br Author contributions br Transparency document br Acknowle

Author contributions
Transparency document
Acknowledgments The financial contributions of the Marie-Curie Action: BIOCONTROL, contract number MCRTN – 33439, the Agence Nationale de la Recherche (projects membraneDNP 12-BSV5-0012, MemPepSyn 14-CE34-0001-01, InMembrane 15-CE11-0017-01 and the LabEx Chemistry of Complex Systems 10-LABX-0026_CSC), the University of Strasbourg, the CNRS, and the RTRA International Center of Frontier Research in Chemistry, are gratefully acknowledged. The authors like to thank the Helmholtz-Zentrum Berlin für Materialien und Energie (Berlin, Germany) for allocating neutron beam time at the Membrane Diffractometer V1. PT and GF thank CALI, Nouvelle Aquitaine Region and INSERM. PT and MO thank the Czech Science Foundation (P208/12/G016) and National Program of Sustainability I from the Ministry-of-Youth, Education and Sports of the Czech Republic (LO1305). B.S. thanks the Hans and Ilse Breuer-Stiftung for support. BB is grateful to the Institut Universitaire de France for providing additional time to be dedicated to research.
Introduction A hallmark of AD is the accumulation of β-amyloid plaque in patient Fosmidomycin sodium salt sale [1, 2]. APP is first cleaved by β-secretase, generating a 99-residue transmembrane fragment known as C99; C99 is then trimmed successively by γ-secretase to produce a number of β-amyloid peptides (Aβ). Longer forms of Aβ such as Aβ42 and Aβ43 are prone to aggregation [3, 4, 5]. The four-component intramembrane γ-secretase, comprising presenilin 1/2 (PS1/2), presenilin enhancer protein 2 (PEN-2), anterior pharynx defective protein 1 (APH-1a/b), and nicastrin, is thought to play a central role in the development of AD by producing Aβ [6, 7, 8]. Genetic screening of AD patients led to the identification of presenilin [9]. An autocatalytic cleavage of the cytoplasmic loop between transmembrane segment (TM) 6 and TM7 of presenilin is required for γ-secretase maturation, generating an amino-terminal fragment (NTF) comprising TMs 1–6, and a carboxyl-terminal fragment (CTF) comprising TMs 7–9 [10]. Two transmembrane aspartate residues are essential for presenilin maturation and substrate cleavage [8, 11, 12, 13]. PEN-2 facilitates the autocatalytic cleavage of presenilin [14]. APH-1 may serve as a scaffold to stabilize γ-secretase [15, 16]. Nicastrin, with a single TM and a large extracellular domain (ECD), is thought to be responsible for substrate recruitment [17, 18]. For much of the past two decades, structural characterization of γ-secretase was lagging behind rapid advances in functional understanding. Until the beginning of 2014, there were only several low-resolution EM structures [19, 20, 21, 22, 23], a nuclear magnetic resonance (NMR) structure of PS1 CTF [24], and a crystal structure of a presenilin archaeal homologue (PSH) []. Breakthroughs in recombinant γ-secretase expression [], together with the application of direct electron detector and improvement of image analysis software [27, 28, 29], allowed successful determination of four near-atomic resolution structures: γ-secretase at 4.5Å [], γ-secretase fused with T4 lysozyme at 4.32Å [], and γ-secretase at 3.4Å [] and soaked with a specific inhibitor DAPT at 4.2Å [].
Expression and purification of human γ-secretase The acquisition of a homogenous membrane protein is always a major bottleneck for structure determination. This problem had been particularly worse for γ-secretase than for most other mammalian membrane proteins, because γ-secretase contains four components and as a protease poses a threat to the host cells upon overexpression. Earlier effort to obtain decent amounts of recombinant γ-secretase relied on stable transfection of CHO and HEK293S cells. Three CHO cell lines, named γ-30, S1 and S20, were stably transfected by plasmids that express components of γ-secretase [33, 34]. Specifically, γ-30 contains PS1, APH-1 and PEN-2; S1 and S20 are derived from γ-30 and contain differentially tagged nicastrin. The yield of recombinant γ-secretase from these CHO cells is relatively low, about 10–20μg per liter of cell culture [33, 34]. Transposon-mediated multi-gene stable integration technology also allowed generation of stable CHO cell lines for γ-secretase expression, which has a markedly higher yield of nearly 1mg per liter of cell culture [35]. In addition, HEK293S cells were stably transfected for γ-secretase expression [22]. Recombinant γ-secretase reported in these studies is proteolytically active.