This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shimizu, H.
Right arrow Articles by Nukina, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shimizu, H.
Right arrow Articles by Nukina, N.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, June 2008, p. 3663-3671, Vol. 28, No. 11
0270-7306/08/$08.00+0     doi:10.1128/MCB.02185-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Crystal Structure of an Active Form of BACE1, an Enzyme Responsible for Amyloid β Protein Production{triangledown}

Hideaki Shimizu,1 Asako Tosaki,1 Kumi Kaneko,1 Tamao Hisano,2 Takashi Sakurai,1 and Nobuyuki Nukina1*

Laboratory for Structural Neuropathology, RIKEN Brain Science Institute, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan,1 SPring-8 Center, RIKEN Harima Institute, Koto 1-1-1, Sayo-cho, Sayo-gun, Hyogo 679-5148, Japan2

Received 11 December 2007/ Returned for modification 28 January 2008/ Accepted 17 March 2008


arrow
ABSTRACT
 
BACE1 (β-secretase) is a transmembrane aspartic protease that cleaves the β-amyloid precursor protein and generates the amyloid β peptide (Aβ). BACE1 cycles between the cell surface and the endosomal system many times and becomes activated interconvertibly during its cellular trafficking, leading to the production of Aβ. Here we report the crystal structure of the catalytically active form of BACE1. The active form has novel structural features involving the conformation of the flap and subsites that promote substrate binding. The functionally essential residues and water molecules are well defined and play a key role in the iterative activation of BACE1. We further describe the crystal structure of the dehydrated form of BACE1, showing that BACE1 activity is dependent on the dynamics of a catalytically required Asp-bound water molecule, which directly affects its catalytic properties. These findings provide insight into a novel regulation of BACE1 activity and elucidate how BACE1 modulates its activity during cellular trafficking.


arrow
INTRODUCTION
 
The transmembrane aspartic protease BACE1 (also called Asp2 or memapsin 2) is reported to be a β-secretase that is involved in Alzheimer's disease (AD) and cleaves the amyloid precursor protein (APP) to generate an amyloid β peptide (Aβ) following cleavage by {gamma}-secretase (17, 31, 36, 41). Aβ is the primary constituent of amyloid plaques found in the brains of AD patients. Its accumulation is supposed to be toxic and supposed to induce AD pathologies, such as accumulation of tau neurofibrillary tangles or neuronal cell death (30). On the basis of this hypothesis of amyloid, the proteins BACE1 and {gamma}-secretase, which mediate the amyloidogenic processing of APP, are thought to be prime drug targets for the treatment of AD (11, 30).

It has been reported that BACE1 molecules are localized within the trans-Golgi network and endosomal system, where they colocalize with APP (4, 16). In fact, APP cleavage by β-secretase is reported to occur in all of these compartments (4, 16, 17, 20, 36). In addition, BACE1 is reinternalized from the cell surface to early endosomes and can be recycled back to the cell surface (16, 38). Considering the fact that BACE1 activity is optimal at an acidic pH, BACE1 should become activated within the late Golgi compartments and endosomes/lysosomes (20, 31, 36, 41). It has recently been reported that endocytosis of lipid raft domains from the cell surface is required for BACE1 and APP to meet and generate Aβ (6). Although there is still some controversy regarding the subcellular compartments where BACE1 acts, these acidic compartments, especially in endosomes, are likely to be major sites for APP cleavage by BACE1, leading to Aβ production (37).

BACE1 has become a major target in the development of drugs for AD, and X-ray crystallography has determined numerous structures of BACE1 complexes (12, 13, 24, 34). These studies indicated that the active site of BACE1 is covered by a flexible antiparallel β-hairpin, called a flap, which is believed to control substrate access to the active site and set the substrate into the correct geometry for the catalytic process. It has been reported that the flap of the inhibitor-bound form is tightly packed in a closed conformation; however, the substrate-free (apo) structure of BACE1 showed that the flap was in an open conformation (14, 24). These results indicate that a conformational change must take place upon binding of the inhibitor/substrate to the active site and may participate kinetically in substrate binding in the closed conformation and product release in the open conformation. In addition, a simulation study revealed that self-inhibited conformation of a conserved Tyr residue was present in the apo structure of BACE1 (8), while other simulation studies suggested that a large-scale conformational change in BACE1 was associated with the modulation of enzymatic activity (5, 40). These observations implicate the regulation of BACE1 activity in the modulation of substrate affinity and catalytic properties caused by large-scale conformational changes between the active and inactive forms. However, the structural details of such changes have not been reported, since the structure of active BACE1 had not been determined previously.

In the aspartic proteases, two completely conserved water molecules were observed. The first water molecule (Wat1) is located between the Asp pair of Asp32 and Asp228 of BACE1. It has been proposed that BACE1 and aspartic proteases catalyze peptide bond hydrolysis through an acid-base catalysis mechanism mediated by Wat1 and the Asp pair as follows. (i) After substrate binding, the free Asp pair activates Wat1 by forming a hydrogen bond with it. (ii) Next, the activated Wat1, it has been proposed, nucleophilically attacks the scissile-bond carbonyl. The resulting geminal diol intermediate is stabilized by hydrogen bonds with the carboxyl group of Asp. (iii) Finally, decomposition of the scissile C-N bond is accompanied by the transfer of a proton from Asp to the leaving amino group (see scheme 1 in reference 32). In addition, kinetics studies suggested that Wat1 competitively inhibits the binding of inhibitor to the active site (21). On the other hand, the second conserved water molecule (Wat2) is involved in the hydrogen bond, with a conserved Tyr residue in the flap. Wat2 also participates in a conserved hydrogen-bonding network Wat2-Ser35-Asp32-Wat1-Asp228 and was proposed to assist in the catalytic reaction (2, 8). Thus, BACE1 activity is directly affected by the behaviors of these water molecules.

In this paper, we describe the crystal structure of the catalytically active form of BACE1 and provide a detailed analysis of structural switching in the activation/inactivation process. Our study reveals new features of the structure in active BACE1 and provides structural insights into the role of active site residues and water molecules in enzymatic properties. We propose a novel regulatory mechanism of enzymatic activity, and discuss the potential physiological significance of the structural switching in BACE1.


arrow
MATERIALS AND METHODS
 
Expression and crystallization of active and inactive structures. Recombinant human BACE1 was produced in Escherichia coli as inclusion bodies, and these were then refolded and purified as previously described (29). The purified BACE1 was concentrated at 8 to 18 mg/ml. The protein solution was estimated to be >95% pure as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Crystallization for apo BACE1 was performed by the sitting-drop vapor diffusion method described previously (24). Equal volumes of protein solution and mother liquor, containing 15 to 22.5% (wt/vol) polyethylene glycol 5000 monomethyl ether (PEG 5000 MME), 200 mM ammonium iodide, 200 mM sodium citrate, pH 6.0 to 7.0, were mixed in a single droplet and equilibrated against 0.5 ml of mother liquor at 20°C. The crystal belongs to the hexagonal system, space group P6122, with one molecule in an asymmetric unit. The crystal for data collection on the structure at pH 7.0 was incubated in soaking buffer (15 to 22.5% [wt/vol] PEG 5000 MME, 200 mM sodium citrate) at pH 7.0 for 24 h. In the case of crystals for active structures, they were transferred to soaking buffer at pH 5.0 for 12 to 24 h, and some of them were further transferred to soaking buffer at pH 4.5 for 24 h or pH 4.0 for 6 h at 20°C.

Structure determination and refinement of active and inactive structures. Before data collection, the crystals were transferred to a new soaking buffer containing 20% glycerol at pH 4.0, 4.5, 5.0, and 7.0 and subsequently flash-frozen in nitrogen gas. X-ray diffraction data sets were collected with the BL26B1 and BL44B2 beamlines at the SPring-8 Center, using X rays with a wavelength of 1.0 Å and a temperature of 100 K. They were integrated and scaled using the HKL2000 and SCALEPACK software packages (23). The structure was solved by a combination of molecular replacements using the CCP4 program MOLREP (42) and the CNS program (3), employing the previously determined apo BACE1 structure (Protein Data Bank [PDB] accession no. 1W50) as a search model. The progress of refinement was monitored by the Rfree value (5% of the data), and neither a {sigma} nor a low-resolution cutoff was applied to the data. The structure was refined by rigid-body fitting followed by the simulated-annealing protocol implemented in the CNS program (3) interspersed with rounds of model building by the program TURBO-FRODO (28). Water molecules were included in the model if they were within hydrogen-bonding distance of chemically reasonable groups, appeared in Fo-Fc maps contoured at 3.0 {sigma}, and had a B factor less than 60 Å2. The regions 158 to 168 and 310 to 316 were not modeled because of weak electron densities.

Crystallization and structure determination of BACE1 structure in complex with OM99-2 at pH 5.0. The crystal of BACE1 in complex with OM99-2 was prepared by mixing the protein solution (7.2 mg/ml) with the OM99-2 in a 3.3- to 6.6-fold molar excess. Crystals were grown using the vapor diffusion method in a hanging-drop setup at 20°C by mixing equal volumes of 20% (wt/vol) PEG 5000 MME, 200 mM ammonium iodide, and 200 mM sodium citrate at pH 6.4. Leaf-shape crystals appeared after 1 to 2 weeks and continued to grow to 0.4 to 0.8 mm within 3 to 6 weeks. Crystals for data collection were incubated in soaking buffer (20% [wt/vol] PEG 5000 MME, 200 mM sodium citrate) at pH 5.5 for 80 min and transferred to soaking buffer at pH 5.0 for 3.5 h at 20°C. Before data collection, the crystal was further transferred to a new soaking buffer containing 20% glycerol and flash-cooled by nitrogen gas at 100 K. The frozen crystal was packed in the sample tray and delivered to the SPring-8 Center by home delivery service (35). The X-ray diffraction data set was collected at 100 K in the BL26B2 beamline in the SPring-8 Center via a mail-in data collection system (35). The operation of BL26B2 was carried out by a remote control system, and a set of image data was transferred via the Internet network. Data processing, structure determination, and refinement were performed according to the method described above.

In all structures, no residues lie in disallowed regions of the Ramachandran plot. Data collection and refinement statistics are shown in Table 1. Root mean square (RMS) deviations between alpha-carbon (C{alpha}) positions among BACE1 structures were calculated by using the program LSQKAB from the CCP4 suite (42).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Data collection and refinement statistics

Measurement of BACE1 activity in solution. The activity assay was carried out on 20 nM of BACE1 solution in reaction buffer (100 mM sodium acetate, pH 3.0 to 7.0) containing a fluorogenic substrate, according to the manufacturer's instructions (BioVision). After 2 h of incubation at 37°C in the dark, the samples were analyzed in a SpectraMax M2 fluorescence microplate reader (Molecular Devices) with excitation and emission wavelengths of 355 nm and 494 nm, respectively. Negative controls were performed with 80 nM OM99-2 inhibitor (Calbiochem).

Cleavage assay in BACE1 crystal. Synthetic peptide EVNLDAEF (1 mg/ml) was incubated and gently mixed with a single BACE1 crystal in 0.3 ml of soaking buffer (20% [wt/vol] PEG 5000 MME, 200 mM sodium citrate, pH 5.0 or pH 7.0) at 20°C. The number of peptides was estimated by the detection of peptide N termini by using the O-phthalaldehyde (OPA) protein assay method. One microliter of peptide solution was mixed with 30 to 100 µl of OPA reagent solution (5.4 mg/ml OPA, 0.03% Brij 35, 0.2% β-mercaptoethanol, 0.4 M boric acid, pH 10.4) for 10 min at 25°C. OPA-derived fluorescence was measured using the Arvo MX fluorescence microplate reader (PerkinElmer) with excitation and emission wavelengths of 340 nm and 460 nm, respectively. Crystals having the same shape and size in a crystallization batch were used for the cleavage assay (two crystals for each condition). After the cleavage assay, the mass volumes of peptides and BACE1 in final assay solutions were checked by matrix-assisted laser desorption-ionization-time of flight (MALDI-TOF) mass spectrometry as previously described (39). The assay solution was desalted with ZipTip C19 and cocrystallized with {alpha}-cyano-4-hydroxy cinnamic acid matrix for peptide detection and was desalted with a hydrophilic Nutip cartridge (M&S Instruments) and cocrystallized with 2,5-dihydroxy benzoic acid matrix for BACE1 detection. No BACE1 fraction was detected in final assay solutions by MALDI-TOF mass spectrometry (Fig. 1C) and silver-stained sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels (data not shown).


Figure 1
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 1. BACE1 cleavage activity. (A) pH dependence of BACE1 activity in solution. (B) Detection of cleaved peptide by BACE1 crystal. The peptide (EVNLDAEF) solutions at pH 5.0 and pH 7.0 were incubated with BACE1 crystal at 25°C. Control peptide solutions were incubated without BACE1 crystal. Numbers of peptides (EVNLDAEF, EVNL, and DAEF) were estimated by OPA-derived fluorescence intensities. (C) MALDI-TOF mass spectra of the final peptide solution of the cleavage assay at pH 5.0 showing the molecular weights of the full-length peptide (EVNLDAEF), BACE1 cleaved peptide (EVNL or DAEF), and BACE1. No BACE1 signal was detected in any cleavage assay.

Binding of BACE1 to OM99-2. Binding analyses of BACE1 and its inhibitor, OM99-2 (Calbiochem), were performed using the T100 system (Biacore AB, Uppsala, Sweden). OM99-2 was immobilized on the Biacore CM5 sensor chip as previously described (7). OM99-2 (0.01 mg/ml) in 90 mM HEPES, pH 7.4, 1.0 M NaCl, and 10% dimethyl sulfoxide was injected into a CM5 sensor chip, and immobilization was carried out by the standard amine-coupling method. The resonance signal reached about 471 resonance units. Interaction studies were performed with 280 nM of BACE1 in a running buffer containing 100 mM sodium citrate in the pH range of 3.0 to 7.0 and 0.005% surfactant P-20 at 25°C using a flow rate of 30 µl/ml. The abolition and acquisition of inhibitor binding were fully reversible, since a prior treatment by acid had no effect on the pH dependence of inhibitor binding (data not shown).

Protein structure accession numbers. The atomic coordinates of BACE1 have been deposited in the Protein Data Bank (www.pdb.org; PDB accession no. 2ZHS for the structure at pH 4.0, 2ZHT for the structure at pH 4.5, 2ZHU for the structure at pH 5.0, and 2ZHV for the structure at pH 7.0). The atomic coordinates for the OM99-2 complex at pH 5.0 were deposited under accession number 2ZHR.


arrow
RESULTS
 
Activation of crystalline BACE1. The activity of BACE1 in solution was measured by using commercially obtained fluorogenic substrates (Fig. 1A), and a bell curve with the maximum activity around pH 4.5 was revealed. Enzymatic activity was lost rapidly as the pH was either decreased or increased, with no activity at a pH less than 3.5 or higher than 5.5.

To determine whether or not crystalline BACE1 also shows proteolytic activity, a cleavage experiment with a substrate peptide was carried out. Fortunately, our synthetic peptide dissolved easily in soaking buffer (22.5% [wt/vol] PEG 5000 MME, 200 mM sodium citrate), although most fluorogenic BACE1-specific substrates were insoluble. Soaking buffer at pH 5.0 or pH 7.0 was used throughout the assay to avoid crystal dissolution, and no BACE1 enzyme activity was detected in assay buffer only at either pH. The assay at pH 5.0 showed an increase in fluorescence intensity characteristic of cleaved peptides, but the assay at pH 7.0 showed no significant fluorescence, indicating that crystalline BACE1 still has proteolytic activity at an acidic pH (Fig. 1B). Therefore, at pH 7.0, BACE1 is in its inactive state, whereas at pH 4.5, it is active. In other words, the BACE1 in our crystal can display the structural transition related to the enzymatic activity.

Structure of active BACE1. We determined a total of four X-ray structures of free BACE1, one each at pH 4.0, 4.5, 5.0, and 7.0, and also determined the structure of OM99-2 complex at pH 5.0. The resolutions of these structures ranged between 2.35 Å and 2.7 Å, and the values of crystallographic Rwork were between 18.5% and 24.6% and Rfree were between 24.0% and 28.9% (Table 1).

To quantify the differences or similarities among these apo structures, we estimated RMS deviations of C{alpha} atoms for superimposition of the active (pH 4.0, pH 4.5, and pH 5.0), inactive (pH 7.0), and apo (1W50; pH 6.6 [24]) structures (Table 2). These estimations clearly show that, on average, the changes in the main-chain conformation between active and inactive structures are more pronounced than those in the pairs of active structures. The RMS deviations among any of the pairs of active and inactive structures were less than 0.35 Å, but the values between active and inactive structures clearly show high deviations with values of 0.87 to 0.96 Å.


View this table:
[in this window]
[in a new window]

  		TABLE 2. RMS deviation in C{alpha} positions among active and inactive structures

When all structures are superimposed, we can see how the overall protein structure changes as activity changes. One of the most striking differences between the active and inactive structures is the change in the location of residues 67 to 77, the so-called flap. The flap has a long, flexible β-hairpin loop and generally adopts a closed conformation in the inhibitor-bound form. Several residues in the flap interact directly with bound inhibitors (12, 13, 34). On the other hand, the flap in the inactive structure shows an open-flap conformation which is almost identical to that in the apo structure with the PDB accession no. 1W50, reported by Patel et al. (24). The phenolic ring of the conserved Tyr71 near the flap points toward Trp76 and creates a weak hydrogen bond with the indole nitrogen of Trp76 with a distance of 3.3 Å, and the hydrogen bond in the β-strands between Tyr71 and Gly74 is not formed. In contrast, the apo structure with the PDB accession no. 1SGZ, reported by Hong et al. (14), shows the open-flap conformation, but Tyr71 adopts a unique orientation to form a hydrogen bond between the backbone carbonyl of Lys107, and the hydrogen bond in the β-strands between Tyr71 and Gly74 is present in this structure.

It is noteworthy that the flaps in the active structures have adopted more-open conformations than those of the inactive structure (Fig. 2A to C). The flaps in the active structures are displaced significantly, with a maximal C{alpha} displacement (Gln73) of almost 2.2 to 2.5 Å from the positions in our inactive structures and that with PDB accession no. 1W50, 4.0 to 4.3 Å from the apo structure corresponding to PDB accession no. 1SGZ, and 6.9 to 7.2 Å from the OM99-2 complex, accompanied by a conformational change in the neighboring segment. Another difference between the active and inactive structures is the loop around residue 328 on the opposite side of the flap (Fig. 2A). The tip of this loop also undergoes a conformational change in which it moves 1.1 to 1.8 Å from its location in the inactive structure, producing a more open pocket at the active site. A quantitative comparison of the active site cavity by VOIDOO (19) reveals that the cavity in the active structure has a large volume (2,751 to 3,280 Å3) compared to that in the inactive structure (1,778 Å3), although the sizes of the protein in active and inactive structures correspond to the same volume (38,040 to 38,130 Å3).


Figure 2
View larger version (47K):
[in this window]
[in a new window]

  		FIG. 2. Conformational changes associated with activation of BACE1. (A) Stereo view of superimposed foldings in the inactive structure (pH 7.0; dark blue) and active structures (pH 4.0, 4.5, and 5.0; yellow) of BACE1. The side chains of Asp32 and Asp228 in the active site are represented by a red ball-and-stick model. The regions 158 to 167 and 310 to 316 are missing. The displacement from inactive to active structures is indicated by red arrows. ter, terminus. (B) Close-up superimposition of flaps (residues 69 to 75) in the inactive structure (dark blue), active structures (yellow), apo structure reported by Hong et al. (1SGZ; light blue), apo structure reported by Patel et al. (1W50; light blue), and OM99-2 complex (red). (C) Front view of flap and Tyr71. (D) Difference in location in the vicinity of the active site cleft and flap between the active structure (yellow) and the inactive structure (blue). The hydrogen-bonding network Wat2-Ser35-Asp32-Wat1-Asp228 in the active structure is indicated by dashed lines.

Conformational switching in the substrate-binding site. The substrate-binding site is located between the N- and C-terminal lobes, so the motions in both lobes directly affect the catalytic properties of BACE1. Also, the substrate-binding site displays significant conformational changes controlled by altering the pH (Fig. 2D). Leu30, which contributes to the stabilization of the bound inhibitor conformation in the S3 subsite (13), flips its side chain by 180°. The side chain of Arg128 in the S2' pocket, which is located very close to OM99-2, forms a new conformation and makes a new space in the subsite. The positions of Tyr71, Phe108, Ile110, Trp115, and the 10s loop (residues 9 to 14) in the active structures, which are located in the N-terminal lobe and face the active site cleft, are displaced about 0.5 to 1.5 Å from the corresponding position in the inactive structure. As a result, the subsites of S2', S1, S3, and S5 are also pulled in the same direction. Another difference is also observed in the vicinity of Tyr198 in the C-terminal lobe. This region is displaced in the opposite direction by a distance of 0.6 to 1.0 Å. In contrast to the presence of these large conformational movements, the active site in the vicinity of Asp32 and Asp228 displays no changes or only small changes in conformation (less than a 0.5-Å shift). It has been suggested that the orientation of the conserved Tyr residue at the corresponding position of Tyr71 upon ligand binding leads to a new hydrogen bond network and a second active site water molecule (Wat2), conserved in all pepsin-like enzymes, that participates in this network to stabilize the closed conformation (2, 8). In addition, it was proposed that Wat2 and Ser35 swap their proton donor and acceptor properties during the catalytic cycle by using this network (2). In the active structures, Tyr71, as well as the neighboring residues of Trp76 and Phe108, are displaced from their original positions by a distance of 0.8 to 1.0 Å. However, Wat2 is observed in active structures and participates in the hydrogen-bonding network Wat2-Ser35-Asp32-Wat1-Asp228 as shown in Fig. 2D. It seems likely that this network plays an important role in the catalytic reaction in BACE1, but it is not required for the regulation of active/inactive states, since the network is conserved in both structures.

Catalytic site and Asp-bound water molecule (Wat1). Figure 3A to D show electron density maps in the vicinity of catalytic water at pH 7.0, 5.0, 4.5, and 4.0. The electron densities of Wat1 at pH 7.0 and 5.0 were clearly detected, with B factors of 24.4 Å2 and 26.2 Å2 and electron density levels (2Fo-Fc composite omit map) of 4.1 {sigma} and 2.8 {sigma}, respectively (Fig. 3A and B). In the structure at pH 4.5, Wat1 had a slightly weak electron density, with a value of 1.9 {sigma} and a B factor of 27.7 Å2, even though the X-ray data for this crystal were of better quality and higher resolution than that at pH 5.0 (Fig. 3C). A kinetics study of BACE1 for the substrate and inhibitor at pH 4.5 indicated that the displacement of Wat1 occurs during inhibitor binding but that no Wat1 replacement occurs during substrate binding (21). This observation implies that the hydrated form is still dominant at pH 4.5, which is consistent with our result that Wat1 in the structure at pH 4.5 gave the electron density that it did. In the case of the structure at pH 4.0, an omit map showed a weak electron density at the corresponding position of Wat1 (data not shown). However, after several refinements with the model, including Wat1 coordination, the Wat1 molecule exhibited a high B factor (more than 60 Å2) and a weak electron density level (Fig. 3D). These observations clearly show a low occupancy or high disorder of Wat1 at this position at pH 4.0.


Figure 3
View larger version (34K):
[in this window]
[in a new window]

  		FIG. 3. Comparison of the 2Fo-Fc electron density (light blue) and 2Fo-Fc composite omit (dark blue) maps of the vicinity of Asp32 and Asp228 at pHs of 7.0 (A), 5.0 (B), 4.5 (C), and 4.0 (D). The electron densities are contoured at 1.5 {sigma}. (E) pH dependence of OM99-2 binding as determined by use of the Biacore T100 system. The responses (in resonance units [RU]) were recorded as a function of time.

One of the possible reasons for these observations is that the conformation of the structure at pH 4.0 might be different from that of the structure at pH 5.0. However, this possibility seems unlikely, because the RMS deviation and structural comparison to the adjacent Asp pair do not show significant structural changes. Another possibility is that the protonation states of the Asp pair are changed with a varying pH. The predicted pKa values for Asp32 and Asp228 in the structure at pH 4.0 calculated by the program PROPKA (http://www.propka.chem.uiowa.edu) were 9.48 and 3.81, respectively. Considering these results, the fraction of diprotonated species at pH 4.0 is expected to be higher than that at a pH level greater than 4.0. The monoprotonate form is always accompanied by the presence of two strong hydrogen bonds between the Asp pair and the H bond donor (Wat1 or inhibitor). However, such H bond interactions may not exist or may decrease in the diprotonated form in acidic conditions, resulting in the destabilization of Wat1.

It has been proposed that a transition between the hydrated and dehydrated forms is also present in the apo form, and Wat1 might be exchanged with other water molecules, since some inhibitors bind only to the dehydrated form of human immunodeficiency virus protease (25), and the same mechanism probably operates in BACE1 (32). Thus, it seems likely that there is a pH-dependent equilibration between the hydrated form and the dehydrated form, and the dehydrated form is expected to be predominant at a pH level less than 4.0. Furthermore, the structure at pH 4.0 is almost identical to the structure at pH 4.5, including the conserved hydrogen-bonding network and the conformation of the Asp pair, although Wat1 is disordered. In addition, kinetics mechanism studies suggest that the enzyme returns to the free form, which still lacks the Wat1 molecule (dehydrated form), and that the formation of a tetrahedral intermediate and the return to the rehydrated form may be rate-limiting steps during the catalytic cycle of BACE1 (32), pepsin (27), and human immunodeficiency virus protease (18). It is also possible, and even likely, that such rigidity in the active site confers an advantage in the rehydration process or assists in the formation of the tetrahedral intermediate and the dehydrated form in the catalytic reaction.

Analysis of BACE1-inhibitor interaction. To examine whether or not these conformational changes are related to the binding affinity for inhibitors, the effect of pH on inhibitor (OM99-2) binding was examined by using a Biacore T100 instrument (Fig. 3E). Inhibitor binding was abolished at the range of pH 6.0 to 7.0, and lowering the pH to 5.0 resulted in the binding of inhibitor in a pH-dependent manner. Notably, the conformational switching and the acquisition of inhibitor binding appear to be simultaneous at pH 5.0, suggesting that conformational switching is important in allowing the inhibitor to bind.

Interestingly, the inhibitor binding signal was much stronger around pH 3.0 to 4.0 (Fig. 3E). Crystallographic studies for the inhibitor complex in BACE1 indicated that the inhibitor hydroxyl group plays the roles of hydrogen bond donor and acceptor with the Asp pair in the active site and that Wat1 replacement occurs during substrate binding (12, 13, 34). Also, kinetics studies suggested that Wat1 competitively inhibits the binding of inhibitor to the active site (21). Thus, the higher binding affinity at pH 3.0 to 4.0 is further evidence that Wat1 is disordered at a pH level less than 4.0.

Structure of OM99-2 complex at pH 5.0. We produced crystals of the BACE1/OM99-2 complex by cocrystallization. Two forms of crystals (P212121 and P21) in complex with OM99-2 were observed in the same batch. This P212121 form contains two molecules in the asymmetric unit, which is isomorphous with the previously reported P212121 form in complex with OM00-3 (13). The bound OM99-2 molecules were observed on all molecules in the asymmetric unit with virtually identical conformation. On the other hand, this P21 form is isomorphous to the previously reported P21 form in the OM99-2 complex (1FKN) (12). However, of the four molecules in the asymmetric unit, only two molecules show the strong electron densities of bound OM99-2. According to these observations, we applied the crystal of the P212121 form to the determination of the structure of the OM99-2 complex at pH 5.0. In addition, these observations might be due to the lower binding affinity of OM99-2 for BACE1 at a neutral pH. The parameters in the crystallization condition, such as the concentrations of OM99-2 and protein, the ratio of OM99-2 to protein, incubation time, and temperature, could facilitate the formation of crystal in complex with OM99-2, although BACE1 shows less binding affinity by Biacore analysis.

The OM99-2 complex in our study represents the electron density of bound OM99-2, which can define the position from P4 to P4' (Fig. 4A). The conformation and the interaction at P4-P2' positions are essentially the same as those of the complex with PDB accession no. 1FKN. However, it is noteworthy that the binding positions of the P3' and P4' regions in our study represent a novel binding mode for OM99-2 (Fig. 4B). The hydroxyl group of Tyr198 is hydrogen bonded to the carbonyl oxygen of P2' Ala, and the guanidyl group of Arg128 forms a new bond with the carbonyl oxygen of P3' Glu. The side chain of P4' Phe has hydrophobic interactions with Ile126 and Trp197, resulting in S4' being located in a new subsite defined by Pro70, Glu125, Ile126, Arg128, Arg195, Trp197, and Tyr198. These interactions and the conformation of bound OM99-2 are essentially the same as those of the previously reported OM00-3 complex, which has the same crystal form (P212121) (13). Furthermore, we confirmed that the conformation of OM99-2 at a neutral pH range shows the same conformation at pH 5.0 (unpublished data). Therefore, the conformation of bound OM99-2 could be interpreted as the result of the conformational variation or weak interactions at P3' and P4' positions in bound OM99-2. It has been reported that both P3' and P4' positions contribute little to the interaction with OM99-2, because poor electron densities at these regions were observed for the complex with the PDB accession no. 1FKN, reported previously (12, 33). Indeed, the P3' and P4' residues in our structure also show relatively high average B factor values, with 45.8 Å2 for the P3'-P4' region compared to 25.4 Å2 for the P4-P4' region.


Figure 4
View larger version (61K):
[in this window]
[in a new window]

  		FIG. 4. Structure of the OM99-2 complex. (A) 2Fo-Fc electron density (light blue) and 2Fo-Fc composite omit (dark blue) maps of bound OM99-2 at pH 5.0 contoured at 1.5 {sigma}. (B) Close-up view of bound OM99-2 at pH 5.0 (red) and 1FKN (at pH 7.4; blue).

In spite of these differences, the overall structure of the OM99-2 complex at pH 5.0 is essentially the same as that for the 1FKN complex, determined at pH 7.4. Quantitative comparison of these structures shows that the RMS difference of C{alpha} atoms between them is estimated to be 0.25 Å, but the values between the OM99-2 complex and apo structures clearly show large deviations, with values of 1.03 to 1.27 Å. Moreover, the conformation of the active site at pH 5.0, except for the region in the vicinity of P3' and P4' residues, is very similar to the 1FKN conformation. These findings could lead to the speculation that the binding of OM99-2 enables single conformations at any pH.


arrow
DISCUSSION
 
Our structural analysis strongly suggests that the conformational switching is highly correlated with inhibitor binding to the enzyme, shown by Biacore. In the inactive state, BACE1 has less binding affinity to the inhibitor, accounting for less substrate binding and the loss of enzymatic activity. In the active state, in which conformational switching occurs, the active site cleft accommodates the substrate as well as it accommodates the inhibitor, leading to activation. On the other hand, our studies also reveal the disorder of Wat1 at a pH less than 4.0. The Wat1 molecule is widely believed to be the nucleophile that attacks the carbonyl carbon of a peptide bond arranged in the active site. Thus, the disorder of Wat1 could lead to the lack of activity at a pH less than 4.0. According to these results, we propose that BACE1 has a dual regulatory mechanism of enzymatic activity at different pH values (Fig. 5). Consequently, BACE1 exhibits the pH profile for optimal activity at pH 4.5, at which both the presence of Wat1 and the structural switching of the active site cleft are observed.


Figure 5
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 5. Schematic representation of the regulatory mechanism of enzymatic activity in BACE1.

In fact, the enzyme-ligand interactions are directly affected by structural features, such as the configuration of the active site. The presence of the open conformation in the active state of BACE1 allows the substrate to enter easily, thereby modulating its activity. The structure of the OM99-2/BACE1 complex represents the single binding mode in the closed conformation at pH 5.0 and pH 7.4. Therefore, the conformational difference between the apo structures in the active and inactive states can elucidate important regions which are closely related to the regulatory mechanism of BACE1 activity. We expect that the conformation of the active state is convertible to the bound conformation, whereas the conformation of the inactive state blocks the conformational transition to the bound form, or sterically inhibits the OM99-2 binding to BACE1. As mentioned before, Tyr71 of the active structure shows significant displacement compared to that of the inactive structure, and this displacement is accompanied by a conformational change in the flap. Additionally, the side chain of Tyr71 in the inactive structure partly occupies the S1 subsite and is located near the position of the side chain of P1 Leu in superimposed OM99-2 at a distance of 2.9 Å (Fig. 6). Recent studies demonstrate that the conserved Tyr residue shows a self-inhibition mode in aspartic proteases (1, 9, 10), and Tyr71 in BACE1 also shows a self-inhibiting conformation in a simulation study (8). It is probable that the conformation of Tyr71 observed in the inactive structure sterically inhibits ligand binding to BACE1. It is also possible that Tyr71 blocks the movement of the flap to form the closed conformation. Also, the side chain of Arg128 occupies a space between Arg128 and Tyr198 in the inactive form. After OM99-2 binding, this side chain is flipped up compared with the apo structure, producing enough space for the ligand between Arg128 and Tyr198. Moreover, the conformation of this residue in the active structure shows an intermediate position between the inactive and OM99-2 complex structures (data not shown). These structural features could lead to speculation that Arg128 is related to a potential inhibitory conformation for ligand binding. On the other hand, the conformation of Arg128 in the 1SGZ apo structure was also flipped up like that in the OM99-2 complex (14). This may reflect one of the states in the equilibration between the open and closed conformations. Therefore, at any pH, inhibitors will find some BACE1 in an open conformation to which they will bind, resulting in a shift to the bound conformation.


Figure 6
View larger version (67K):
[in this window]
[in a new window]

  		FIG. 6. Composite surface representation in the active site cleft with superimposed OM99-2 (at pH 5.0; white-stick model). (A) Active structure (yellow). (B) Inactive structure (blue). (C) Superimposition of active and inactive structures (OM99-2 is shown at van der Waals radii in lower panels).

Recent X-ray crystallographic studies have also confirmed that aspartic proteases exhibit significant pH-dependent structural changes. The most dramatic change was reported for cathepsin D (22). In the inactive form of cathepsin D, the N terminus relocates into the active site and blocks the Asp residues. This observation is important for understanding the molecular mechanism of protease activation/inhibition. Moreover, the crystallographic study of rhizopuspepsin at different pH values reported that the pH-sensitive flexible region, showing an increase in the mobility of loop and a change in the water structure, is related to the stability of the protein structure (26). By comparison with these reports, our findings reveal a novel feature of aspartic proteases and the behavior of Asp-water and inhibitory conformations raise new possibilities for the mechanism of activation/inactivation of aspartic proteases.

Our study reveals that BACE1 has alternative conformations corresponding to the enzymatic activity. This could well have physiological significance. It has been reported that BACE1 is reinternalized from the cell surface to early endosomes and can recycle back to the cell surface via the trans-Golgi network (16, 38). At the plasma membrane, both BACE1 and APP are located in lipid rafts on the cell surface where, at a neutral pH (pH ~7), BACE1 is expected to show poor activity (15). BACE1 and APP follow similar trafficking routes and meet within endosomes (6, 15), where BACE1 is activated due to an acidic environment (pH 4.5 to 6) (38). When BACE1 is recycled to the cell surface, it becomes inactive again. Considering the long half-life of BACE1 and the recycling rate, BACE1 moves between the cell surface and the endosomal system many times through the course of its life span (16), showing an activation/inactivation transition during cellular trafficking. The conformational transition of active and inactive states in BACE1 represents interconvertibility under physiological conditions and plays an important role in the localization-dependent activity of BACE1. In fact, the pH-dependent conformational change makes BACE1 act in the appropriate subcellular compartment, such as endosomes, and cleave APP for further processing by presenilin. Furthermore, there is a possibility that the dehydrated form observed at pH 4.0 might have physiological significance, since local acidic compartments where BACE1 exists, for example, the endosomal system and trans-Golgi network (4, 16), could drop their pH value to less than 4.0. More-detailed studies should clarify the existence of such an acidic compartment and give insight into the significance of dehydrated BACE1.

In conclusion, our analysis reveals the molecular mechanism of activation/inactivation in BACE1. The control of BACE1 activity using these features, especially the displacement of the active site and the behavior of Wat1, provides insight into the novel mechanism of BACE1 and may improve efforts to design inhibitors. Indeed, several challenging problems regarding inhibitor design of BACE1 remain. Inhibitor size should be below 500 Da to allow the crossing of the blood-brain barrier and deep permeation of neurons to reach the lumen of endosomes. Selectivity against other aspartic proteases and specificity to other substrates should be necessary for desirable inhibitors to avoid secondary effects. The discovery of the active conformation in BACE1 raises a new possibility for the discovery of novel inhibitors. For example, compounds that can block the conformational transition to the active form also have the potential, as inhibitors, to slow down BACE1 activity. Furthermore, as a template for novel inhibitor designs, the most suitable structure of BACE1 should be selected depending on the condition of the cellular compartment where the inhibitor reaches and binds to BACE1. Accordingly, structural information on active BACE1 would be of value in designing further selective drugs.


arrow
ACKNOWLEDGMENTS
 
We thank Joanna Doumanis for her kind help in preparing the manuscript. We thank Hideyuki Miyatake and Naoshi Dohmae (Biomolecular Characterization Team, RIKEN) for support in the mail-in data collection system. We thank the staff of beamlines BL26B1, BL26B2, and BL44B2 at the SPring-8 Center for help in data collection and the staff of the Research Resources Center (RIKEN Brain Science Institute) for mass spectrometry analysis and DNA sequencing analysis.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Laboratory for Structural Neuropathology, RIKEN Brain Science Institute, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan. Phone: 81-48-462-9702. Fax: 81-48-462-4796. E-mail: nukina{at}brain.riken.jp Back

{triangledown} Published ahead of print on 31 March 2008. Back


arrow
REFERENCES
 
    1
  1. Andreeva, N., J. Dill, and G. L. Gilliland. 1992. Can enzymes adopt a self-inhibited form? Results of X-ray crystallographic studies of chymosin. Biochem. Biophys. Res. Commun. 184:1074-1081.[CrossRef][Medline]
    2. 2
  2. Andreeva, N. S., and L. D. Rumsh. 2001. Analysis of crystal structures of aspartic proteinases: on the role of amino acid residues adjacent to the catalytic site of pepsin-like enzymes. Protein Sci. 10:2439-2450.[CrossRef][Medline]
    3. 3
  3. Brünger, A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, R. J. Read, L. M. Rice, T. Simonson, and G. L. Warren. 1998. Crystallography & NMR System: a new software suite for macromolecular structure determination. Acta Crystallogr. D 54:905-921.[CrossRef][Medline]
    4. 4
  4. Capell, A., H. Steiner, M. Willem, H. Kaiser, C. Meyer, J. Walter, S. Lammich, G. Multhaup, and C. Haass. 2000. Maturation and pro-peptide cleavage of beta-secretase. J. Biol. Chem. 275:30849-30854.[Abstract/Free Full Text]
    5. 5
  5. Cascella, M., C. Micheletti, U. Rothlisberger, and P. Carloni. 2005. Evolutionarily conserved functional mechanics across pepsin-like and retroviral aspartic proteases. J. Am. Chem. Soc. 127:3734-3742.[CrossRef][Medline]
    6. 6
  6. Ehehalt, R., P. Keller, C. Haass, C. Thiele, and K. Simons. 2003. Amyloidogenic processing of the Alzheimer beta-amyloid precursor protein depends on lipid rafts. J. Cell Biol. 160:113-123.[Abstract/Free Full Text]
    7. 7
  7. Ermolieff, J., J. A. Loy, G. Koelsch, and J. Tang. 2000. Proteolytic activation of recombinant pro-memapsin 2 (pro-beta-secretase) studied with new fluorogenic substrates. Biochemistry 39:12450-12456.[CrossRef][Medline]
    8. 8
  8. Gorfe, A. A., and A. Caflisch. 2005. Functional plasticity in the substrate binding site of beta-secretase. Structure 13:1487-1498.[Medline]
    9. 9
  9. Gustchina, A., M. Li, L. H. Phylip, W. E. Lees, J. Kay, and A. Wlodawer. 2002. An unusual orientation for Tyr75 in the active site of the aspartic proteinase from Saccharomyces cerevisiae. Biochem. Biophys. Res. Commun. 295:1020-1026.[CrossRef][Medline]
    10. 10
  10. Gustchina, A., M. Li, S. Wunschmann, M. D. Chapman, A. Pomes, and A. Wlodawer. 2005. Crystal structure of cockroach allergen Bla g 2, an unusual zinc binding aspartic protease with a novel mode of self-inhibition. J. Mol. Biol. 348:433-444.[CrossRef][Medline]
    11. 11
  11. Hardy, J., and D. J. Selkoe. 2002. The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science 297:353-356.[Abstract/Free Full Text]
    12. 12
  12. Hong, L., G. Koelsch, X. Lin, S. Wu, S. Terzyan, A. K. Ghosh, X. C. Zhang, and J. Tang. 2000. Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor. Science 290:150-153.[Abstract/Free Full Text]
    13. 13
  13. Hong, L., R. T. Turner III, G. Koelsch, D. Shin, A. K. Ghosh, and J. Tang. 2002. Crystal structure of memapsin 2 (beta-secretase) in complex with an inhibitor OM00-3. Biochemistry 41:10963-10967.[CrossRef][Medline]
    14. 14
  14. Hong, L., and J. Tang. 2004. Flap position of free memapsin 2 (beta-secretase), a model for flap opening in aspartic protease catalysis. Biochemistry 43:4689-4695.[CrossRef][Medline]
    15. 15
  15. Huang, X. P., W. P. Chang, G. Koelsch, R. T. Turner III, F. Lupu, and J. Tang. 2004. Internalization of exogenously added memapsin 2 (beta-secretase) ectodomain by cells is mediated by amyloid precursor protein. J. Biol. Chem. 279:37886-37894.[Abstract/Free Full Text]
    16. 16
  16. Huse, J. T., D. S. Pijak, G. J. Leslie, V. M. Lee, and R. W. Doms. 2000. Maturation and endosomal targeting of beta-site amyloid precursor protein-cleaving enzyme. The Alzheimer's disease beta-secretase. J. Biol. Chem. 275:33729-33737.[Abstract/Free Full Text]
    17. 17
  17. Hussain, I., D. Powell, D. R. Howlett, D. G. Tew, T. D. Meek, C. Chapman, I. S. Gloger, K. E. Murphy, C. D. Southan, D. M. Ryan, T. S. Smith, D. L. Simmons, F. S. Walsh, C. Dingwall, and G. Christie. 1999. Identification of a novel aspartic protease (Asp 2) as beta-secretase. Mol. Cell. Neurosci. 14:419-427.[CrossRef][Medline]
    18. 18
  18. Hyland, L. J., T. A. J. Tomaszek, and T. D. Meek. 1991. Human immunodeficiency virus-1 protease. 2. Use of pH rate studies and solvent kinetic isotope effects to elucidate details of chemical mechanism. Biochemistry 30:8454-8463.[CrossRef][Medline]
    19. 19
  19. Kleywegt, G. J., and T. A. Jones. 1994. Detection, delineation, measurement and display of cavities in macromolecular structures. Acta Crystallogr. D 50:178-185.[CrossRef][Medline]
    20. 20
  20. Lin, X., G. Koelsch, S. Wu, D. Downs, A. Dashti, and J. Tang. 2000. Human aspartic protease memapsin 2 cleaves the beta-secretase site of beta-amyloid precursor protein. Proc. Natl. Acad. Sci. USA 97:1456-1460.[Abstract/Free Full Text]
    21. 21
  21. Marcinkeviciene, J., Y. Luo, N. R. Graciani, A. P. Combs, and R. A. Copeland. 2001. Mechanism of inhibition of beta-site amyloid precursor protein-cleaving enzyme (BACE) by a statine-based peptide. J. Biol. Chem. 276:23790-23794.[Abstract/Free Full Text]
    22. 22
  22. Merrick, J. M., A. Osman, J. Tsai, J. Quackenbush, P. T. LoVerde, and N. H. Lee. 2003. The Schistosoma mansoni gene index: gene discovery and biology by reconstruction and analysis of expressed gene sequences. J. Parasitol. 89:261-269.[CrossRef][Medline]
    23. 23
  23. Otwinowski, Z., and W. Minor. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276:307-326.[CrossRef]
    24. 24
  24. Patel, S., L. Vuillard, A. Cleasby, C. W. Murray, and J. Yon. 2004. Apo and inhibitor complex structures of BACE (beta-secretase). J. Mol. Biol. 343:407-416.[CrossRef][Medline]
    25. 25
  25. Piana, S., D. Sebastiani, P. Carloni, and M. Parrinello. 2001. Ab initio molecular dynamics-based assignment of the protonation state of pepstatin A/HIV-1 protease cleavage site. J. Am. Chem. Soc. 123:8730-8737.[CrossRef][Medline]
    26. 26
  26. Prasad, B. V., and K. Suguna. 2003. Effect of pH on the structure of rhizopuspepsin. Acta Crystallogr. D 59:1755-1761.[CrossRef][Medline]
    27. 27
  27. Rebholz, K. L., and D. B. Northrop. 1991. Slow step after bond-breaking by porcine pepsin identified using solvent deuterium isotope effects. Biochem. Biophys. Res. Commun. 176:65-69.[CrossRef][Medline]
    28. 28
  28. Roussel, A., and C. Cambillan. 1989. TURBO-FRODO, p. 77-78. In Silicon Graphics Geometry partner directory. Silicon Graphics, Mountain View, CA.
    29. 29
  29. Sardana, V., B. Xu, J. Zugay-Murphy, Z. Chen, M. Sardana, P. L. Darke, S. Munshi, and L. C. Kuo. 2004. A general procedure for the purification of human beta-secretase expressed in Escherichia coli. Protein Expr. Purif. 34:190-196.[CrossRef][Medline]
    30. 30
  30. Selkoe, D. J. 2001. Alzheimer's disease: genes, proteins, and therapy. Physiol. Rev. 81:741-766.[Abstract/Free Full Text]
    31. 31
  31. Sinha, S., J. P. Anderson, R. Barbour, G. S. Basi, R. Caccavello, D. Davis, M. Doan, H. F. Dovey, N. Frigon, J. Hong, K. Jacobson-Croak, N. Jewett, P. Keim, J. Knops, I. Lieberburg, M. Power, H. Tan, G. Tatsuno, J. Tung, D. Schenk, P. Seubert, S. M. Suomensaari, S. Wang, D. Walker, J. Zhao, L. McConlogue, and V. John. 1999. Purification and cloning of amyloid precursor protein beta-secretase from human brain. Nature 402:537-540.[CrossRef][Medline]
    32. 32
  32. Toulokhonova, L., W. J. Metzler, M. R. Witmer, R. A. Copeland, and J. Marcinkeviciene. 2003. Kinetic studies on beta-site amyloid precursor protein-cleaving enzyme (BACE). Confirmation of an iso mechanism. J. Biol. Chem. 278:4582-4589.[Abstract/Free Full Text]
    33. 33
  33. Turner, R. T., III, J. A. Loy, C. Nguyen, T. Devasamudram, A. K. Ghosh, G. Koelsch, and J. Tang. 2002. Specificity of memapsin 1 and its implications on the design of memapsin 2 (beta-secretase) inhibitor selectivity. Biochemistry 41:8742-8746.[CrossRef][Medline]
    34. 34
  34. Turner, R. T., III, L. Hong, G. Koelsch, A. K. Ghosh, and J. Tang. 2005. Structural locations and functional roles of new subsites S5, S6, and S7 in memapsin 2 (beta-secretase). Biochemistry 44:105-112.[CrossRef][Medline]
    35. 35
  35. Ueno, G., H. Kanda, R. Hirose, K. Ida, T. Kumasaka, and M. Yamamoto. 2006. RIKEN structural genomics beamlines at the SPring-8; high throughput protein crystallography with automated beamline operation. J. Struct. Funct. Genomics 7:15-22.[CrossRef][Medline]
    36. 36
  36. Vassar, R., B. D. Bennett, S. Babu-Khan, S. Kahn, E. A. Mendiaz, P. Denis, D. B. Teplow, S. Ross, P. Amarante, R. Loeloff, Y. Luo, S. Fisher, J. Fuller, S. Edenson, J. Lile, M. A. Jarosinski, A. L. Biere, E. Curran, T. Burgess, J. C. Louis, F. Collins, J. Treanor, G. Rogers, and M. Citron. 1999. Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science 286:735-741.[Abstract/Free Full Text]
    37. 37
  37. Vassar, R. 2001. The beta-secretase, BACE: a prime drug target for Alzheimer's disease. J. Mol. Neurosci. 17:157-170.[CrossRef][Medline]
    38. 38
  38. Walter, J., R. Fluhrer, B. Hartung, M. Willem, C. Kaether, A. Capell, S. Lammich, G. Multhaup, and C. Haass. 2001. Phosphorylation regulates intracellular trafficking of beta-secretase. J. Biol. Chem. 276:14634-14641.[Abstract/Free Full Text]
    39. 39
  39. Wong, H. K., T. Sakurai, F. Oyama, K. Kaneko, K. Wada, H. Miyazaki, M. Kurosawa, B. De Strooper, P. Saftig, and N. Nukina. 2005. β-Subunits of voltage-gated sodium channels are novel substrates of β-site amyloid precursor protein-cleaving enzyme (BACE1) and gamma-secretase. J. Biol. Chem. 280:23009-23017.[Abstract/Free Full Text]
    40. 40
  40. Xiong, B., X. Q. Huang, L. L. Shen, J. H. Shen, X. M. Luo, X. Shen, H. L. Jiang, and K. X. Chen. 2004. Conformational flexibility of beta-secretase: molecular dynamics simulation and essential dynamics analysis. Acta Pharmacol. Sin. 25:705-713.[Medline]
    41. 41
  41. Yan, R., M. J. Bienkowski, M. E. Shuck, H. Miao, M. C. Tory, A. M. Pauley, J. R. Brashier, N. C. Stratman, W. R. Mathews, A. E. Buhl, D. B. Carter, A. G. Tomasselli, L. A. Parodi, R. L. Heinrikson, and M. E. Gurney. 1999. Membrane-anchored aspartyl protease with Alzheimer's disease beta-secretase activity. Nature 402:533-537.[CrossRef][Medline]
    42. 42
  42. Yu-dong, L., I. Harvey, G. Yuan-xin, Z. Chao-de, H. Yi-zong, F. Hai-fu, S. S. Hasnain, and Q. Hao. 1994. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D 50:760-763.[CrossRef][Medline]

Molecular and Cellular Biology, June 2008, p. 3663-3671, Vol. 28, No. 11
0270-7306/08/$08.00+0     doi:10.1128/MCB.02185-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Sakurai, T., Kaneko, K., Okuno, M., Wada, K., Kashiyama, T., Shimizu, H., Akagi, T., Hashikawa, T., Nukina, N. (2008). Membrane microdomain switching: a regulatory mechanism of amyloid precursor protein processing. JCB 183: 339-352 [Abstract] [Full Text]  
  • Vanoni, O., Paganetti, P., Molinari, M. (2008). Consequences of Individual N-glycan Deletions and of Proteasomal Inhibition on Secretion of Active BACE. Mol. Biol. Cell 19: 4086-4098 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shimizu, H.
Right arrow Articles by Nukina, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shimizu, H.
Right arrow Articles by Nukina, N.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»