupled to agarose beads to isolate and precipitate O-GlcNAcylated proteins in the cell lysate. Following incubation of the cell lysate with the sWGA-bead slurry, O-GlcNAcylated proteins are precipitated by centrifugation and released from the beads by boiling in SDS-sample buffer. The levels of APP or ADAM10 that precipitated with sWGA beads were then analyzed by western blot. In paper III, we used immunoprecipitation with an antibody targeted against O-GlcNAc for the same purpose. As discussed in paper IV and in the results section of this thesis, the fact that a protein is precipitated with sWGA or an O-GlcNAc antibody does not necessary mean that the protein is O-GlcNAcylated. The precipitated protein may instead be in complex with a protein that is O-GlcNAcylated. In our study, this is further investigated by pre-treating the cell lysate with SDS, thereby breaking all complexes, before sWGA precipitation.

2.9 Biotinylation assay In paper IV we wanted to study APP trafficking in response to increased O- GlcNAcylation. Since the cell surface is the main site for α-secretase cleavage, we used a biotinylation assay to analyze APP levels on the cell surface. This assay makes use of sulfo-NHS-SS-biotin, which is a non-membrane permeable compound that reacts with amines. The cells are incubated in PBS containing sulfo-NHS-SS-biotin at 4oC (to inhibit endocytosis), resulting in selective biotin labeling of all cell surface proteins via a disulfide bond. Cells are then lysed and biotin-labeled proteins are isolated by their selective binding to streptavidine that has been conjugated to agarose beads. The proteins are then released from the streptavidine-bound biotin label through dithiothreitol (DTT) reduction of the disulfide bond. Cell surface APP can then be detected using western blot. To verify membrane integrity during the biotinylation protocol and that only surface proteins have been biotinylated, the western blot membrane can be stripped and re-blotted against intracellular proteins such as tubulin or microtubule-associated protein 1a (MAP1a).

2.10 Design and cloning of the APP/APLP2 chimer The initial strategy of this project was to first create APP/APLP2 chimers where large parts of the proteins had been replaced. After analyzing the effects of these substitutions on ADAM10/TACE specificity during IGF-1 stimulation, the domain that was found to be determining the specificity

39 would be further investigated by narrowing down the substituted part. We decided to focus on the ectoplasmic part of APP, since this is the part that is cleaved. Two chimeric constructs were created, substituting either the E1 or E2 domain in APP with the corresponding part of APLP2. The constructs were created by inserting unique restriction sites flanking the domain to be replaced. The sites were inserted using Pfu Ultra mediated site-directed mu- tagenesis. The same restriction sites were also incorporated, using PCR in the 5’- and 3’-end of the specific APLP2 domain, enabling ligation (after restriction site cleavage) into the mutated vector. Unfortunately, we later realized that the E2 domain in APLP2 that we inserted into APP contained a KPI domain. Since this domain is not present in the APP695 isoform, it would be preferable to replace the domain with the corresponding domain in an APLP2 isoform that also does not contain a KPI domain (e.g., the 707 isoform). This APP/E2/APLP2 construct was successfully expressed in SH- SY5Y cells but the secretion was not stimulated by IGF-1 and displayed broad “smeared” bands on the western blot using an anti-FLAG antibody (data not shown). A new E2 chimer is currently being constructed.

Since our focus was to analyze secretion of the constructs, we wanted an N- terminal tag. However, APP is normally expressed with a signal sequence, directing it to the plasma membrane, which is then cleaved off. To avoid removal of the N-terminal tag, the signal sequence in APP was deleted. Fur- thermore we used the pCMV8 vector (Sigma Aldrich) which contains the preprotrypsin leader sequence and a 3xFLAG-tag preceding the multiple cloning site. A drawback of using this vector in our study is the CMV promoter, which is a very strong promoter causing high degree of expression. This is further discussed in the results section and paper V.

40 3. Aims

sAPPα levels in the brain are reduced in both familial and sporadic AD (Lannfelt et al., 1995), and several studies have shown that α-secretase processing is negatively correlated with Aβ production. Furthermore, mutations in ADAM10 have been linked to late onset AD (Suh, et al., 2013). Together, this illustrates the importance of studying the regulation of α-site cleavage of APP, since enhancers of this processing pathway clearly constitutes a potential therapeutic strategy. The aim of this thesis was to elucidate signaling pathways involved in stimulated APP α-site processing. In addition, we wanted to further investigate the previously shown difference in regulation of APP and APLP2 processing (Adlerz, et al., 2007).

The more specific aims of this thesis were to:

. Determine the signaling pathways involved in IGF-1- and RA- induced processing of APP and APLP2. . Determine if the same α-secretase cleaves both APP and APLP2. . Investigate the role of O-GlcNAcylation on APP and APLP2 processing. . Investigate which part of APP and APLP2 that determines their α- secretase selectivity.

41 4. Results and discussion

All five papers are focused on the α-secretase processing of the APP family during stimulated conditions, and more specifically the distinct mechanisms regulating processing of APP and APLP2. The most important results from these articles are discussed below.

4.1 Stimulated α-secretase processing of APP is mediated by ADAM10 in a PI3K-dependent manner (Paper I and II) It has previously been shown that RA and IGF-1 induce α-secretase processing of all three members of the APP family in SH-SY5Y cells (Adlerz, et al., 2007, Holback, et al., 2005). Furthermore, the signaling pathway behind IGF-1-induced APP processing was investigated and was found to be completely dependent on PI3K (Adlerz, et al., 2007). Based on these studies we aimed to determine if also the RA-induced APP processing was dependent on the same signaling pathway. We used an antibody that recognizes the first 17 amino acids in the Aβ region, making it possible to specifically detect sAPPα in conditioned cell medium from RA-differentiated SH-SY5Y cells. Our results show that inhibition of PI3K reduced the RA-induced sAPPα secretion by ~50% (Paper I: Fig. 1B). So why is the RA-induced APP ectodomain shedding not completely blocked by the PI3K inhibitor, as seen for the IGF-1-induced processing? It should be pointed out that RA was added for 6 days, as compared to 18 h with IGF-1, and the PI3K inhibitor was only present the last 18 h. RA can mediate its effects both by transcriptional effects through activation of nuclear retinoid receptors as well as direct effects on protein kinases (see section 1.4.2). The PI3K-independent effects on RA-induced APP processing are probably a result of transcriptional effects, such as up-regulation of α-secretases, which have been induced during several days.

Next, the involvement of PKC during RA- and IGF-1-induced APP ectodomain shedding was investigated using a PKC inhibitor. The RA-induced processing was found to be completely unaffected by PKC inhibition where-

42 shown to be strongly affected by PI3K inhibition (Paper II: Fig. 2). This suggests that PI3K is involved in stabilizing mature ADAM10 and not involved in the transcriptional effects of RA on ADAM10. This stabilization of mature ADAM10 is subsequently important for the increased APP α-site processing seen during stimulated conditions.

4.2 Stimulated α-secretase processing of APLP2 is mediated by TACE in a PKC-dependent manner (Paper I and II) In contrast to APP, which previously was shown to be processed in a PI3K- dependent manner during IGF-1 stimulation, APLP2 processing was unaffected by blockage of this signaling pathway during the same conditions (Adlerz, et al., 2007). To elaborate this work, we used a PKC inhibitor to determine if signaling pathways including this kinase was important for stimulated APLP2 processing. Activation of PKC by phorbol esters has previously been shown to induce both APP and APLP2 processing (Caporaso, et al., 1992, Xu, et al., 2001). Our results demonstrate that the IGF-1-induced APLP2 processing was completely blocked by PKC inhibition (Paper I: Fig. 2). Furthermore, the RA-induced sAPLP2 secretion was also shown to be completely dependent on PKC (Paper 1: Fig. 1B). Our results suggest that IGF-1 induces activation of a pathway involving PKC that is independent of PI3K, probably through receptor activation of PLCγ. This is not a consensus signaling pathway for IGF-1R, but has been established in some cell lines (c.f., (Chattopadhyay and Carpenter, 2002). The signaling behind RA- induced APLP2 processing might involve up-regulation of PKC or direct interaction and thereby activation of PKC (see section 1.4.2). Together our data suggest that stimulated APLP2 processing is PKC-dependent during these conditions.

Based on the difference in signaling behind stimulated APP and APLP2 processing, we speculated that the two proteins were cleaved by different enzymes. Indeed, knocking down the expression of TACE using siRNA we found a strong correlation between the degree of TACE down-regulation and the degree of IGF-1-induced sAPLP2 secretion (Paper I: Fig. 4). On the other hand, inhibition of ADAM10 had no effect on the IGF-1-induced processing of APLP2 (Paper I: Fig. 3). The RA-induced sAPLP2 secretion was also blocked by TACE knock-down. Interestingly, we could also detect an increased phosphorylation of TACE in response to IGF-1 using 32P- phosphate labeling followed by immunoprecipitation with a TACE antibody. This increased phosphorylation was also shown to be dependent on PKC, since it was blocked using the PKC inhibitor (Paper I: Fig. 5). There are

44 contradictory reports regarding the connection between TACE activity and phosphorylation (see section 1.5.2.1), but based on our results we speculate that the increased TACE activity during IGF-1 stimulation is dependent on TACE phosphorylation.

In addition to TACE phosphorylation, we also wanted to study the maturation of TACE (i.e., its processing) during stimulated conditions. Similar to ADAM10, the protein levels of TACE were also shown to be up-regulated in response to 6 days exposure of RA (Paper II: Fig. 1). However, it was mainly the immature form of TACE that was increased, although the mature form was increased as well, but not to the same extent. The PI3K inhibitor had no effect on the level of neither immature nor mature TACE. Inhibition of PKC had a small effect on the immature level of TACE, but more importantly it strongly reduced the level of mature TACE (Paper II: Fig. 2 and table 1). Our result suggest that PKC is involved in stabilizing the mature form of TACE in response to RA, perhaps through regulation of TACE phosphorylation, leading to increased APLP2 processing.

In conclusion, we have used two physiological stimuli, RA and IGF-1 to induce α-secretase processing of APP and APLP2 in a human neuroblastoma cell line. Our results suggest that stimulated APP processing is PI3K- dependent and is mediated by ADAM10, whereas APLP2 processing is PKC-dependent and mediated by TACE during these relevant conditions.

4.3 O-GlcNAcylation induces α-secretase processing of APP but not of APLP2 (Paper III and IV) There are many studies on APP phosphorylation, but the reported effects on APP processing are contradictory. Many studies have introduced a mutation to evaluate the effect of phosphorylation at a specific site. However, what has not been considered in these studies is the possibility that another post- translational modification, namely O-GlcNAcylation also can occur at the same site as phosphorylation. O-GlcNAcylation is also a rapid modification that cycle on and off the protein (see section 1.5.2). Many cytosolic and nuclear proteins have been shown to be O-GlcNAcylated, thereby affecting their activity and/or localization. APP was the first membrane protein reported to be O-GlcNAcylated (Griffith, et al., 1995), but the consequence of this has not been investigated. We aimed to determine if APP localization was affected by O-GlcNAcylation, and since our research mainly concerns the α- secretase processing we focused on the effect on plasma membrane localization where the main part of α-site processing occurs. We treated SH-SY5Y cells with PUGNAc, an inhibitor of OGA-the enzyme that catalyzes the re-

45 moval of O-GlcNAc from target protein, thereby increasing the amount of O-GlcNAcylated proteins in the cell. Immunoprecipitation followed by western blot using antibodies against APP demonstrated that the level of APP that precipitated with an O-GlcNAc antibody increased in response to PUGNAc (Paper III: Fig. 1). The levels of APP at the cell surface were analyzed by a biotinylation assay followed by western blot. The results show that cell surface localization of mature APP was greatly enhanced in response to PUGNAc without affecting the total level of APP (Paper IV: Fig. 2). This indicates that O-GlcNAcylation affects the trafficking of APP targeting it to the cell surface.

Based on the observed PUGNAc-induced cell surface localization of APP, we speculated that the degree of its α-secretase processing was affected. To test this we treated cells with PUGNAc and analyzed by western blot the secretion of sAPPα in conditioned cell medium. Indeed, PUGNAc increased the sAPPα secretion with approximately 150% and reduced the Aβ secretion (Paper III: Fig. 2), suggesting that O-GlcNAcylation induces a shift in APP processing towards increased α-secretase processing through enhanced cell surface localization. Interestingly, when analyzing the conditioned cell medium, we found that the sAPLP2 secretion was unaffected by PUGNAc (Pa- per IV: Fig. 1). Thus, our result shows that it is possible to induce α- secretase processing of APP without affecting the processing of APLP2.

In addition to the effects on α-secretase processing, it has further been shown that O-GlcNAcylation of the γ-secretase component nicastrin regulates Aβ production and inhibition of OGA reduced plaque formation and rescued memory impairment in an AD mouse model (Kim et al., 2013). OGA and OGT are both enriched and highly active in the brain, and most abundant in the hippocampus, a region important for learning and memory. Interestingly, LTP was shown to be enhanced in vivo upon inhibition of OGA (Tallent, et al., 2009). The study further demonstrated that the effect was meditated through O-GlcNAcylation of several pre-synaptic proteins. O- GlcNAcylation has also been directly linked to AD since the level of protein O-GlcNAcylation is decreased by approximately 20% in AD brains (Liu et al., 2004a). Furthermore, it has been demonstrated that decreased O- GlcNAcylation of tau is correlated to tau hyperphosphorylation and thereby increased formation of neurofibrillary tangles (Liu, et al., 2004a).

4.3.1 O-GlcNAcylation selectively enhances α-secretase processing of APP in neuron-like cells (Paper IV) We also analyzed the effect of O-GlcNAcylation on APP processing in different cell lines. In addition to SH-SY5Y cells, we used the human neuro-

46 regarding the enzyme selectivity between APP and APLP2 and the fact that only APP and not APLP2 processing was affected by O-GlcNAcylation, we speculated that perhaps ADAM10 was O-GlcNAcylated. Indeed, we could demonstrate that the level of ADAM10 that was precipitated with sWGA or with an O-GlcNAc antibody was increased in response to PUGNAc. When pre-treating the cell lysate with SDS (breaking up complexes), ADAM10 but not APP is still precipitated with the O-GlcNAc antibody (Paper IV: Fig. 5), suggesting that ADAM10 is in fact O-GlcNAcylated whereas APP is not. This will be further investigated using mass spectrometry. Together with our previous results we now speculated that O-GlcNAcylation is a neuron- specific modification of ADAM10 resulting in increased interaction and subsequent α-site cleavage of APP. Further studies showed however that also this hypothesis had its faults, since ADAM10 was shown to be O- GlcNAcylated in HEK293 cells as well (Paper IV: Fig 6), which do not result in increased α-site processing. In addition, the level of cell surface APP was unchanged in HEK293 cells in response to PUGNAc (Paper IV: Fig. 6). At this point, the mechanism of O-GlcNAc-regulated APP processing is still unknown, but our results indicate that it involves a neuron-specific protein that either stabilizes APP at the cell surface or increases it transport to the cell surface.

Together our result suggest that although APP is not O-GlcNAcylated, an increase of O-GlcNAcylation will affect its trafficking towards the cell surface where it will be cleaved at the α-site, resulting in a shift in its processing generating more sAPPα and less Aβ. ADAM10 is O-GlcNAcylated, but this is not sufficient for increased α-secretase processing of APP. We also show that O-GlcNAcylation only affects APP processing in neuron-like cells without affecting APLP2 processing, which is important in the search for therapeutic treatment aimed to selectively modify APP processing in the brain.

4.4 The difference in regulation of α-secretase processing between APP and APLP2 is determined by their E1 domain (Paper III) Having established a clear difference in the regulation of α-site processing between APP and APLP2, including cleavage by different enzymes, we sought to determine which part of these two homologous proteins defines their selectivity for either ADAM10 or TACE, respectively. We started by constructing an APP/E1/APLP2 chimer, in which the entire E1 domain in APP was substituted with the E1 domain from APLP2. We then analyzed the

48 processing of the chimeric protein in response to IGF-1. The secretion of the chimeric protein was enhanced upon IGF-1 stimulation (Paper V: Fig. I), as seen for both APP and APLP2. However, in contrast to APP, this increased secretion of sAPP/E1/APLP2 was unaffected by PI3K inhibition (Paper V: Fig. 2), as previously shown for APLP2. We further investigated the IGF-1- induced processing of APP/E1/APLP2 showing that it was PKC-dependent (Paper V: Fig. 3), as seen for APLP2. It should be noted that this experiment only has been performed once and needs to be repeated to draw any statisti- cally reliable conclusions. Our results indicate that the chimer is cleaved by TACE, and not by ADAM10. Since our results show that the chimeric protein behaves more like APLP2 during IGF-1 stimulation, we wanted to investigate if this would be the case during increased O-GlcNAcylation. We treated transfected cells with PUGNAc and in accordance to what was shown for IGF-1, the processing of the chimeric protein was similar to APLP2 processing; i.e., it was unaffected (Paper V: Fig. 4). Also this experiment needs to be repeated. Our results indicate that differences in the APP and APLP2 E1 domain determine their selectivity for ADAM10 and TACE, respectively. It should be noted that the E1 domain of APP and APLP2 are 62% identical on amino acid level and 82% similar, i.e., they are very homologous. It has previously been shown that the secondary structure of the substrate is important for ADAM10 and TACE recognition and plays a role in the substrate selectivity of the two enzymes (Stawikowska, et al., 2013). We therefore speculate that the E1 domain of APP might differ in its secondary structure, targeting them to either enzyme. Another possibility could be a difference in the loop sequence in the proteins. This loop in APP, created by a disulfide bridge between Cys98 and Cys105 (Cys116 and Cys123 in APLP2) contains mostly positively charged amino acids, except one glycine (Fig. 5). In the corresponding loop in APLP2, there is an Asp instead of the glycine, introducing a negative charge, which we speculate might be important for the selective TACE interaction and subsequent cleavage. It would be interesting to perform a point mutation at this site and analyze the stimulated processing.

50 APP SDPSGTKTCIDTKEGILQYCQEVYPELQITNVVEANQPVTIQNWCKRGRKQCKTHPHFVIPYRCLVGEFVSDALLVPDKCKFLHQERMDVCETHLH

DP+GTK+C +TKE +LQYCQE+YPELQITNV+EANQ V+I NWC+R +KQCK+ FV P++CLVGEFVSD LLVP+KC+F H+ERM+VCE H H

68 APLP2 PDPTGTKSCFETKEEVLQYCQEMYPELQITNVMEANQRVSIDNWCRRDKKQCKS— RFVTPFKCLVGEFVSDVLLVPEKCQFFHKERMEVCENHQH

Figure 5. Alignment of a sequence in the E1 domain of APP695 with the corresponding sequence of APLP2 763. The cysteine loop is indicated with enlarged characters.

5. Conclusions

In our studies, we have demonstrated that;

 Both RA- and IGF-1-induced processing of APLP2 is mediated by TACE in a PKC-dependent manner, while the stimulated processing of APP is mediated by ADAM10 in a PI3K-dependent manner. We further show that RA up-regulated the mRNA and protein levels of ADAM10 and TACE, as previously seen for the APP family (Beckman and Iverfeldt, 1997). The maturation of ADAM10 and TACE during RA treatment was PI3K and PKC dependent, respectively.

 Enhancing cellular O-GlcNAcylation through pharmacological inhibition or siRNA knock-down of OGA increases the α-site processing of APP, concomitantly with decreased Aβ production. In addition, we show that APLP2 processing is unaffected by increased O-GlcNAcylation. We have further shown that APP is not O-GlcNAcylated and that the mechanism rather involves O- GlcNAcylation of ADAM10 as well as increased cell surface localization of APP, possibly mediated through O-GlcNAcylation of an unknown neuron-specific protein.

 The E1 domains determine the difference in α-secretase processing between APP and APLP2 during certain stimulated conditions. This was demonstrated by creating a chimeric APP/APLP2 construct.

Insulin receptor signaling has been shown to involve PI3K-dependent OGT translocation from the nucleus to the plasma membrane, where it O-GlcNAc- modifies Akt and IRS-1, thereby regulating their phosphorylation status, resulting in attenuated insulin signaling (Yang et al., 2008). Based on this, it is not unlikely that the effect of IGF-1 on APP processing is dependent on OGT and subsequent O-GlcNAcylation of ADAM10 and the unknown neuron-specific protein that regulates APP trafficking. In addition to regulating OGT translocation, we also show that PI3K regulates ADAM10 maturation. Although the mechanism behind this is still unclear, it does not seem to involve O-GlcNAcylation. APLP2 processing is, however, independent of PI3K and O-GlcNAcylation. This mechanism rather involves PKC, leading

51 to phosphorylation of TACE (not yet investigated during RA stimulation), which possibly affects its subcellular localization and/or maturation. Alt- hough ADAM10 today is considered the constitutive α-secretase for APP and APLP2 (Hogl, et al., 2011, Kuhn, et al., 2010), little is known about the mechanism involved during stimulated conditions. Our studies indicate that TACE preferably cleaves APLP2 during conditions when the activity of both enzymes clearly is increased. This could be due to many different factors, including the possibilities that APLP2 and TACE are spatially segregated during non-stimulated conditions, or that APLP2 during certain stimulations undergoes conformational changes (in the E1 domain) making it a better substrate for TACE.

The overall conclusion of our studies presented in this thesis is that the α- secretase processing of APP and APLP2 display clear divergences during stimulated conditions (Fig. 7). This includes different signaling pathways and processing enzymes. Moreover, our results provide insight to what could determine these differences in regulation. These finding may be important for a general understanding of the regulation of ectodomain shedding; the initial and rate-limiting step of RIP signaling (se section 1.2.3). Many proteins, including growth factors, cytokines, cell adhesion proteins and receptors undergo RIP. Regulation of this form of signal transduction thereby controls several important cellular functions. However, little is known about substrate recognition and specificity of RIP proteases, which are important factors when aiming to selectively modify the processing of a specific protein. More specifically our results contribute to the understanding of the normal function of APP/APLP2 and how their processing may be regulated. This might be useful in the devolvement of selective enhancers of APP α- secretase processing that does not affect the processing of the physiologically important homolog APLP2.

Figure 7. Schematic illustration con- cluding the results presented in this thesis. During stimulated conditions a PI3K- and a PKC-dependent pathway is activated. The PI3K pathway results in OGT mediated ADAM10 O- GlcNAcylation and APP cleavage, generating sAPPα. The PKC pathway leads to TACE phosphorylation and APLP2 cleavage, generating sAPLP2.

52 6. Populärvetenskaplig sammanfattning på svenska

Alzheimers sjukdom är en demenssjukdom som ca 70 000 människor lider av i Sverige. Sjukdomen innebär att nervceller i specifika delar av hjärnan dör, vilket resulterar i symptom som t.ex. problem med minnesfunktioner, nedsatt uppmärksamhetsförmåga och språkförståelse, svårare att uttrycka sig, sämre lokalsinne samt svårare att utföra ändamålsenliga kroppsrörelser. Det är ännu inte känt vad som orsakar Alzheimers sjukdom, men hög ålder är den i särklass mest bidragande riskfaktorn. Det som dödar nervcellerna i hjärnan är framförallt ett ämne som kallas amyloid-β (Aβ). Aβ finns normalt i hjärnan och bildas genom att ett större protein, APP klyvs av specifika enzymer. Normalt sätt så klyvs APP till största del av ett annat enzym, som hindrar bildningen av Aβ. Denna klyvning kallas α-klyvning och de enzymer som medierar detta kallas α-secretaser. Vid Alzheimers sjukdom bildas av någon anledning Aβ till mycket större grad och α-klyvningen minskar.

Förutom dess bidragande roll i Alzheimer, så har även APP många viktiga funktioner i hjärnan, vilket inte har studerats lika väl. APP ingår i en s.k familj utav liknande proteiner som även innefattar de två homologa proteinerna APLP1 och APLP2. Forskning har visat att det finns ett samspel mellan dessa tre proteiner där APLP2 spelar en särskilt stor roll. En möjlig strategi för att utveckla ett läkemedel som skulle skydda mot Alzheimers sjukdom är att stimulera α-klyvningen av APP. För att minimera oönskade effekter på APLP2, som klyvs på ett likande sätt som APP, vore det att föredra om man selektivt kunde påverka α-klyvningen av APP.

I vår forskning som presenteras i denna avhandling har vi studerat α- klyvningen av APP och APLP2 i en human neuroblastoma cell-linje (SH- SY5Y). I våra studier har vi behandlat cellerna med IGF-1 och RA, två ämnen som tidigare visats öka α-klyvningen av både APP och APLP2 och som det dessutom är brist på vid Alzheimers sjukdom. Vi har sett att den stimulerade α-klyvningen av APP och APLP2 är reglerad via olika signaleringsvägar i cellen och att de även klyvs av olika α-secretaser. APP klyvs av ADAM10 och APLP2 av TACE. Våra studier visar även att nivåerna av ADAM10 och TACE är beroende av olika signaleringsvägar.

53 En annan del av vår forskning har fokuserat på O-GlcNAcylering, en process där ett sorts socker kopplas på proteiner vilket påverkar deras funktion. Även O-GlcNAcylering har visats minska i hjärnan hos patienter med Alzheimers sjukdom. Vi har sett att om man ökar O-GlcNAcyleringen i vårt modellsystem, så ökar α-klyvningen av APP, men påverkar inte klyvningen av APLP2. Vi visar även att det är ADAM10 och inte APP som O- GlcNAcyleras.

För att förstå varför dessa två liknande proteiner, APP och APLP2 uppvisar en skillnad i hur och när de klyvs vid deras α-site har vi klonat fram ett chimert protein där en del av APP ersatts av den motsvarande delen av APLP2. Poängen var att få APP α-klyvningen att bete sig som α-klyvningen för APLP2. Det visade sig vara den mest N-terminala delen av proteinerna, den s.k. E1 domänen som avgjorde skillnanden i α-klyvning. Tillsammans visar dessa studier att α-klyvningen av APP och APLP2 är reglerad genom distinkta mekanismer och därmed kan särskiljas.

54 7. Acknowledgements

Det har gått nästan 6 år sen jag började här och nu helt plötsligt så är den här tiden nästan över. Det är sjukt hur snabbt tiden går. När jag började tänkte jag att 5 år var hur lång tid som helst, och jag skulle vara 30 år när jag var klar-alltså jätte gammal! Nu blev det inte så, jag är ännu äldre. Vet inte hur många gånger jag funderat på att sluta, speciellt nu i slutet för att undvika disputationen  Anledningen till att man stannat kvar har till stor del varit en massa härliga männsikor som jag härmed vill tacka:

Först och främst och mest av allt, ett stort tack till Kerstin Iverfeldt. Tack för att du antog mig som doktorand och för att du trott på mig under alla dessa år. Tack för alla givande och roliga diskussioner och för att du lyssnar och hjälper när det inte går så bra eller man inte mår så bra. Du är en fantastisk handledare som gett mig frihet men även stöd och hjälp när jag behövt det. Tack för att du alltid tar dig tid, även när du har fullt upp.

Ett stort tack även till alla i gruppen. Tack Sofia Holback för handledning redan som D-kursare men även under min första tid här. Du har alltid varit en inspirationskälla. Tack Tom Gatsinzi för att du är en så himla glad och nyfiken kille, det har varit kul att retas med dig. Hoppas vi kan jobba tillsammans igen snart. Linda Tracy min vapendragare under alla dessa år som jag delat många tårar och glädjestunder med. Hade inte klarat det här utan dig. Vi började samma dag, och slutar nästan samma dag. Tack för att du alltid lyssnar och bryr dig. Det har varit så himla roligt att dela kontor och lab med dig. Tack även till min vän och medförfattare Niina Koistinen du är så mycket bättre än du någonsin kommer våga tro. Tack för alla tidiga frukostar, träningspass, pratstunder och hjälp med projekten. Jag kommer verkligen sakna att jobba med dig. Thank you Elena Ivanova your passion for science is inspiring! Ylva Strååt min O-GlcNAc kollega; det har varit roligt att jobba och bolla idéer med dig. Anna Edlund det känns jätte roligt att du börjat i gruppen.

Ett vamt tack går såklart även till alla mina fina rumskamrater i ”tjejrummet”. Christin Svensson som nog har skrattat och gråtit lika mycket som jag inne på kontoret och Marie Danielsson som var en lugn punkt i stormen-jag har saknat er! Tack Jessica Lundqvist för att du är en sån fantastisk människa som alltid ställer upp. Tack även för moraliskt stöd

55 under skrivandet av avhandlingen och all hjälp med korrekturläsningen. Kristina Attoff som har samma humor som jag-jätte rolig alltså! Det har varit så roligt och mysigt att dela kontor med er alla.

Thank you to all PhD students at the department; it has been a joy to work with all of you. Thanks for all the fun Christmas parties, summer barbeques and other festivities that we have had at the department during these years. You are an amazing bunch of people and it feels great that so many of you are dedicated not only to your own research but also for making this department an even better place to work at. A special thanks to: Kristin Webling för att du är världens gulligaste, Staffan Lindberg för alla dryga kommentarer , Andrés Muñoz-Alarcón för att du alltid är glad och för att du kämpar för oss doktorander i din roll som doktorandordförande, Rania Abdo för att du en trevlig tjej med skinn på näsan, Henrik Helmfors för att du aldrig skrattar åt mina ”problem” med datorn och för att du är en himla snäll kille, Veronica Larsson för att du är en väldigt snäll och stark tjej, Santhosh Gudise for alway being so happy, Xin Yu for your friendship-it is amazing everything that you have overcomed and accomplished in your life.

Ett stort tack till alla lärare, Ülo Langel, Anders Undén, Anna Forsby, Einar Hallberg and Anna-Lena Ström, för att ni engagerar er i alla oss doktorander och hjälper oss på vägen. Ett speciellt tack till Anna-Lena för att du styrde upp Tisdags seminarierna och för att du korrekturläste min avhandling. Tack även till Bengt Mannervik för att du tar dig tid att kommenterar ens insatser på seminarierna.

Det är även många utanför institutionens väggar som bidragit, framförallt med moraliskt stöd under dessa år. Tack till mina kära föräldrar som alltit har trott på mig, jag skulle aldrig ens påbörjat det här utan ert stöd. Mina älskade systrar för att ni alltid finns där för mig, även fast ni alltid är så långt borta. Mina vänner, Monique, Katti, Märta, Jenny, Camilla och Linda, för er villkorslösa vänskap och för alla trevliga tjejmiddagar. Alla klätterkompisar, Lollo, Erik och Anton, som jag tyvärr inte träffar så ofta längre men som ändå betytt mycket under åren. Tack även till min nya familj-Molindrarna; Lars, Britt och Tomas för att ni alltid ställer upp när det behövs. Sist av allt, ett enormt tack till min man Anders, ord går inte att beskriva hur mycket jag uppskattar allt stöd och all kärlek som du gett mig denna tid. Jag skulle verkligen inte klarat det här utan dig. Jag älskar dig. Tack för att du gett mig en liten busunge!

56 8. References

Adlerz, L., Beckman, M., Holback, S., Tehranian, R., Cortes Toro, V., Iverfeldt, K., 2003. Accumulation of the amyloid precursor-like protein APLP2 and reduction of APLP1 in retinoic acid- differentiated human neuroblastoma cells upon curcumin-induced neurite retraction. Brain Res Mol Brain Res 119, 62-72. Adlerz, L., Holback, S., Multhaup, G., Iverfeldt, K., 2007. IGF-1-induced processing of the amyloid precursor protein family is mediated by different signaling pathways. J Biol Chem 282, 10203-10209. Adrain, C., Zettl, M., Christova, Y., Taylor, N., Freeman, M., 2012. Tumor necrosis factor signaling requires iRhom2 to promote trafficking and activation of TACE. Science 335, 225-228. Alfa Cisse, M., Sunyach, C., Slack, B.E., Fisher, A., Vincent, B., Checler, F., 2007. M1 and M3 muscarinic receptors control physiological processing of cellular prion by modulating ADAM17 phosphorylation and activity. J Neurosci 27, 4083-4092. Allinson, T.M., Parkin, E.T., Turner, A.J., Hooper, N.M., 2003. ADAMs family members as amyloid precursor protein alpha-secretases. J Neurosci Res 74, 342-352. Anders, A., Gilbert, S., Garten, W., Postina, R., Fahrenholz, F., 2001. Regulation of the -secretase ADAM10 by its prodomain and proprotein convertases. Faseb J 15, 1837-1839. Andersen, O.M., Reiche, J., Schmidt, V., Gotthardt, M., Spoelgen, R., Behlke, J., von Arnim, C.A., Breiderhoff, T., Jansen, P., Wu, X., Bales, K.R., Cappai, R., Masters, C.L., Gliemann, J., Mufson, E.J., Hyman, B.T., Paul, S.M., Nykjaer, A., Willnow, T.E., 2005. Neuronal sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proc Natl Acad Sci U S A 102, 13461-13466. Ando, K., Iijima, K., Elliott, J.I., Kirino, Y., Suzuki, T., 2001. Phosphorylation-dependent regulation of the interaction of amyloid precursor protein with Fe65 affects the production of -amyloid. J Biol Chem 276, 40353-40361. Aplin, A.E., Gibb, G.M., Jacobsen, J.S., Gallo, J.M., Anderton, B.H., 1996. In vitro phosphorylation of the cytoplasmic domain of the amyloid precursor protein by glycogen synthase kinase-3. J Neurochem 67, 699-707. Barnham, K.J., McKinstry, W.J., Multhaup, G., Galatis, D., Morton, C.J., Curtain, C.C., Williamson, N.A., White, A.R., Hinds, M.G., Norton, R.S., Beyreuther, K., Masters, C.L., Parker, M.W., Cappai, R., 2003.

57 Structure of the Alzheimer's disease amyloid precursor protein copper binding domain. A regulator of neuronal copper homeostasis. J Biol Chem 278, 17401-17407. Barrett, P.J., Song, Y., Van Horn, W.D., Hustedt, E.J., Schafer, J.M., Hadziselimovic, A., Beel, A.J., Sanders, C.R., 2012. The amyloid precursor protein has a flexible transmembrane domain and binds cholesterol. Science 336, 1168-1171. Beckman, M., Iverfeldt, K., 1997. Increased gene expression of -amyloid precursor protein and its homologues APLP1 and APLP2 in human neuroblastoma cells in response to retinoic acid. Neurosci Lett 221, 73-76. Belyaev, N.D., Kellett, K.A., Beckett, C., Makova, N.Z., Revett, T.J., Nalivaeva, N.N., Hooper, N.M., Turner, A.J., 2010. The transcriptionally active amyloid precursor protein (APP) intracellular domain is preferentially produced from the 695 isoform of APP in a {beta}-secretase-dependent pathway. J Biol Chem 285, 41443-41454. Berditchevski, F., Odintsova, E., 2007. Tetraspanins as regulators of protein trafficking. Traffic 8, 89-96. Biedler, J.L., Roffler-Tarlov, S., Schachner, M., Freedman, L.S., 1978. Multiple neurotransmitter synthesis by human neuroblastoma cell lines and clones. Cancer Res 38, 3751-3757. Black, R.A., Rauch, C.T., Kozlosky, C.J., Peschon, J.J., Slack, J.L., Wolfson, M.F., Castner, B.J., Stocking, K.L., Reddy, P., Srinivasan, S., Nelson, N., Boiani, N., Schooley, K.A., Gerhart, M., Davis, R., Fitzner, J.N., Johnson, R.S., Paxton, R.J., March, C.J., Cerretti, D.P., 1997. A metalloproteinase disintegrin that releases tumour-necrosis factor- from cells. Nature 385, 729-733. Borg, J.P., Yang, Y., De Taddéo-Borg, M., Margolis, B., Turner, R.S., 1998. The X11 protein slows cellular amyloid precursor protein processing and reduces A40 and A42 secretion. J Biol Chem 273, 14761-14766. Borroto, A., Ruiz-Paz, S., de la Torre, T.V., Borrell-Pages, M., Merlos- Suarez, A., Pandiella, A., Blobel, C.P., Baselga, J., Arribas, J., 2003. Impaired trafficking and activation of tumor necrosis factor-alpha- converting enzyme in cell mutants defective in protein ectodomain shedding. J Biol Chem 278, 25933-25939. Bozkulak, E.C., Weinmaster, G., 2009. Selective use of ADAM10 and ADAM17 in activation of Notch1 signaling. Mol Cell Biol 29, 5679- 5695. Buxbaum, J.D., Koo, E.H., Greengard, P., 1993. Protein phosphorylation inhibits production of Alzheimer amyloid /A4 peptide. Proc Natl Acad Sci U S A 90, 9195-9198. Buxbaum, J.D., Liu, K.N., Luo, Y., Slack, J.L., Stocking, K.L., Peschon, J.J., Johnson, R.S., Castner, B.J., Cerretti, D.P., Black, R.A., 1998. Evidence that tumor necrosis factor  converting enzyme is involved

58 in regulated -secretase cleavage of the Alzheimer amyloid protein precursor. J Biol Chem 273, 27765-27767. Caescu, C.I., Jeschke, G.R., Turk, B.E., 2009. Active-site determinants of substrate recognition by the metalloproteinases TACE and ADAM10. Biochem J 424, 79-88. Cam, J.A., Zerbinatti, C.V., Li, Y., Bu, G., 2005. Rapid endocytosis of the low density lipoprotein receptor-related protein modulates cell surface distribution and processing of the beta-amyloid precursor protein. J Biol Chem 280, 15464-15470. Cao, X., Südhof, T.C., 2001. A transcriptionally active complex of APP with Fe65 and histone acetyltransferase Tip60. Science 293, 115-120. Caporaso, G.L., Gandy, S.E., Buxbaum, J.D., Ramabhadran, T.V., Greengard, P., 1992. Protein phosphorylation rgeulates secretion of Alzheimer /A4 amyloid precursor protein. Proc Natl Acad Sci U S A 89, 3055-3059. Carey, R.M., Balcz, B.A., Lopez-Coviella, I., Slack, B.E., 2005. Inhibition of dynamin-dependent endocytosis increases shedding of the amyloid precursor protein ectodomain and reduces generation of amyloid beta protein. BMC Cell Biol 6, 30. Carey, R.M., Blusztajn, J.K., Slack, B.E., 2011. Surface expression and limited proteolysis of ADAM10 are increased by a dominant negative inhibitor of dynamin. BMC Cell Biol 12, 20. Carro, E., Torres-Aleman, I., 2004. The role of insulin and insulin-like growth factor I in the molecular and cellular mechanisms underlying the pathology of Alzheimer's disease. Eur J Pharmacol 490, 127- 133. Carro, E., Trejo, J.L., Gerber, A., Loetscher, H., Torrado, J., Metzger, F., Torres-Aleman, I., 2006. Therapeutic actions of insulin-like growth factor I on APP/PS2 mice with severe brain amyloidosis. Neurobiol Aging 27, 1250-1257. Carro, E., Trejo, J.L., Gomez-Isla, T., LeRoith, D., Torres-Aleman, I., 2002. Serum insulin-like growth factor I regulates brain amyloid-beta levels. Nat Med 8, 1390-1397. Chabrier, M.A., Blurton-Jones, M., Agazaryan, A.A., Nerhus, J.L., Martinez-Coria, H., LaFerla, F.M., 2012. Soluble abeta promotes wild-type tau pathology in vivo. J Neurosci 32, 17345-17350. Chattopadhyay, A., Carpenter, G., 2002. PLC-gamma1 is required for IGF-I protection from cell death induced by loss of extracellular matrix adhesion. J Cell Sci 115, 2233-2239. Chen, A.C., Guo, L.Y., Ostaszewski, B.L., Selkoe, D.J., LaVoie, M.J., 2010. Aph-1 associates directly with full-length and C-terminal fragments of gamma-secretase substrates. J Biol Chem 285, 11378-11391. Chyung, J.H., Selkoe, D.J., 2003. Inhibition of receptor-mediated endocytosis demonstrates generation of amyloid beta-protein at the cell surface. J Biol Chem 278, 51035-51043. Cisse, M.A., Sunyach, C., Lefranc-Jullien, S., Postina, R., Vincent, B., Checler, F., 2005. The disintegrin ADAM9 indirectly contributes to

59 the physiological processing of cellular prion by modulating ADAM10 activity. J Biol Chem 280, 40624-40631. Colombo, A., Wang, H., Kuhn, P.H., Page, R., Kremmer, E., Dempsey, P.J., Crawford, H.C., Lichtenthaler, S.F., 2012. Constitutive alpha- and beta-secretase cleavages of the amyloid precursor protein are partially coupled in neurons, but not in frequently used cell lines. Neurobiol Dis 49C, 137-147. Corcoran, J.P., So, P.L., Maden, M., 2004. Disruption of the retinoid signalling pathway causes a deposition of amyloid beta in the adult rat brain. Eur J Neurosci 20, 896-902. Dahms, S.O., Hoefgen, S., Roeser, D., Schlott, B., Guhrs, K.H., Than, M.E., 2010. Structure and biochemical analysis of the heparin-induced E1 dimer of the amyloid precursor protein. Proc Natl Acad Sci U S A 107, 5381-5386. Daigle, I., Li, C., 1993. apl-1, a Caenorhabditis elegans gene encoding a protein related to the human -amyloid protein precursor. Proc Natl Acad Sci U S A 90, 12045-12049. Das, U., Scott, D.A., Ganguly, A., Koo, E.H., Tang, Y., Roy, S., 2013. Activity-Induced Convergence of APP and BACE-1 in Acidic Microdomains via an Endocytosis-Dependent Pathway. Neuron 79, 447-460. Davies, S.P., Reddy, H., Caivano, M., Cohen, P., 2000. Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351, 95-105. Dawson, G.R., Seabrook, G.R., Zheng, H., Smith, D.W., Graham, S., O'Dowd, G., Bowery, B.J., Boyce, S., Trumbauer, M.E., Chen, H.Y., Van der Ploeg, L.H., Sirinathsinghji, D.J., 1999. Age-related cognitive deficits, impaired long-term potentiation and reduction in synaptic marker density in mice lacking the beta-amyloid precursor protein. Neuroscience 90, 1-13. De Strooper, B., Iwatsubo, T., Wolfe, M.S., 2012. Presenilins and gamma- secretase: structure, function, and role in Alzheimer Disease. Cold Spring Harb Perspect Med 2, a006304. Deuss, M., Reiss, K., Hartmann, D., 2008. Part-time alpha-secretases: the functional biology of ADAM 9, 10 and 17. Curr Alzheimer Res 5, 187-201. Dong, D.L., Hart, G.W., 1994. Purification and characterization of an O- GlcNAc selective N-acetyl-beta-D-glucosaminidase from rat spleen cytosol. J Biol Chem 269, 19321-19330. Dornier, E., Coumailleau, F., Ottavi, J.F., Moretti, J., Boucheix, C., Mauduit, P., Schweisguth, F., Rubinstein, E., 2012. TspanC8 tetraspanins regulate ADAM10/Kuzbanian trafficking and promote Notch activation in flies and mammals. J Cell Biol 199, 481-496. Dries, D.R., Yu, G., 2008. Assembly, maturation, and trafficking of the gamma-secretase complex in Alzheimer's disease. Curr Alzheimer Res 5, 132-146.

60 Duyckaerts, C., Delatour, B., Potier, M.C., 2009. Classification and basic pathology of Alzheimer disease. Acta Neuropathol 118, 5-36. Dyrks, T., Weidemann, A., Multhaup, G., Salbaum, J.M., Lemaire, H.G., Kang, J., Müller-Hill, B., Masters, C.L., Beyreuther, K., 1988. Identification, transmembrane orientation and biogenesis of the amyloid A4 precursor of Alzheimer's disease. EMBO J 7, 949-957. Edbauer, D., Winkler, E., Regula, J.T., Pesold, B., Steiner, H., Haass, C., 2003. Reconstitution of -secretase activity. Nat Cell Biol 5, 486- 488. Edwards, D.R., Handsley, M.M., Pennington, C.J., 2008. The ADAM metalloproteinases. Mol Aspects Med 29, 258-289. Eggert, S., Midthune, B., Cottrell, B., Koo, E.H., 2009. Induced dimerization of the amyloid precursor protein leads to decreased amyloid-beta protein production. J Biol Chem 284, 28943-28952. Eggert, S., Paliga, K., Soba, P., Evin, G., Masters, C.L., Weidemann, A., Beyreuther, K., 2004. The proteolytic processing of the amyloid precursor protein gene family members APLP-1 and APLP-2 involves -, -, -, and -like cleavages: modulation of APLP-1 processing by N-glycosylation. J Biol Chem 279, 18146-18156. Ehehalt, R., Keller, P., Haass, C., Thiele, C., Simons, K., 2003. Amyloidogenic processing of the Alzheimer -amyloid precursor protein depends on lipid rafts. J Cell Biol 160, 113-123. Encinas, M., Iglesias, M., Liu, Y., Wang, H., Muhaisen, A., Ceña, V., Gallego, C., Comella, J.X., 2000. Sequential treatment of SH-SY5Y cells with retinoic acid and brain-derived neurotrophic factor gives rise to fully differentiated, neurotrophic factor-dependent, human neuron-like cells. J Neurochem 75, 991-1003. Endres, K., Postina, R., Schroeder, A., Mueller, U., Fahrenholz, F., 2005. Shedding of the amyloid precursor protein-like protein APLP2 by disintegrin-metalloproteinases. FEBS J 272, 5808-5820. Epis, R., Marcello, E., Gardoni, F., Vastagh, C., Malinverno, M., Balducci, C., Colombo, A., Borroni, B., Vara, H., Dell'Agli, M., Cattabeni, F., Giustetto, M., Borsello, T., Forloni, G., Padovani, A., Di Luca, M., 2010. Blocking ADAM10 synaptic trafficking generates a model of sporadic Alzheimer's disease. Brain 133, 3323-3335. Esch, F.S., Keim, P.S., Beattie, E.C., Blacher, R.W., Culwell, A.R., Oltersdorf, T., McClure, D., Ward, P.J., 1990. Cleavage of amyloid-  peptide during constitutive processing of its precursor. Science 248, 1122-1124. Escrevente, C., Morais, V.A., Keller, S., Soares, C.M., Altevogt, P., Costa, J., 2008. Functional role of N-glycosylation from ADAM10 in processing, localization and activity of the enzyme. Biochim Biophys Acta 1780, 905-913. Finder, V.H., Glockshuber, R., 2007. Amyloid-beta aggregation. Neurodegener Dis 4, 13-27. Fjorback, A.W., Seaman, M., Gustafsen, C., Mehmedbasic, A., Gokool, S., Wu, C., Militz, D., Schmidt, V., Madsen, P., Nyengaard, J.R.,

61 Willnow, T.E., Christensen, E.I., Mobley, W.B., Nykjaer, A., Andersen, O.M., 2012. Retromer binds the FANSHY sorting motif in SorLA to regulate amyloid precursor protein sorting and processing. J Neurosci 32, 1467-1480. Flammang, B., Pardossi-Piquard, R., Sevalle, J., Debayle, D., Dabert-Gay, A.S., Thevenet, A., Lauritzen, I., Checler, F., 2012. Evidence that the amyloid-beta protein precursor intracellular domain, AICD, derives from beta-secretase-generated C-terminal fragment. J Alzheimers Dis 30, 145-153. Francis, R., McGrath, G., Zhang, J., Ruddy, D.A., Sym, M., Apfeld, J., Nicoll, M., Maxwell, M., Hai, B., Ellis, M.C., Parks, A.L., Xu, W., Li, J., Gurney, M., Myers, R.L., Himes, C.S., Hiebsch, R., Ruble, C., Nye, J.S., Curtis, D., 2002. aph-1 and pen-2 are required for Notch pathway signaling, gamma-secretase cleavage of betaAPP, and presenilin protein accumulation. Dev Cell 3, 85-97. Fukumori, A., Okochi, M., Tagami, S., Jiang, J., Itoh, N., Nakayama, T., Yanagida, K., Ishizuka-Katsura, Y., Morihara, T., Kamino, K., Tanaka, T., Kudo, T., Tanii, H., Ikuta, A., Haass, C., Takeda, M., 2006. Presenilin-dependent gamma-secretase on plasma membrane and endosomes is functionally distinct. Biochemistry 45, 4907-4914. Gandy, S., Czernik, A.J., Greengard, P., 1988. Phosphorylation of Alzheimer disease amyloid precursor peptide by protein kinase C and Ca2+/calmodulin-dependent protein kinase II. Proc Natl Acad Sci U S A 85, 6218-6221. Gervais, F.G., Xu, D., Robertson, G.S., Vaillancourt, J.P., Zhu, Y., Huang, J., LeBlanc, A., Smith, D., Rigby, M., Shearman, M.S., Clarke, E.E., Zheng, H., Van Der Ploeg, L.H., Ruffolo, S.C., Thornberry, N.A., Xanthoudakis, S., Zamboni, R.J., Roy, S., Nicholson, D.W., 1999. Involvement of caspases in proteolytic cleavage of Alzheimer's amyloid-beta precursor protein and amyloidogenic A beta peptide formation. Cell 97, 395-406. Goodman, Y., Mattson, M.P., 1994. Secreted forms of beta-amyloid precursor protein protect hippocampal neurons against amyloid beta- peptide-induced oxidative injury. Exp Neurol 128, 1-12. Gooz, M., 2010. ADAM-17: the enzyme that does it all. Crit Rev Biochem Mol Biol 45, 146-169. Goutte, C., Hepler, W., Mickey, K.M., Priess, J.R., 2000. aph-2 encodes a novel extracellular protein required for GLP-1-mediated signaling. Development 127, 2481-2492. Graham, F.L., Smiley, J., Russell, W.C., Nairn, R., 1977. Characteristics of a human cell line transformed by DNA from human adenovirus type 5. J Gen Virol 36, 59-74. Gralle, M., Ferreira, S.T., 2007. Structure and functions of the human amyloid precursor protein: The whole is more than the sum of its parts. Prog Neurobiol 82, 11-32.

62 Griffith, L.S., Mathes, M., Schmitz, B., 1995. Beta-amyloid precursor protein is modified with O-linked N-acetylglucosamine. J Neurosci Res 41, 270-278. Groves, J.A., Lee, A., Yildirir, G., Zachara, N.E., 2013. Dynamic O- GlcNAcylation and its roles in the cellular stress response and homeostasis. Cell Stress Chaperones 18, 535-558. Gu, Y., Misonou, H., Sato, T., Dohmae, N., Takio, K., Ihara, Y., 2001. Distinct intramembrane cleavage of the -amyloid precursor protein family resembling -secretase-like cleavage of Notch. J Biol Chem 276, 35235-35238. Gustafsen, C., Glerup, S., Pallesen, L.T., Olsen, D., Andersen, O.M., Nykjaer, A., Madsen, P., Petersen, C.M., 2013. Sortilin and SorLA display distinct roles in processing and trafficking of amyloid precursor protein. J Neurosci 33, 64-71. Haass, C., Koo, E.H., Mellon, A., Hung, A.Y., Selkoe, D.J., 1992. Targeting of cell-surface beta-amyloid precursor protein to lysosomes: alternative processing into amyloid-bearing fragments. Nature 357, 500-503. Haltiwanger, R.S., Blomberg, M.A., Hart, G.W., 1992. Glycosylation of nuclear and cytoplasmic proteins. Purification and characterization of a uridine diphospho-N-acetylglucosamine:polypeptide beta-N- acetylglucosaminyltransferase. J Biol Chem 267, 9005-9013. Hanover, J.A., Krause, M.W., Love, D.C., 2012. Bittersweet memories: linking metabolism to epigenetics through O-GlcNAcylation. Nat Rev Mol Cell Biol 13, 312-321. Hardy, J., 2009. The amyloid hypothesis for Alzheimer's disease: a critical reappraisal. J Neurochem 110, 1129-1134. Hardy, J.A., Higgins, G.A., 1992. Alzheimer's disease: the amyloid cascade hypothesis. Science 256, 184-185. Hart, G.W., Slawson, C., Ramirez-Correa, G., Lagerlof, O., 2011. Cross talk between O-GlcNAcylation and phosphorylation: roles in signaling, transcription, and chronic disease. Annu Rev Biochem 80, 825-858. Heber, S., Herms, J., Gajic, V., Hainfellner, J., Aguzzi, A., Rülicke, T., von Kretzschmar, H., von Koch, C., Sisodia, S., Tremml, P., Lipp, H.P., Wolfer, D.P., Müller, U., 2000. Mice with combined gene knock- outs reveal essential and partially redundant functions of amyloid precursor protein family members. J Neurosci 20, 7951-7963. Herms, J., Anliker, B., Heber, S., Ring, S., Fuhrmann, M., Kretzschmar, H., Sisodia, S., Müller, U., 2004. Cortical dysplasia resembling human type 2 lissencephaly in mice lacking all three APP family members. EMBO J 23, 4106-4115. Hikita, A., Tanaka, N., Yamane, S., Ikeda, Y., Furukawa, H., Tohma, S., Suzuki, R., Tanaka, S., Mitomi, H., Fukui, N., 2009. Involvement of a disintegrin and metalloproteinase 10 and 17 in shedding of tumor necrosis factor-alpha. Biochem Cell Biol 87, 581-593. Hogl, S., Kuhn, P.H., Colombo, A., Lichtenthaler, S.F., 2011. Determination of the proteolytic cleavage sites of the amyloid precursor-like

63 protein 2 by the proteases ADAM10, BACE1 and gamma-secretase. PLoS One 6, e21337. Holback, S., Adlerz, L., Gatsinzi, T., Jacobsen, K.T., Iverfeldt, K., 2007. PI3-K- and PKC-dependent up-regulation of APP processing enzymes by retinoic acid. Biochem Biophys Res Commun 365, 298- 303. Holback, S., Adlerz, L., Iverfeldt, K., 2005. Increased processing of APLP2 and APP with concomitant formation of APP intracellular domains in BDNF and retinoic acid-differentiated human neuroblastoma cells. J Neurochem 95, 1059-1068. Howard, L., Lu, X., Mitchell, S., Griffiths, S., Glynn, P., 1996. Molecular cloning of MADM: a catalytically active mammalian disintegrin- metalloprotease expressed in various cell types. Biochem J 317, 45- 50. Hoyos, B., Imam, A., Chua, R., Swenson, C., Tong, G.X., Levi, E., Noy, N., Hammerling, U., 2000. The cysteine-rich regions of the regulatory domains of Raf and protein kinase C as retinoid receptors. J Exp Med 192, 835-845. Hundhausen, C., Misztela, D., Berkhout, T.A., Broadway, N., Saftig, P., Reiss, K., Hartmann, D., Fahrenholz, F., Postina, R., Matthews, V., Kallen, K., Rose-John, S., Ludwig, A., 2003. The disintegrin-like metalloproteinaseADAM10 is involved in constitutive cleavage of CX3CL1 (fractalkine) and regulates CX3CL1-mediated cell-cell adhesion. BLOOD 102, 1186-1195. Hussain, I., Powell, D., Howlett, D.R., Tew, D.G., Meek, T.D., Chapman, C., Gloger, I.S., Murphy, K.E., Southan, C.D., Ryan, D.M., Smith, T.S., Simmons, D.L., Walsh, F.S., Dingwall, C., Christie, G., 1999. Identification of a novel aspartic protease (Asp 2) as -secretase. Mol Cell Neurosci 14, 419-427. Iijima, K., Ando, K., Takeda, S., Satoh, Y., Seki, T., Itohara, S., Greengard, P., Kirino, Y., Nairn, A.C., Suzuki, T., 2000. Neuron-specific phosphorylation of Alzheimer's -amyloid precursor protein by cyclin-dependent kinase 5. J Neurochem 75, 1085-1091. Jacobsen, K.T., Iverfeldt, K., 2009. Amyloid precursor protein and its homologues: a family of proteolysis-dependent receptors. Cell Mol Life Sci 66, 2299-2318. Kaden, D., Munter, L., Joshi, M., Treiber, C., Weise, C., Bethge, T., Voigt, P., Schaefer, M., Beyermann, M., Reif, B., Multhaup, G., 2008. Homophilic interactions of the amyloid precursor protein (APP) ectodomain are regulated by the loop region and affect -secretase cleavage of APP. J Biol Chem 283, 7271-7279. Kaden, D., Voigt, P., Munter, L.M., Bobowski, K.D., Schaefer, M., Multhaup, G., 2009. Subcellular localization and dimerization of APLP1 are strikingly different from APP and APLP2. J Cell Sci 122, 368-377. Kang, J., Lemaire, H.G., Unterbeck, A., Salbaum, J.M., Masters, C.L., Grzeschik, K.H., Multhaup, G., Beyreuther, K., Müller-Hill, B.,

64 1987. The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor. Nature 325, 733-736. Kayed, R., Lasagna-Reeves, C.A., 2013. Molecular mechanisms of amyloid oligomers toxicity. J Alzheimers Dis 33 Suppl 1, S67-78. Kim, C., Nam, D.W., Park, S.Y., Song, H., Hong, H.S., Boo, J.H., Jung, E.S., Kim, Y., Baek, J.Y., Kim, K.S., Cho, J.W., Mook-Jung, I., 2013. O-linked beta-N-acetylglucosaminidase inhibitor attenuates beta-amyloid plaque and rescues memory impairment. Neurobiol Aging 34, 275-285. Kinoshita, A., Whelan, C.M., Berezovska, O., Hyman, B.T., 2002. The  secretase carboxyl-terminal domain of the amyloid precursor protein induces apoptrosis via Top60 in H4 cells. J Biol Chem 277, 28530- 28536. Kins, S., Lauther, N., Szodorai, A., Beyreuther, K., 2006. Subcellular trafficking of the amyloid precursor protein gene family and its pathogenic role in Alzheimer's disease. Neurodegener Dis 3, 218- 226. Klyubin, I., Walsh, D.M., Lemere, C.A., Cullen, W.K., Shankar, G.M., Betts, V., Spooner, E.T., Jiang, L., Anwyl, R., Selkoe, D.J., Rowan, M.J., 2005. Amyloid beta protein immunotherapy neutralizes Abeta oligomers that disrupt synaptic plasticity in vivo. Nat Med 11, 556- 561. Knops, J., Gandy, S., Greengard, P., Lieberburg, I., Sinha, S., 1993. Serine Phosphorylation of the Secreted Extracellular Domain of APP. Biochem Biophys Res Commun 197, 380-385. Koike, H., Tomioka, S., Sorimachi, H., Saido, T.C., Maruyama, K., Okuyama, A., Fujisawa-Sehara, A., Ohno, S., Suzuki, K., Ishiura, S., 1999. Membrane-anchored metalloprotease MDC9 has an alpha- secretase activity responsible for processing the amyloid precursor protein. Biochem J 343, 371-375. Koike, M.A., Lin, A.J., Pham, J., Nguyen, E., Yeh, J.J., Rahimian, R., Tromberg, B.J., Choi, B., Green, K.N., LaFerla, F.M., 2012. APP knockout mice experience acute mortality as the result of ischemia. PLoS One 7, e42665. Kuhn, P.H., Wang, H., Dislich, B., Colombo, A., Zeitschel, U., Ellwart, J.W., Kremmer, E., Rossner, S., Lichtenthaler, S.F., 2010. ADAM10 is the physiologically relevant, constitutive alpha-secretase of the amyloid precursor protein in primary neurons. EMBO J 29, 3020- 3032. Lai, A., Sisodia, S.S., Trowbridge, I.S., 1995. Characterization of sorting signals in the beta-amyloid precursor protein cytoplasmic domain. J Biol Chem 270, 3565-3573. Lannfelt, L., Basun, H., Wahlund, L.O., Rowe, B.A., Wagner, S.L., 1995. Decreased alpha-secretase-cleaved amyloid precursor protein as a diagnostic marker for Alzheimer's disease. Nat Med 1, 829-832. Le Gall, S.M., Bobe, P., Reiss, K., Horiuchi, K., Niu, X.D., Lundell, D., Gibb, D.R., Conrad, D., Saftig, P., Blobel, C.P., 2009. ADAMs 10

65 and 17 represent differentially regulated components of a general shedding machinery for membrane proteins such as transforming growth factor alpha, L-selectin, and tumor necrosis factor alpha. Mol Biol Cell 20, 1785-1794. Lee, J.H., Lau, K.F., Perkinton, M.S., Standen, C.L., Shemilt, S.J.A., Mercken, L., Cooper, J.D., McLoughlin, D.M., Miller, C.C.J., 2003a. The neuronal adaptor protein X11 reduces A levels in the brains of Alzheimer's APPswe Tg2576 transgenic Mice. J Biol Chem 278, 47025-47029. Lee, M.S., Kao, S.C., Lemere, C.A., Xia, W., Tseng, H.C., Zhou, Y., Neve, R., Ahlijanian, M.K., Tsai, L.H., 2003b. APP processing is regulated by cytoplasmic phosphorylation. J Cell Biol 163, 83-95. Lee, S., Xue, Y., Hu, J., Wang, Y., Liu, X., Demeler, B., Ha, Y., 2011. The E2 domains of APP and APLP1 share a conserved mode of dimerization. Biochemistry 50, 5453-5464. Li, H., Wang, B., Wang, Z., Guo, Q., Tabuchi, K., Hammer, R.E., Sudhof, T.C., Zheng, H., 2010. Soluble amyloid precursor protein (APP) regulates transthyretin and Klotho gene expression without rescuing the essential function of APP. Proc Natl Acad Sci U S A 107, 17362-17367. Li, Q., Südhof, T.C., 2004. Cleavage of amyloid- precursor protein and amyloid- precursor-like protein by BACE 1. J Biol Chem 279, 10542-10550. Lichtenthaler, S.F., Haass, C., Steiner, H., 2011. Regulated intramembrane proteolysis--lessons from amyloid precursor protein processing. J Neurochem 117, 779-796. Liou, Y.C., Sun, A., Ryo, A., Zhou, X.Z., Yu, Z.X., Huang, H.K., Uchida, T., Bronson, R., Bing, G., Li, X., Hunter, T., Lu, K.P., 2003. Role of the isomerase Pin1 in protecting against age-dependent neurododegeneration. Nature 424, 556-561. Liu, F., Iqbal, K., Grundke-Iqbal, I., Hart, G.W., Gong, C.X., 2004a. O- GlcNAcylation regulates phosphorylation of tau: a mechanism involved in Alzheimer's disease. Proc Natl Acad Sci U S A 101, 10804-10809. Liu, K., Paterson, A.J., Zhang, F., McAndrew, J., Fukuchi, K., Wyss, J.M., Peng, L., Hu, Y., Kudlow, J.E., 2004b. Accumulation of protein O- GlcNAc modification inhibits proteasomes in the brain and coincides with neuronal apoptosis in brain areas with high O- GlcNAc metabolism. J Neurochem 89, 1044-1055. Lorent, K., Overbergh, L., Moechars, D., De Strooper, B., Van Leuven, F., Van den Berghe, H., 1995. Expression in mouse embryos and in adult mouse brain of three members of the amyloid precursor protein family, of the 2-macroglobulin receptor/low density lipoprotein receptor-related protein and of its ligands apolipoprotein E, lipoprotein lipase, 2-macroglobulin and the 40,000 molecular weight receptor-associated protein. Neuroscience 65, 1009-1025.

66 Lu, D.C., Rabizadeh, S., Chandra, S., Shayya, R.F., Ellerby, L.M., Ye, X., Salvesen, G.S., Koo, E.H., Bredesen, D.E., 2000. A second cytotoxic proteolytic peptide derived from amyloid beta-protein precursor. Nat Med 6, 397-404. Luo, Y., Bolon, B., Kahn, S., Bennett, B.D., Babu-Khan, S., Denis, P., Fan, W., Kha, H., Zhang, J., Gong, Y., Martin, L., Louis, J.C., Yan, Q., Richards, W.G., Citron, M., Vassar, R., 2001. Mice deficient in BACE1, the Alzheimer's -secretase, have normal phenotype and abolished -amyloid generation. Nat Neurosci 4, 231-232. Maden, M., 2007. Retinoic acid in the development, regeneration and maintenance of the nervous system. Nat Rev Neurosci 8, 755-765. Maniar, R., Pecherskaya, A., Ila, R., Solem, S., 2005. PKC alpha-dependent regulation of the IGF1 receptor in adult and embryonic rat cardiomyocytes. Mol Cell Biochem 275, 15-24. Marcello, E., Gardoni, F., Di Luca, M., Perez-Otano, I., 2010. An arginine stretch limits ADAM10 exit from the endoplasmic reticulum. J Biol Chem 285, 10376-10384. Marcello, E., Gardoni, F., Mauceri, D., Romorini, S., Jeromin, A., Epis, R., Borroni, B., Cattabeni, F., Sala, C., Padovani, A., Di Luca, M., 2007. Synapse-associated protein-97 mediates alpha-secretase ADAM10 trafficking and promotes its activity. J Neurosci 27, 1682-1691. Maretzky, T., Reiss, K., Ludwig, A., Buchholz, J., Scholz, F., Proksch, E., de Strooper, B., Hartmann, D., Saftig, P., 2005. ADAM10 mediates E-cadherin shedding and regulates epithelial cell-cell adhesion, migration, and beta-catenin translocation. Proc Natl Acad Sci U S A 102, 9182-9187. Masia, S., Alvarez, S., de Lera, A.R., Barettino, D., 2007. Rapid, nongenomic actions of retinoic acid on phosphatidylinositol-3- kinase signaling pathway mediated by the retinoic acid receptor. Mol Endocrinol 21, 2391-2402. Mattson, M.P., Cheng, B., Culwell, A.R., Esch, F.S., Lieberburg, I., Rydel, R.E., 1993. Evidence for excitoprotective and intraneuronal calcium- regulating roles for secreted forms of the -amyloid precursor protein. Neuron 10, 243-254. McCarthy, J.V., Twomey, C., Wujek, P., 2009. Presenilin-dependent regulated intramembrane proteolysis and gamma-secretase activity. Cell Mol Life Sci 66, 1534-1555. McFarlane, I., Georgopoulou, N., Coughlan, C.M., Gillian, A.M., Breen, K.C., 1999. The role of the protein glycosylation state in the control of cellular transport of the amyloid beta precursor protein. Neuroscience 90, 15-25. McLoughlin, D.M., Miller, C.C., 2008. The FE65 proteins and Alzheimer's disease. J Neurosci Res 86, 744-754. Merlos-Suarez, A., Ruiz-Paz, S., Baselga, J., Arribas, J., 2001. Metalloprotease-dependent protransforming growth factor-alpha ectodomain shedding in the absence of tumor necrosis factor-alpha- converting enzyme. J Biol Chem 276, 48510-48517.

67 Milward, E.A., Papadopoulos, R., Fuller, S.J., Moir, R.D., Small, D., Beyreuther, K., Masters, C.L., 1992. The amyloid protein precursor of Alzheimer's disease is a mediator of the effects of nerve growth factor on neurite outgrowth. Neuron 9, 129-137. Moss, M.L., Powell, G., Miller, M.A., Edwards, L., Qi, B., Sang, Q.X., De Strooper, B., Tesseur, I., Lichtenthaler, S.F., Taverna, M., Zhong, J.L., Dingwall, C., Ferdous, T., Schlomann, U., Zhou, P., Griffith, L.G., Lauffenburger, D.A., Petrovich, R., Bartsch, J.W., 2011. ADAM9 inhibition increases membrane activity of ADAM10 and controls alpha-secretase processing of amyloid precursor protein. J Biol Chem 286, 40443-40451. Muller, U., Cristina, N., Li, Z.W., Wolfer, D.P., Lipp, H.P., Rulicke, T., Brandner, S., Aguzzi, A., Weissmann, C., 1994. Behavioral and anatomical deficits in mice homozygous for a modified beta- amyloid precursor protein gene. Cell 79, 755-765. Munter, L.M., Voigt, P., Harmeier, A., Kaden, D., Gottschalk, K.E., Weise, C., Pipkorn, R., Schaefer, M., Langosch, D., Multhaup, G., 2007. GxxxG motifs within the amyloid precursor protein transmembrane sequence are critical for the etiology of Abeta42. EMBO J 26, 1702- 1712. Müller, T., Meyer, H.E., Egensperger, R., Marcus, K., 2008. The amyloid precursor protein intarcellular domain (AICD) as a modulator of gene transcription, apoptosis, and cytoskeletal dynamics-Relevance for Alzheimer's disease. Prog Neurobiol 85, 393-406. Nagano, O., Murakami, D., Hartmann, D., De Strooper, B., Saftig, P., Iwatsubo, T., Nakajima, M., Shinohara, M., Saya, H., 2004. Cell- matrix interaction via CD44 is independently regulated by different metalloproteinases activated in response to extracellular Ca(2+) influx and PKC activation. J Cell Biol 165, 893-902. Nakayama, K., Ohkawara, T., Hiratochi, M., Koh, C.S., Nagase, H., 2008. The intracellular domain of amyloid precursor protein induces neuron-specific apoptosis. Neurosci Lett 44, 127-131. Narindrasorasak, S., Lowery, D.E., Altman, R.A., Gonzalez-DeWhitt, P., Greenberg, B., Kisilevsky, R., 1992. Characterization of high affinity binding between laminin and Alzheimer's disease amyloid precursor proteins. Lab Invest 67, 643-652. Naruse, S., Thinakaran, G., Luo, J.J., Kusiak, J.W., Tomita, T., Iwatsubo, T., Qian, X., Ginty, D.D., Price, D.L., Borchelt, D.R., Wong, P.C., Sisodia, S.S., 1998. Effects of PS1 deficiency on membrane protein trafficking in neurons. Neuron 21, 1213-1221. Navarro, I., Leibush, B., Moon, T.W., Plisetskaya, E.M., Banos, N., Mendez, E., Planas, J.V., Gutierrez, J., 1999. Insulin, insulin-like growth factor-I (IGF-I) and glucagon: the evolution of their receptors. Comp Biochem Physiol B Biochem Mol Biol 122, 137-153. Nikolaev, A., McLaughlin, T., O'Leary, D.D., Tessier-Lavigne, M., 2009. APP binds DR6 to trigger axon pruning and neuron death via distinct caspases. Nature 457, 981-989.

68 Nitsch, R.M., Slack, B.E., Wurtman, R.J., Growdon, J.H., 1992. Release of Alzheimer amyloid precursor derivatives stimulated by activation of muscarinic acetylcholine receptors. Science 258, 304-207. Pahlman, S., Ruusala, A.I., Abrahamsson, L., Mattsson, M.E., Esscher, T., 1984. Retinoic acid-induced differentiation of cultured human neuroblastoma cells: a comparison with phorbolester-induced differentiation. Cell Differ 14, 135-144. Pahlsson, P., Shakin-Eshleman, S.H., Spitalnik, S.L., 1992. N-linked glycosylation of beta-amyloid precursor protein. Biochem Biophys Res Commun 189, 1667-1673. Pahlsson, P., Spitalnik, S.L., 1996. The role of glycosylation in synthesis and secretion of beta-amyloid precursor protein by Chinese hamster ovary cells. Arch Biochem Biophys 331, 177-186. Paliga, K., Peraus, G., Kreger, S., Dürrwang, U., Hesse, L., Multhaup, G., Masters, C.L., Beyreuther, K., Weidemann, A., 1997. Human amyloid precursor-like protein 1 - cDNA cloning, ectopic expression in COS-7 cells and identification of soluble forms in the cerebrospinal fluid. Eur J Biochem 250, 354-363. Pastorino, L., Ikin, A.F., Lamprianou, S., Vacaresse, N., Revelli, J.P., Platt, K., Paganetti, P., Mathews, P.M., Harroch, S., Buxbaum, J.D., 2004. BACE (-secretase) modulates the processing of APLP2 in vivo. Mol Cell Neurosci 25, 642-649. Pastorino, L., Sun, A., Lu, P.J., Zhou, X.Z., Balastik, M., Finn, G., Wulf, G., Lim, J., Li, S.H., Li, X., Xia, W., Nicholson, L.K., Lu, K.P., 2006. The prolyl isomerase Pin1 regulates amyloid precursor protein processing and amyloid- production. Nature 440, 528-533. Peiretti, F., Canault, M., Deprez-Beauclair, P., Berthet, V., Bonardo, B., Juhan-Vague, I., Nalbone, G., 2003. Intracellular maturation and transport of tumor necrosis factor alpha converting enzyme. Exp Cell Res 285, 278-285. Perdivara, I., Petrovich, R., Allinquant, B., Deterding, L.J., Tomer, K.B., Przybylski, M., 2009. Elucidation of O-glycosylation structures of the beta-amyloid precursor protein by liquid chromatography-mass spectrometry using electron transfer dissociation and collision induced dissociation. J Proteome Res 8, 631-642. Perez, R.G., Zheng, H., Van der Ploeg, L.H., Koo, E.H., 1997. The - amyloid precursor protein of Alzheimer's disease enhances neuron viability and modulates neuronal polarity. J Neurosci 17, 9407-9414. Peschon, J.J., Slack, J.L., Reddy, P., Stocking, K.L., Sunnarborg, S.W., Lee, D.C., Russell, W.E., Castner, B.J., Johnson, R.S., Fitzner, J.N., Boyce, R.W., Nelson, N., Kozlosky, C.J., Wolfson, M.F., Rauch, C.T., Cerretti, D.P., Paxton, R.J., March, C.J., Black, R.A., 1998. An essential role for ectodomain shedding in mammalian development. Science 282, 1281-1284. Pietrzik, C.U., Yoon, I.S., Jaeger, S., Busse, T., Weggen, S., Koo, E.H., 2004. FE65 constitutes the functional link between the low-density

69 lipoprotein receptor-related protein and the amyloid precursor protein. J Neurosci 24, 4259-4265. Postina, R., Schroeder, A., Dewachter, I., Bohl, J., Schmitt, U., Kojro, E., Prinzen, C., Endres, K., Hiemke, C., Blessing, M., Flamez, P., Dequenne, A., Godaux, E., van Leuven, F., Fahrenholz, F., 2004. A disintegrin-metalloproteinase prevents amyloid plaque formation and hippocampal defects in an Alzheimer disease mouse model. J Clin Invest 113, 1456-1464. Qi-Takahara, Y., Morishima-Kawashima, M., Tanimura, Y., Dolios, G., Hirotani, N., Horikoshi, Y., Kametani, F., Maeda, M., Saido, T.C., Wang, R., Ihara, Y., 2005. Longer forms of amyloid beta protein: implications for the mechanism of intramembrane cleavage by gamma-secretase. J Neurosci 25, 436-445. Querfurth, H.W., LaFerla, F.M., 2010. Alzheimer's disease. N Engl J Med 362, 329-344. Reddy, P., Slack, J.L., Davis, R., Cerretti, D.P., Kozlosky, C.J., Blanton, R.A., Shows, D., Peschon, J.J., Black, R.A., 2000. Functional analysis of the domain structure of tumor necrosis factor-alpha converting enzyme. J Biol Chem 275, 14608-14614. Ring, S., Weyer, S.W., Kilian, S.B., Waldron, E., Pietrzik, C.U., Filippov, M.A., Herms, J., Buchholz, C., Eckman, C.B., Korte, M., Wolfer, D.P., Muller, U.C., 2007. The secreted beta-amyloid precursor protein ectodomain APPs alpha is sufficient to rescue the anatomical, behavioral, and electrophysiological abnormalities of APP-deficient mice. J Neurosci 27, 7817-7826. Roberds, S.L., Anderson, J., Basi, G., Bienkowski, M.J., Branstetter, D.G., Chen, K.S., Freedman, S.B., Frigon, N.L., Games, D., Hu, K., Johnson-Wood, K., Kappenman, K.E., Kawabe, T.T., Kola, I., Kuehn, R., Lee, M., Liu, W., Motter, R., Nichols, N.F., Power, M., Robertson, D.W., Schenk, D., Schoor, M., Shopp, G.M., Shuck, M.E., Sinha, S., Svensson, K.A., Tatsuno, G., Tintrup, H., Wijsman, J., Wright, S., McConlogue, L., 2001. BACE knockout mice are healthy despite lacking the primary beta-secretase activity in brain: implications for Alzheimer's disease therapeutics. Hum Mol Genet 10, 1317-1324. Roberts, S.B., Ripellino, J.A., Ingalls, K.M., Robakis, N.K., Felsenstein, K.M., 1994. Non-amyloidogenic cleavage of the -amyloid precursor protein by an integral membrane metalloendopeptidase. J Biol Chem 269, 3111-3116. Rosen, D.R., Martin-Morris, L., Luo, L.Q., White, K., 1989. A Drosophila gene encoding a protein resembling the human -amyloid protein precursor. Proc Natl Acad Sci U S A 86, 2478-2482. Rossjohn, J., Cappai, R., Feil, S.C., Henry, A., McKinstry, W.J., Galatis, D., Hesse, L., Multhaup, G., Beyreuther, K., Masters, C.L., Parker, M.W., 1999. Crystal structure of the N-terminal, growth factor-like domain of Alzheimer amyloid precursor protein. Nat Struct Biol 6, 327-331.

70 Russ, W.P., Engelman, D.M., 2000. The GxxxG motif: a framework for transmembrane helix-helix association. J Mol Biol 296, 911-919. Schlondorff, J., Becherer, J.D., Blobel, C.P., 2000. Intracellular maturation and localization of the tumour necrosis factor alpha convertase (TACE). Biochem J 347 Pt 1, 131-138. Seabrook, G.R., Smith, D.W., Bowery, B.J., Easter, A., Reynolds, T., Fitzjohn, S.M., Morton, R.A., Zheng, H., Dawson, G.R., Sirinathsinghji, D.J., Davies, C.H., Collingridge, G.L., Hill, R.G., 1999. Mechanisms contributing to the deficits in hippocampal synaptic plasticity in mice lacking amyloid precursor protein. Neuropharmacology 38, 349-359. Shafi, R., Iyer, S.P., Ellies, L.G., O'Donnell, N., Marek, K.W., Chui, D., Hart, G.W., Marth, J.D., 2000. The O-GlcNAc transferase gene resides on the X chromosome and is essential for embryonic stem cell viability and mouse ontogeny. Proc Natl Acad Sci U S A 97, 5735-5739. Shah, S., Lee, S.F., Tabuchi, K., Hao, Y.H., Yu, C., LaPlant, Q., Ball, H., Dann, C.E., Sudhof, T., Yu, G., 2005. Nicastrin functions as a gamma-secretase substrate receptor. Cell 122, 435-447. Shankar, G.M., Bloodgood, B.L., Townsend, M., Walsh, D.M., Selkoe, D.J., Sabatini, B.L., 2007. Natural oligomers of the Alzheimer amyloid- beta protein induce reversible synapse loss by modulating an NMDA-type glutamate receptor-dependent signaling pathway. J Neurosci 27, 2866-2875. Simanek, E.E., McGarvey, G.J., Jablonowski, J.A., Wong, C.H., 1998. Selectinminus signCarbohydrate Interactions: From Natural Ligands to Designed Mimics. Chem Rev 98, 833-862. Sinha, S., Anderson, J.P., Barbour, R., Basi, G.S., Caccavello, R., Davis, D., Doan, M., Dovey, H.F., Frigon, N., Hong, J., Jacobson-Croak, K., Jewett, N., Keim, P., Knops, J., Lieberburg, I., Power, M., Tan, H., Tatsuno, G., Tung, J., Schenk, D., Seubert, P., Suomensaari, S.M., Wang, S., Walker, D., Zhao, J., McConlogue, L., John, V., 1999. Purification and cloning of amyloid precursor protein -secretase from human brain. Nature 402, 537-540. Sisodia, S.S., 1992. Beta-amyloid precursor protein cleavage by a membrane-bound protease. Proc Natl Acad Sci U S A 89, 6075- 6079. Sisodia, S.S., Koo, E.H., Beyreuther, K., Unterbeck, A., Price, D.L., 1990. Evidence that -amyloid protein in Alzheimer's disease is not derived by normal processing. Science 248, 492-495. Skovronsky, D.M., Moore, D.B., Milla, M.E., Doms, R.W., Lee, V.M., 2000. Protein kinase C-dependent alpha-secretase competes with beta-secretase for cleavage of amyloid-beta precursor protein in the trans-golgi network. J Biol Chem 275, 2568-2575. Slack, B.E., Ma, L.K., Seah, C.C., 2001. Constitutive shedding of the amyloid precursor protein ectodomain is up-regulated by tumour necrosis factor- converting enzyme. Biochem J 357, 787-794.

71 Slunt, H.H., Thinakaran, G., Von Koch, C., Lo, A.C., Tanzi, R.E., Sisodia, S.S., 1994. Expression of a ubiquitous, cross-reactive homologue of the mouse -amyloid precursor protein (APP). J Biol Chem 269, 2637-2644. Small, D., Nurcombe, V., Reed, G., Clarris, H., Moir, R., Beyreuther, K., Masters, C.L., 1994. A heparin-binding domain in the amyloid protein precursor of Alzheimer’s disease is involved in the regulation of neurite outgrowth. J Neurosci 14, 2117-2127. Soba, P., Eggert, S., Wagner, K., Zentgraf, H., Siehl, K., Kreger, S., Löwer, A., Langer, A., Merdes, G., Paro, R., Masters, C.L., Müller, U., Kins, S., Beyreuther, K., 2005. Homo- and heterodimerization of APP family members promotes intercellular adhesion. EMBO J 24, 3624-3634. Soond, S.M., Everson, B., Riches, D.W., Murphy, G., 2005. ERK-mediated phosphorylation of Thr735 in TNFalpha-converting enzyme and its potential role in TACE protein trafficking. J Cell Sci 118, 2371- 2380. Srour, N., Lebel, A., McMahon, S., Fournier, I., Fugere, M., Day, R., Dubois, C.M., 2003. TACE/ADAM-17 maturation and activation of sheddase activity require proprotein convertase activity. FEBS Lett 554, 275-283. Stawikowska, R., Cudic, M., Giulianotti, M., Houghten, R.A., Fields, G.B., Minond, D., 2013. Activity of ADAM17 (a disintegrin and metalloprotease 17) is regulated by its noncatalytic domains and secondary structure of its substrates. J Biol Chem 288, 22871-22879. Steen, E., Terry, B.M., Rivera, E.J., Cannon, J.L., Neely, T.R., Tavares, R., Xu, X.J., Wands, J.R., de la Monte, S.M., 2005. Impaired insulin and insulin-like growth factor expression and signaling mechanisms in Alzheimer's disease--is this type 3 diabetes? J Alzheimers Dis 7, 63-80. Suh, J., Choi, S.H., Romano, D.M., Gannon, M.A., Lesinski, A.N., Kim, D.Y., Tanzi, R.E., 2013. ADAM10 Missense Mutations Potentiate beta-Amyloid Accumulation by Impairing Prodomain Chaperone Function. Neuron. Suzuki, T., Ando, K., Isohara, T., Oishi, M., Lim, G.S., Satoh, Y., Wasco, W., Tanzi, R.E., Nairn, A.C., Greengard, P., Gandy, S.E., Kirino, Y., 1997. Phosphorylation of Alzheimer -amyloid precursor-like proteins. Biochemistry 36, 4643-4649. Suzuki, T., Oishi, M., Marshak, D.R., Czernik, A.J., Nairn, A.C., Greengard, P., 1994. Cell cycle-dependent regulation of the phosphorylation and metabolism of the Alzheimer amyloid precursor protein. EMBO J 13, 1114-1122. Takahashi, K., Niidome, T., Akaike, A., Kihara, T., Sugimoto, H., 2008. Phosphorylation of amyloid protein (APP) at Tyr687 regulates APP processing by  and secretase. Biochem Biophys Res Commun 377, 544-549.

72 Takami, M., Nagashima, Y., Sano, Y., Ishihara, S., Morishima-Kawashima, M., Funamoto, S., Ihara, Y., 2009. gamma-Secretase: successive tripeptide and tetrapeptide release from the transmembrane domain of beta-carboxyl terminal fragment. J Neurosci 29, 13042-13052. Tallent, M.K., Varghis, N., Skorobogatko, Y., Hernandez-Cuebas, L., Whelan, K., Vocadlo, D.J., Vosseller, K., 2009. In vivo modulation of O-GlcNAc levels regulates hippocampal synaptic plasticity through interplay with phosphorylation. J Biol Chem 284, 174-181. Taru, H., Suzuki, T., 2004. Facilitation of stress-induced phosphorylation of beta-amyloid precursor protein family members by X11-like/Mint2 protein. J Biol Chem 279, 21628-21636. Tomita, S., Kirino, Y., Suzuki, T., 1998. Cleavage of Alzheimer's amyloid precursor protein (APP) by secretases occurs after O-glycosylation of APP in the protein secretory pathway. Identification of intracellular compartments in which APP cleavage occurs without using toxic agents that interfere with protein metabolism. J Biol Chem 273, 6277-6284. Tsakadze, N.L., Sithu, S.D., Sen, U., English, W.R., Murphy, G., D'Souza, S.E., 2006. Tumor necrosis factor-alpha-converting enzyme (TACE/ADAM-17) mediates the ectodomain cleavage of intercellular adhesion molecule-1 (ICAM-1). J Biol Chem 281, 3157-3164. Walsh, D.M., Klyubin, I., Fadeeva, J.V., Cullen, W.K., Anwyl, R., Wolfe, M.S., Rowan, M.J., Selkoe, D.J., 2002. Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature 416, 535-539. Wang, P., Yang, G., Mosier, D.R., Chang, P., Zaidi, T., Gong, Y.D., Zhao, N.M., Dominguez, B., Lee, K.F., Gan, W.B., Zheng, H., 2005. Defective neuromuscular synapses in mice lacking amyloid precursor protein (APP) and APP-Like protein 2. J Neurosci 25, 1219-1225. Wang, Y., Ha, Y., 2004. The X-ray structure of an antiparallel dimer of the human amyloid precursor protein E2 domain. Mol Cell 15, 343-353. Wang, Z., Wang, B., Yang, L., Guo, Q., Aithmitti, N., Songyang, Z., Zheng, H., 2009. Presynaptic and postsynaptic interaction of the amyloid precursor protein promotes peripheral and central synaptogenesis. J Neurosci 29, 10788-10801. Wasco, W., Bupp, K., Magendantz, M., Gusella, J.F., Tanzi, R.E., Solomon, F., 1992. Identification of a mouse brain cDNA that encodes a protein related to the Alzheimer disease-associated amyloid beta protein precursor. Proc Natl Acad Sci U S A 89, 10758-10762. Wasco, W., Gurubhagavatula, S., Paradis, M.D., Romano, D.M., Sisodia, S.S., Hyman, B.T., Neve, R.L., Tanzi, R.E., 1993. Isolation and characterization of APLP2 encoding a homologue of the Alzheimer's associated amyloid beta protein precursor. Nat Genet 5, 95-100. Vassar, R., Bennett, B.D., Babu-Khan, S., Kahn, S., Mendiaz, E.A., Denis, P., Teplow, D.B., Ross, S., Amarante, P., Loeloff, R., Luo, Y.,

73 Fisher, S., Fuller, J., Edenson, S., Lile, J., Jarosinski, M.A., Biere, A.L., Curran, E., Burgess, T., Louis, J.C., Collins, F., Treanor, J., Rogers, G., Citron, M., 1999. -secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science 286, 735-741. Weidemann, A., Eggert, S., Reinhard, F.B., Vogel, M., Paliga, K., Baier, G., Masters, C.L., Beyreuther, K., Evin, G., 2002. A novel -cleavage within the transmembrane domain of the Alzheimer amyloid precursor protein demonstrates homology with Notch processing. Biochemistry 41, 2825-2835. Vella, L.J., Cappai, R., 2012. Identification of a novel amyloid precursor protein processing pathway that generates secreted N-terminal fragments. FASEB J 26, 2930-2940. Vetrivel, K.S., Meckler, X., Chen, Y., Nguyen, P.D., Seidah, N.G., Vassar, R., Wong, P.C., Fukata, M., Kounnas, M.Z., Thinakaran, G., 2009. Alzheimer disease Abeta production in the absence of S- palmitoylation-dependent targeting of BACE1 to lipid rafts. J Biol Chem 284, 3793-3803. Weyer, S.W., Klevanski, M., Delekate, A., Voikar, V., Aydin, D., Hick, M., Filippov, M., Drost, N., Schaller, K.L., Saar, M., Vogt, M.A., Gass, P., Samanta, A., Jaschke, A., Korte, M., Wolfer, D.P., Caldwell, J.H., Muller, U.C., 2011. APP and APLP2 are essential at PNS and CNS synapses for transmission, spatial learning and LTP. EMBO J 30, 2266-2280. White, A.R., Reyes, R., Mercer, J.F., Camakaris, J., Zheng, H., Bush, A.I., Multhaup, G., Beyreuther, K., Masters, C.L., Cappai, R., 1999. Copper levels are increased in the cerebral cortex and liver of APP and APLP2 knockout mice. Brain Res 842, 439-444. Willem, M., Garratt, A.N., Novak, B., Citron, M., Kaufmann, S., Rittger, A., DeStrooper, B., Saftig, P., Birchmeier, C., Haass, C., 2006. Control of peripheral nerve myelination by the beta-secretase BACE1. Science 314, 664-666. Wolfe, M.S., 2006. The -secretase complex: membrane-embedded proteolytic ensemble. Biochemistry 45, 7931-7939. Wolfe, M.S., Xia, W., Ostaszewski, B.L., Diehl, T.S., Kimberly, W.T., Selkoe, D.J., 1999. Two transmembrane aspartates in presenilin-1 required for presenilin endoproteolysis and -secretase activity. Nature 398, 513-517. von Koch, C.S., Zheng, H., Chen, H., Trumbauer, M., Thinakaran, G., van der Ploeg, L.H., Price, D.L., Sisodia, S.S., 1997. Generation of APLP2 KO mice and early postnatal lethality in APLP2/APP double KO mice. Neurobiol Aging 18, 661-669. Xu, K.P., Zoukhri, D., Zieske, J.D., Dartt, D.A., Sergheraert, C., Loing, E., Yu, F.S., 2001. A role for MAP kinase in regulating ectodomain shedding of APLP2 in corneal epithelial cells. Am J Physiol Cell Physiol 281, 603-614.

74 Xue, Y., Lee, S., Ha, Y., 2011. Crystal structure of amyloid precursor-like protein 1 and heparin complex suggests a dual role of heparin in E2 dimerization. Proc Natl Acad Sci U S A 108, 16229-16234. Yan, R., Bienkowski, M.J., Shuck, M.E., Miao, H., Tory, M.C., Pauley, A.M., Brashier, J.R., Stratman, N.C., Mathews, W.R., Buhl, A.E., Carter, D.B., Tomasselli, A.G., Parodi, L.A., Heinrikson, R.L., Gurney, M.E., 1999. Membrane-anchored aspartyl protease with Alzheimer's disease -secretase activity. Nature 402, 533-537. Yang, X., Ongusaha, P.P., Miles, P.D., Havstad, J.C., Zhang, F., So, W.V., Kudlow, J.E., Michell, R.H., Olefsky, J.M., Field, S.J., Evans, R.M., 2008. Phosphoinositide signalling links O-GlcNAc transferase to insulin resistance. Nature 451, 964-969. Yang, Y.R., Song, M., Lee, H., Jeon, Y., Choi, E.J., Jang, H.J., Moon, H.Y., Byun, H.Y., Kim, E.K., Kim, D.H., Lee, M.N., Koh, A., Ghim, J., Choi, J.H., Lee-Kwon, W., Kim, K.T., Ryu, S.H., Suh, P.G., 2012. O-GlcNAcase is essential for embryonic development and maintenance of genomic stability. Aging Cell 11, 439-448. Yazaki, M., Tagawa, K., Maruyama, K., Sorimachi, H., Tsuchiya, T., Ishiura, S., Suzuki, K., 1996. Mutation of potential N-linked glycosylation sites in the Alzheimer's disease amyloid precursor protein (APP). Neurosci Lett 221, 57-60. Young-Pearse, T.L., Chen, A.C., Chang, R., Marquez, C., Selkoe, D.J., 2008. Secreted APP regulates the function of full-length APP in neurite outgrowth through interaction with integrin beta1. Neural Dev 3, 15. Yu, C., Kim, S.H., Ikeuchi, T., Xu, H., Gasparini, L., Wang, R., Sisodia, S.S., 2001. Characterization of a presenilin-mediated amyloid precursor protein carboxyl-terminal fragment . Evidence for distinct mechanisms involved in -secretase processing of the APP and Notch1 transmembrane domains. J Biol Chem 276, 43756-43760. Yu, G., Chen, F., Levesque, G., Nishimura, M., Zhang, D.M., Levesque, L., Rogaeva, E., Xu, D., Liang, Y., Duthie, M., St George-Hyslop, P.H., Fraser, P.E., 1998. The presenilin 1 protein is a component of a high molecular weight intracellular complex that contains beta-catenin. J Biol Chem 273, 16470-16475. Yu, G., Nishimura, M., Arawaka, S., Levitan, D., Zhang, L., Tandon, A., Song, Y.Q., Rogaeva, E., Chen, F., Kawarai, T., Supala, A., Levesque, L., Yu, H., Yang, D.S., Holmes, E., Milman, P., Liang, Y., Zhang, D.M., Xu, D.H., Sato, C., Rogaev, E., Smith, M., Janus, C., Zhang, Y., Aebersold, R., Farrer, L.S., Sorbi, S., Bruni, A., Fraser, P., St George-Hyslop, P., 2000. Nicastrin modulates presenilin-mediated notch/glp-1 signal transduction and APP processing. Nature 407, 48-54.