Article A Bacteriophage Capsid Protein Provides a General Amyloid Interaction Motif (GAIM) That Binds and Remodels Misfolded Protein Assemblies Rajaraman Krishnan 1 , Haim Tsubery 1 , Ming Y. Proschitsky 1 , Eva Asp 1 , Michal Lulu 1 , Sharon Gilead 1 , Myra Gartner 1 , Jonathan P. Waltho 2, 3 , Peter J. Davis 3 , Andrea M. Hounslow 3 , Daniel A. Kirschner 4 , Hideyo Inouye 4 , David G. Myszka 5 , Jason Wright 1 , Beka Solomon 6 and Richard A. Fisher 1 1 - Neurophage Pharmaceuticals, 222 Third Street, Suite 3120, Cambridge, MA 02142, USA 2 - Faculty of Life Sciences, Manchester Institute of Biotechnology, The University of Manchester, 131 Princess Street, Manchester M1 7DN, UK 3 - Department of Molecular Biology and Biotechnology, University of Sheffield, Firth Court, Western Bank, Sheffield S10 2TN, UK 4 - Department of Biology, Boston College, Chestnut Hill, MA 02467, USA 5 - Biosensor Tools LLC, 1588 East Connecticut Drive, Salt Lake City, UT 84103, USA 6 - Department of Molecular Microbiology and Biotechnology, Tel Aviv University, Tel Aviv 69978, Israel Correspondence to Rajaraman Krishnan and Richard A. Fisher: [email protected]; [email protected] http://dx.doi.org/10.1016/j.jmb.2014.04.015 Edited by R. Wetzel Abstract Misfolded protein aggregates, characterized by a canonical amyloid fold, play a central role in the pathobiology of neurodegenerative diseases. Agents that bind and sequester neurotoxic intermediates of amyloid assembly, inhibit the assembly or promote the destabilization of such protein aggregates are in clinical testing. Here, we show that the gene 3 protein (g3p) of filamentous bacteriophage mediates potent generic binding to the amyloid fold. We have characterized the amyloid binding and conformational remodeling activities using an array of techniques, including X-ray fiber diffraction and NMR. The mechanism for g3p binding with amyloid appears to reflect its physiological role during infection of Escherichia coli, which is dependent on temperature-sensitive interdomain unfolding and cis–trans prolyl isomerization of g3p. In addition, a natural receptor for g3p, TolA-C, competitively interferes with Aβ binding to g3p. NMR studies show that g3p binding to Aβ fibers is predominantly through middle and C-terminal residues of the Aβ subunit, indicating β strand–g3p interactions. A recombinant bivalent g3p molecule, an immunoglobulin Fc (Ig) fusion of the two N-terminal g3p domains, (1) potently binds Aβ fibers (fAβ) (KD = 9.4 nM); (2); blocks fAβ assembly (IC50 ~ 50 nM) and (3) dissociates fAβ (EC50 = 40–100 nM). The binding of g3p to misfolded protein assemblies is generic, and amyloid-targeted activities can be demonstrated using other misfolded protein systems. Taken together, our studies show that g3p(N1N2) acts as a general amyloid interaction motif. © 2014 Elsevier Ltd. All rights reserved. Introduction In neurodegenerative diseases, such as Alzheimer's disease (AD), Parkinson's disease and numerous non-AD dementias, there is progressive accumulation of extracellular and intracellular misfolded protein deposits. The brain deposits of these misfolded proteins contain amyloid fibrils characterized by a fiber core composed of β strands running 0022-2836/© 2014 Elsevier Ltd. All rights reserved. perpendicular to the axis of the fibers, the product of a complex assembly pathway [1,2]. A variety of therapies targeting specific protein aggregation pathways, such as those for amyloid-β, the microtubule-associated protein tau and pre-synaptic protein α-synuclein, are in all stages of drug development, and these drug candidates include vaccines, monoclonal antibodies (MAbs), antisense oligonucleotides and small molecules [3–6]. Recent studies J. Mol. Biol. (2014) 426, 2500–2519 2501 G3P remodels amyloid fibers (a) (b) 8x106 Alexa488 FI (A.U.) Response units (A.U.) 1000 750 500 250 0 6 4x106 2x106 0 4000 8000 12000 10-10 Time (s) (c) 6x106 10-9 M13 (M) 10-8 (d) 5x104 400 4x104 ThT Fl (A.U.) ThT Fl (A.U.) 500 3x104 300 200 100 2x104 0 0 1 2 3 4 5 6 480 Time (days) (e) 500 520 540 560 Wavelength (nm) Cellulose acetate membrane Nitrocellulose membrane fAβ+6.5x1013/ml M13 fAβ+1.3x1013/ml M13 fAβ+1.3x1012/ml M13 fAβ+1.3x1011/ml M13 fAβ only [T0] fAβ+1.3x1013/ml M13 Aβ monomer only 100 67 44 30 20 100 67 44 30 20 Aβ (ng/well) Fig. 1. M13 binds to and remodels Aβ fibers (fAβ). (a) SPR binding profiles show M13 binding to immobilized fAβ (black line; estimated KD ~ 3.6 nM) but not Aβ monomers (red line). (b) Concentration-dependent binding of M13 to fAβ measured by co-sedimentation of Alexa488-M13 with fAβ (estimated KD ~ 7 nM). (c) Prolonged incubation at 37 °C of aggregated Aβ (2.5 μM) with M13 (1 × 10 12 phage/ml) reduces ThT binding fluorescence at 450 nm (red line). Aβ fibers incubated under identical conditions show no reduction in ThT binding (black line). (d) M13‐mediated remodeling of fAβ monitored by ThT binding fluorescence spectra. fAβ (2.5 μM) was pre‐incubated with (red continuous line) or without (black continuous line) 1 × 10 12 M13 phage/ml for 3 days at 37 °C, and the fluorescence spectra were recorded in the presence of 5‐fold excess ThT. Fluorescence spectra of buffer alone (grey broken line) and the fAβ‐M13 mixture without any pre‐incubation (red broken line) are plotted for comparison. (e) Concentration-dependent remodeling activity of M13 monitored by the CA filter retention assay. Aggregated Aβ (2.5 μM) was co‐incubated with varying concentrations of M13 for 3 days at 37 °C. Samples were serially diluted and spotted on a CA membrane, washed under vacuum and probed with MAb 6E10. A nitrocellulose membrane blot, which non-selectively retains total Aβ from each reaction, is shown as a control. show two or more species of misfolded protein can underlie neurodegenerative diseases, which potentially complicates understanding disease patho- physiology and proving single target (i.e., single misfolded protein) approaches [7,8]. For example, brains from AD patients contain two cardinal protein 2502 G3P remodels amyloid fibers deposits, extraneuronal Aβ plaques and neurofibrillary tangles of tau, and frequently a third deposit of aggregated α-synuclein, normally associated with Parkinson's disease [9]. Because a combination of these three neuropathologies could contribute to disease progression, approaches that can target amyloid deposition generically could offer a therapeutic advantage over single target approaches. A serendipitous discovery showed that filamentous bacteriophage M13 can mediate disruption of amyloid assemblies consisting of the non-amyloid component of amyloid plaque (NAC), a hydrophobic region (residues 65–95) of human α-synuclein [10,11]. The NAC peptide shares sequence homology to the nucleating sequences of several aggregates, including Aβ and the mammalian prion sequence [12–14]. Here, we confirm and extend this finding by characterizing the activities of M13 on multiple amyloid targets. We show that highly purified preparations of native M13 potently and broadly mediate binding to and disruption of a variety of misfolded protein assemblies, including Aβ, α-synuclein, tau and yeast prion Sup35. Characterization of amyloid fiber binding and remodeling indicates that the bacteriophage minor capsid protein, gene 3 protein (g3p), is critical for this activity. We functionally define the two N-terminal domains of g3p that facilitate binding and disruption of amyloids as a general amyloid interaction motif (GAIM). Results Filamentous bacteriophage M13 mediates amyloid remodeling activity through binding interactions To characterize the interactions between M13 and misfolded protein aggregates, we used fAβ as a model aggregate. Surface plasmon resonance (SPR) binding assays show that highly purified preparations of native M13 potently bind immobilized fAβ but not Aβ monomers (Fig. 1a). The apparent equilibrium dissociation constant (KD) calculated from SPR binding studies is 3.6 nM (Table 1), a KD that is similar to fAβ binding of anti-N-terminal Aβ Table 1. SPR kinetic parameters of commercial antibodies, M13, GAIM-based molecules binding to fAβ Analyte fAβ binding parameter kon (M−1 s−1) M13 3D6 [16] 6E10a [15] G3P Ig-G3P 2.6 6.6 2. 1 1.5 1.7 × × × × × 3 10 104 105 103 104 koff (s−1) 9.2 2 2.9 2.4 1.6 × × × × × −6 10 10−4 10−4 10−4 10−4 KD (nM) 3.6 3 1.4 150 9.4 a Aβ1–40 fibers were used to calculate the binding parameters for 6E10. MAbs, such as 3D6 and 6E10 [15,16]. However, unlike these MAbs, the fAβ binding association (kon) and dissociation (koff) rates of M13 binding differ significantly. M13 binding to fAβ is ~ 100-fold slower and effectively irreversible (koff = 9.2 × 10 − 6 /s) when compared to the kinetics determined for the anti-Aβ MAbs (Table 1). We took advantage of the differential sedimentation properties of fAβ and M13 to develop a quantitative solution binding assay (Fig. S1a). Fluorescently labeled M13 (Alexa488-M13) sediments at low centrifugal forces (7500g) only when in a complex with fAβ, while unbound Alexa488-M13 remains in the supernatant. Alexa488-M13 sedimentation depends on the concentration of fAβ present in the reaction (Fig. S1b) suggesting first-order binding kinetics. The apparent KD estimated by quantifying the pelleted Alexa488-M13:fAβ complex (KD = 7 nM) is similar to that calculated using SPR assays of M13 binding to immobilized fAβ (Fig. 1b). To investigate whether M13 binding to fAβ alters the conformation of the fibers, we utilized both thioflavin T (ThT) binding [17] and cellulose acetate (CA) filter retention assays [18,19]. ThT binds cross-β sheet fibers to produce characteristic high fluorescence emission maxima around 485 nm, but ThT binds poorly to non-fibrillar aggregates and monomeric peptides [20,21]. Based on the fluorescence intensity, ThT binding to the M13:fAβ complex was unchanged at 0, 4 or 12 h, which shows that M13 binding does not alter the fiber conformation for incubation times ≤ 24 h (Fig. 1c). However, longer incubation times (2–5 days) of the M13:fAβ complex lead to significant loss of ThT fluorescence, indicating M13-mediated structural changes to the amyloid fold (Fig. 1c and d). We used a CA filter retention assay [19] to measure loss of amyloid conformation (remodeling) mediated by M13. Unlike nitrocellulose membranes that retain equally all forms of Aβ (monomers, oligomers and fibrillar aggregates), CA membranes with 0.2 μM pores retain amyloid fibers, which are quantitatively detected by anti-N-terminal Aβ MAbs, such as 6E10. Non-fibrillar aggregates and peptides flow through the membrane. CA retention assays show that M13 potently remodels Aβ fiber structure in a concentration-dependent manner (EC50 = 1 to 5 × 10 12 phage particles per milliliter or 1.7–8.5 nM particle concentration), consistent with the apparent M13:fAβ binding KD (Table 1). The reductions in CA membrane retention of M13-treated fAβ occurs only after 24 h (Fig. S2a), consistent with the ThT amyloid dissociation assay data (Fig. 1c). By comparison with fAβ quantification at each time point on nitrocellulose (Fig. 1e), we show that the apparent activity measured by CA retention is not due to M13 interference with 6E10 binding. To more directly measure M13-mediated fAβ remodeling, we measured the fiber X-ray diffraction pattern of the M13:fAβ complex at the end of 24 and 48 h. Since amyloid fibers and M13 have distinct G3P remodels amyloid fibers diffraction patterns, changes in conformation of either structure can be detected differentially. These diffraction studies show that the amyloid fibers, but not the 2503 M13 filaments, gradually lose their total β sheet content over 48 h, an effect consistent with fiber remodeling (Fig. S2b and c). Fig. 2. M13 remodels a variety of amyloid fibers with unrelated misfolded protein subunits. (a) CA filter retention assays showing remodeling of ftau (2.5 μM), fAS (2.5 μM), fMTBR (2.5 μM) and fNM (2.5 μM) in the presence of 1 × 10 13 phage/ ml. The tau, α-synuclein, myc-tagged MTBR and his-tagged Sup35NM fibers retained on the membrane were detected using tau5 (anti-Tau, 1:10,000 dilution), 4D6 (anti-α-synuclein, 1:10,000), anti-myc (9B11, 1:10,000) and MMS-156P (anti-his tag, 1:15,000) antibodies, respectively. Monomeric proteins were directly diluted from a urea (8 M) to prevent oligomerization and non-specific binding. Control nitrocellulose filter blots of each corresponding remodeling reaction are shown.(b) Detergent solubility of fNM treated with and without 1 × 10 13 phage/ml for 3 days at 37 °C. Samples were incubated with SDS (1–6%) for 6 h and centrifuged at 30,000 rpm. The fluorescence in the supernatant was used to calculate soluble Sup35NM release. (c) SDS solubility of α-synuclein fibers treated with and without 1 × 10 13 phage/ml for 3 days at 37 °C. Samples were incubated with SDS (1–6%) for 1 h and then centrifuged at 30,000 rpm. The fluorescence in the supernatant was used to calculate soluble α-synuclein released. (d–f) A time course study showing increased solubility of fluorescently labeled α-synuclein fibers after co-incubation with M13. The phage slowly remodels the fibers to a non-amyloid conformer with concomitant increasing sensitivity to SDS solubilization. (g) Solubility of fAβ fibers treated with 1 × 10 13 phage/ml in 2%, 3% and 4% RIPA buffer without SDS. fAβ:M13 complexes were incubated with 2–4% RIPA buffer without SDS for 30 min and centrifuged at 40,000 rpm for 10 min. The supernatants (soluble fraction) were spotted on a nitrocellulose membrane and quantitated using MAb 6E10 for detection and a fluorescent secondary antibody quantified by densitometry. 2504 G3P remodels amyloid fibers Fig. 2 (continued). Taken together, these studies directed to measure amyloid fiber binding and dissociation show that M13 potently binds fAβ and disrupts the amyloid conformation. The binding event is completed in about 3–4 h, while amyloid fiber remodeling takes place more slowly over several days. M13 recognizes the canonical amyloid conformation We tested for M13-mediated amyloid fiber remodeling activity with three non-Aβ fiber assemblies, all with unrelated primary sequences and a broad range of monomer molecular masses (15–440 kDa). The three proteins, tau, α-synuclein and the yeast Sup35 prion domain, NM, have different lengths of disordered regions and aggregate by either hydrophobic (tau and α-synuclein) or glutamine-asparagine-rich sequences (Sup35NM). Regardless of the protein species, SPR binding, assembly inhibition and CA filter retention assays show that all three amyloid fibers are targets for M13-mediated binding and remodeling (Fig. 2a and Table 2) suggesting that interaction with the phage is conformation dependent and is not protein primary sequence dependent. To determine if amyloid remodeling following M13 binding produces low molecular weight (b 100 kDa) products, we filtered M13-bound aggregates (incubated for 3 days) through size-exclusion membranes with four different molecular mass cutoffs (10, 30, 50 and 100 kDa). No detectable Aβ passed through the membranes, suggesting that phage treatment does not release low molecular weight aggregates, oligomers or monomers (data not shown). 2505 G3P remodels amyloid fibers Table 2. Comparison of activities of GAIM-based molecules with α-synuclein, tau and Aβ assemblies GAIM construct M13 G3P Ig-G3P Binding Assembly inhibition KD (nM) fAβ 3.6 150 9.4 ftau/fMTBR ND ND 63/10.4 Remodeling IC50 (nM) fAS 6 360 49 Aβ ~3 20–100 50 Since M13 binding appears to gradually alter β sheet content, we measured the differential detergent solubility of amyloids with or without M13 co-incubation to uncover alteration of the fiber conformation. Figure 2b and c show M13-mediated changes in solubility of fluorescently labeled Sup35NM and α-synuclein fibers, respectively. Sup35NM monomers were co-assembled with 10% fluorescently labeled cy5-labeled Sup35NM monomers. Likewise, α-synuclein was co-assembled with 10% Alexa488-labeled α-synuclein monomers. Under these conditions, the assembly kinetics of the unlabeled and the 10% labeled proteins were identical (data not shown). For both fibers, M13 treatment significantly increases detergent solubility. Further, this increase in fiber solubility is time dependent (Fig. 2d–f). Similar solubility studies with fAβ were not performed due to higher inherent detergent solubility of these fibers to SDS. Instead, we performed detergent solubility assays of fAβ with SDS-free radioimmunoprecipitation assay (RIPA) buffer (containing 0.5% sodium deoxycholate and 1% NP-40). Quantification of solubilized Aβ shows that M13 treatment significantly increases fAβ sensitivity to detergents (Fig. 2g). Capsid protein g3p mediates M13:Aβ interactions To determine whether hydrophobic interactions could be involved in M13 binding to amyloid fibers, we measured the effect of temperature changes for M13-mediated fAβ binding and for fiber remodeling. Pre-heating M13 prior to activity tests shows that M13 binding to and remodeling of fAβ increases with increasing temperature (Fig. 3a and b), suggesting that hydrophobic interactions are critical. Importantly, the temperature approximating the melting temperature (Tm) that promotes separation of the two N-terminal domains (N1 and N2) of the minor capsid protein g3p (Tm = 48.1 °C) [22] is within the temperature transition (Fig. 3a) for greater Aβ fiber binding and disaggregation activities. To investigate more directly the role of g3p for amyloid targeting, we measured the amyloid binding activities of phage treated to disable g3p function. The first M13 variant was prepared by incubating phage at 95 °C for 10 min, which denatures g3p but does not affect the filament architecture (Fig. S3a) [22–24]. The second g3p-disabled variant is a tau/MTBR 10 ND ND EC50 (nM) AS 10–20 100–150 ND fAβ 1.7–8.5 N2500 40–100 ftau/fMTBR 1–10 N 2000 150 fAS 1–10 ND ND protease-treated preparation of M13 with selectively cleaved g3pN1N2 domains (described in Fig. S3b). Consistent with g3p having a central role for M13:fAβ interactions, denaturation or proteolysis of g3p significantly reduces both M13 binding to fAβ (Fig. 3c) and fAβ remodeling activity (Fig. 3d). To directly demonstrate g3p targeting of amyloid, we expressed and purified a functionally active recombinant soluble g3p fragment (called G3P) containing g3p N1 and N2 domains and the glycinerich hinge sequence connecting them [23,24]. G3P competes with M13 for binding to fAβ, as measured by the binding competition assay (Table 1 and Fig. 3e). G3P also shows potent fAβ binding by SPR (Table 1 and Fig. S8c). To determine if G3P could block Aβ aggregation, we co-incubated G3P with Aβ under fiber assembly conditions. Based on ThT binding and circular dichroism analyses, G3P effectively blocks fiber assembly with 50% inhibitory concentration (IC50) = 20–100 nM (Fig. 3f and Fig. S4). Valence affects G3P potency for amyloid interactions Compared to M13, G3P shows reduced potency for fAβ binding and fiber remodeling (Table 2 and 3). As shown for M13, G3P binds Aβ fibers over monomers, but the fiber binding (K D = 100– 150 nM) is 25- to 40-fold less potent than the intact phage. G3P did not effectively remodel Aβ fibers in the 3-day remodeling assay, even at concentrations as high as 2.5 μM (Fig. 4). These quantitative and qualitative differences between G3P and native phage are addressed below. Because g3p is oligomeric (n = 3–5) at the phage tip, the potency differences between M13 and G3P for binding and remodeling activities could be due to valence effects. To address this possibility, we created two multivalent G3P molecules: (1) dimeric G3P made from an immunoglobulin (Ig) Fc-G3P genetic fusion and (2) a multivalent G3P conjugate made using streptavidin (SA) complexed with biotin-G3P. Ig-G3P was expressed from mammalian cells and purified using protein A chromatography. The SA-G3P multimer was prepared by biotinylation of G3P. The G3P-biotin conjugate was reacted with SA to form G3P multimers (n = 2–4), which were purified by size-exclusion chromatography. Both multimeric 2506 G3P remodels amyloid fibers (a) (b) Bound M13 Fl (A.U.) 1.2x104 200 100 9.0x103 50 25 12.5 Aβ(ng/well) fAβ only open 6.0x103 closed 3.0x103 0 15 30 45 60 Temperature (°C) (c) Alexa488-M13 competed (%) Alexa488-M13 competed (%) (d) 100 100 200 80 100 50 25 12.5 Aβ (ng/well) fAβ only 60 +ArgC treated M13 40 + M13 20 0 (e) 0 10 00 00 13 M 3P G 0 G3P [nM] no co m pe tit o r 0 25 10 20 50 10 0 40 75 10 60 1 80 100 0. 1 fAβ assemblyinhibition(%) (f) Fig. 3. M13 g3p is necessary and sufficient for amyloid targeting by M13. (a) Binding of M13 to 10 μM fAβ at different temperatures monitored by the co‐sedimentation assay. The data were fitted to a sigmoidal curve and the estimated Tm is 46 ± 3 °C. The cartoons of g3pN1 and N2 domains show temperature dependency of interdomain unfolding. (b) Remodeling activity of M13 is temperature dependent. Aggregated Aβ (2.5 μM) was co‐incubated with M13 (1 × 10 13 phage/ml) for 3 days at 4 °C and 37 °C and the loss of fiber measured by the CA membrane retention assay. (c) Binding competition of various M13 variants monitored by the fluorescent co-sedimentation assay (see also Fig. S1c). Alexa488‐M13 (5 × 10 11 phage/ml) was co‐incubated for 1 h with 15‐fold excess of preparations of M13 that had been heat inactivated for 10 min at 95 °C, treated with endoproteinase ArgC, or native phage was pre‐incubated at either 37 °C or 50 °C, respectively. Binding competition was measured after centrifugation (7500g) to pellet M13:fAβ complexes, and the loss of fluorescence from the pellets was used to calculate % binding competition. (d) fAβ remodeling activity of M13 is dependent on intact g3p. Aggregated Aβ (2.5 μM) was co‐incubated with endoproteinase ArgC‐treated M13 (1 × 10 13 phage/ml) for 3 days at 37 °C and measured for loss of amyloid in the CA filter retention assay. (e) Binding competition of G3P monitored by the fluorescence co-sedimentation assay. Alexa488‐M13 (5 × 10 11 phage/ml) was co-incubated with 50-fold excess of soluble G3P or unlabeled M13 followed by measurement of Alexa488 fluorescence in the co-sedimentation binding assay. (f) G3P inhibits Aβ assembly in a concentration-dependent manner. Plot showing inhibition of fiber assembly monitored by ThT binding fluorescence. 2507 G3P remodels amyloid fibers Table 3. Valence of GAIM and activity Molecule M13 G3P Ig-G3P Valence (G3P) fAβ binding (SPR), KD (nM) fAβ binding (co-sedimentation), KD (nM) (estimate) fAβ remodeling, EC50 (nM) Aβ assembly inhibition, IC50 (nM) 3–5 1 2 3.6 150 9.4 7 250 ND 1.7–8.5 N2500 40–100 ~3 20–100 50 constructs were tested for fAβ remodeling. The Ig-G3P and SA-G3P constructs exhibited relatively potent amyloid fiber remodeling activities compared to G3P (Fig. 4), suggesting that g3p valence plays a role in amyloid fiber binding and remodeling activities. To directly demonstrate that the G3P dimer created by Ig-G3P exhibits higher avidity to amyloid fibers than G3P monomers, we performed SPR binding and CA filter retention experiments. Table 3 shows that Ig-G3P has significantly increased binding (Ig-G3P apparent KD = 9.4 nM) and fiber remodeling (EC50 = 40–100 nM) activities compared to the G3P monomer (G3P binding KD = 150 nM; EC50 N 2500 nM). Taken together, the binding and fiber dissociation assays of both native phage and the G3P multimers indicate that multimerization of G3P increases amyloid targeting activities. To show that the Ig-G3P fusion, like the phage, can potently bind amyloids assembled from unrelated proteins, we measured binding to assembled fibers of tau [both full-length tau and 4-repeat tau (MTBR; C-terminally myc-tagged microtubule binding region of tau protein, residues 243–375)] and of full-length 200 100 50 25 12.5 α-synuclein. Ig-G3P potently binds tau fibers (apparent KD = 63 nM for full-length tau fibers and apparent KD = 10.4 nM for MTBR fibers) and α-synuclein fibers (apparent KD = 49 nM) (Table 2), which supports that M13 generic amyloid fiber binding is mediated by g3p. Ig-G3P does not bind non-amyloid protein aggregates and blocks binding of amyloid-specific polythiophenes To examine the targeting specificity of G3P, we carried out Ig-G3P binding studies with casein, elastin, gelatin, bovine and human collagen, bovine serum albumin (BSA) and monomeric α-synuclein. This group includes natively disordered proteins, denatured proteins, glycine-proline-rich proteins and proteins with exposed hydrophobic surfaces. Ig-G3P does not avidly bind to any of these test proteins over a broad concentration range (0–225 nM), consistent with the hypothesis that g3p specifically recognizes proteins with an amyloid conformation (Fig. 5a and b and Fig. S5). Aβ (ng/well) fAβ only +2.5µM G3P G3P (monomer) + 0.5µMG3P + 800nM Ig-G3P + 8nM Ig-G3P Ig-G3P (dimer) + 2.5µM SA-G3P + 0.36µM SA (no G3P) + 0.5µM SA-G3P SA-G3P (multimer) + 0.072µM SA (no G3P) Fig. 4. G3P valence is important for Aβ fiber remodeling. CA filter retention assay shows lack of remodeling effect of monomeric G3P after 3 days of co-incubation with fAβ at 37 °C. Multimeric G3P (Ig-G3P dimer; SA-G3P, n = 2–4) shows a dose-dependent remodeling of fAβ under identical conditions. 2508 G3P remodels amyloid fibers (a) (b) 140 Fluorescence Intensity (A.U.) (c) Relative absorbance at 450nm (%) 120 100 80 60 40 120 100 80 60 40 20 20 0 0 (d) 35000 100 30000 fAβ 25000 20000 15000 10000 Aβ monomer 5000 Fluorescence (%) Relative absorbance at 450nm (%) 140 80 60 40 20 0 0 540 580 620 660 Emission Wavelength (nm) 0 1 2 3 4 Ig-G3P Concentration [μM] Fig. 5. G3P binding is specific for amyloid aggregates. (a) Binding of Ig-G3P to various non-amyloid proteins measured by ELISA assays. Ig-G3P (200 nM) binding to the various substrates is normalized to Ig-G3P binding to fAβ. (b) Ig-G3P (100 nM) binds specifically to fAS and shows no binding to AS monomers. (c) Fluorescence emission spectra of PTMI (0.5 μg/ml) co-incubated with either fAβ (2.5 μM) or Aβ monomers (2.5 μM). (d) Competition between PTMI and Ig-G3P for fAβ binding. The fluorescence intensity of PTMI decreases with increasing concentrations of Ig-G3P competitor. Different concentrations of Ig-G3P were added to 2.5 μM fAβ and incubated for 30 min. PTMI was then added to the complexes and fluorescence spectra recorded. The intensity at 590 nm was plotted against the concentration of Ig-G3P (black line). As a control, PTMI was first added to 2.5 μM fAβ and then incubated for 30 min with Ig-G3P. The fluorescent intensity of the fibers complexed with the dye was plotted (red line). Ig-G3P shows no significant binding to PTMI (data not shown). The binding stoichiometry calculated by SPR studies between G3P and fAβ monomers is approximately 1:50 (G3P:Aβ monomers) (Table 4). This calculation was made based on the amount of Aβ deposited on the biosensor chip and the amount of captured G3P (Table 4). This suggests that G3P binds a large epitope along the amyloid fiber that repeats Table 4. GAIM:Aβ binding stoichiometry calculated based on SPR binding studies Analyte Molecular mass (kDa) G3P G3P (IF1) Ig-G3P 25 25 110 RU mol/mm2a 400 1300 830 1.6 × 10−14 5.2 × 10−14 7.45 × 10−15 Aβ monomers: GAIM ratio 55 17 117 a Calculation is based on 1000 RU = 1 ng protein per area unit (mm2) and immobilized fAβ 8.88 × 10−13 mol/mm2. over several β strands (interstrand distance = 0.48 nm). This kind of binding is likely to be very d if f e r e nt f r o m s m a l l m ol e c ul e (m o le c u l ar mass b 500 Da) binding that is limited by interactions with one or two strands at a time [25,26]. To our knowledge, the only other example of a molecule that specifically binds amyloids over such large distances are the luminescent conjugated polythiophenes (LCPs), polymers of thiophenesline containing 25– 40 subunits [25,26]. We tested these molecules for fAβ binding competition with Ig-G3P in a solution-based assay. Figure 5c shows the fluorescence emission spectra of a specific LCP [polythiophene methyl imidazole (PTMI)] bound to either Aβ fibers or monomers. When Ig-G3P is pre-incubated with Aβ fibers, PTMI binding is blocked (Fig. 5d). Because the binding competition is highly concentration dependent, PTMI and G3P could share binding sites on fAβ. G3P remodels amyloid fibers G3P recognizes Aβ fibers specifically through hydrophobic core residues A binding stoichiometry of 1:30 (G3P:Aβ) was measured using 1H NMR (Fig. S6a), which compares with the G3P:Aβ stoichiometry estimated by SPR. As expected from the KD value, the complex is in the tight binding limit at 300 μM Aβ and dissociates in the slow exchange regime on the NMR timescale. The effects of G3P binding to Aβ fibers were examined by measuring its influence on the hydrogen/deuterium (H/D) exchange behavior of fAβ, using 15N transverse relaxation optimized spectroscopy (TROSY) NMR following dissolution of individually quenched samples in 95% DMSO/5% D2O. The exchange profile of fAβ in the absence of G3P (Fig. 6a and Fig. S6b) shows that two stretches of Aβ sequence, residues 17–25 and 31–40, are measurably protected from H/D exchange, and these peptide sequences span the β strands that form the core of the assembled fiber. Protection is also evident for residues 11–12 in the N-terminal region of fAβ. In total, 18 residues have measurable protection (Fig. S6b). The observed non-exponential (bimodal) time dependence of H/D exchange for protected residues is consistent with the previously proposed dynamic recycling model [27], where exchange occurs when Aβ monomers dissociate from the ends of the fibers. A non-exponential (bimodal) time dependence of H/D exchange is maintained on co-incubation of fAβ with G3P, but increased protection for residues 11–12, 17–25 and 31–40 is clearly visible at longer exchange times (Fig. 6b). On binding of G3P, no residues in fAβ become unprotected and no new residues become protected. This is consistent with H/D exchange still occurring through dynamic recycling of Aβ monomers but where their dissociation rate from the fiber ends is decreased. This further supports the data showing that monomeric G3P does not remodel amyloid fibers. The ability to interact with multiple misfolded protein assemblies, together with the binding regions defined by H/D exchange studies, suggest that G3P may recognize β edge strands that are capable of selftemplating [28,29]. Amyloid fibers and TolA-C bind to a common region of g3p In the closed state, the N1 and N2 domains of g3p are fully folded [24] (Fig. 7a). The interface between the two g3p outer domains is rich in β strands that self-anneal and prevent the g3pN1 domain from binding its cognate membrane-associated E.coli receptor, the C-terminal domain of TolA. Infection initiates when the g3pN2 domain binds to a predominantly α helical assembly, the F-pilus, with very high specificity (KD = 3 pM) [30]. This g3pN2-F-pilus binding triggers dissociation of N1, which contains the TolA-C binding site, from N2, and 2509 concomitantly proline 213 in the hinge between N2 and N3 isomerizes from the strained cis conformation to a less strained trans conformation. This prolyl isomerization acts as a switch to maintain g3pN1N2 in an open and infection-competent state (Fig. 7a) [31,32]. Incubating either phage or g3p at elevated temperatures to dissociate N1 and N2 shifts the equilibrium from the cis (closed) to the trans (open) proline conformation [22,24]. To determine if fAβ binding might involve either the TolA or F-pilus binding sites on g3p, we compared amyloid fiber binding of phage g3p mutants that affect g3p N1–N2 association. The IIHY mutant has a stabilized closed-state g3p conformation [Tm (IIHY g3p) = 61.8 °C, Tm (wild-type g3p) = 48.1 °C]. If TolA-C binding sites are involved in g3p:fAβ interactions, then the closed IIHY mutant should show relatively lower potency binding to fAβ. Alternatively, if the F-pilus binding region is involved in fAβ binding, then the binding should be unaffected by a constitutively closed N1N2 conformation. Binding competition studies show that the IIHY mutant poorly competes with wild-type phage for fAβ binding (Fig. 7b). Unlike wild-type phage, increasing the temperature during the binding reaction to 50 °C does not increase IIHY-fAβ binding (Fig. 7b). These data are consistent with direct g3pN1:fAβ interactions. We next tested a phage mutant that displays reduced binding to F-pili and carries a constitutively open conformation of g3pN1N2, both effects caused by three g3pN2 alanine substitutions [33,34]. Because the AAA mutations create a more exposed TolA-C binding region through g3p N1–N2 dissociation, this phage displays less temperature sensitivity for TolA binding. Consistent with its TolA-C binding behavior, the AAA mutant is not affected by temperature shifts for fAβ binding when compared to wild-type phage. Taken together, these mutant phage studies suggest that an open g3p N1– N2 conformation promotes amyloid binding and that the TolA-C binding region of g3pN1 is involved in amyloid fiber recognition. To confirm and extend the observation that g3pN1 is involved directly in amyloid fiber recognition, we tested the filamentous phage IF1, which displays a g3p with N1N2 in a constitutively open conformation [22], for fAβ binding. Studies with both the intact IF1 phage (Fig. S1d) and G3P derived from IF1 (Fig. 7c) confirm that the open g3p conformation promotes greater binding to fAβ. Since TolA-C is the natural substrate of open state g3p, we carried out a binding competition study to determine if soluble TolA-C can prevent Ig-G3P binding to fAβ. Figure 7d shows Ig-G3P binding to fAβ without and with TolA-C pre-incubation. Consistent with the g3p-soluble TolA-C binding potency (KD = 22 μM) [35], TolA interferes with G3P:fAβ binding with IC50 ~ 15 μM. 2510 G3P remodels amyloid fibers Fig. 6. Time course NMR H/D exchange of G3P:fAβ. (a) Profiles of residues R5, G9, F19, E22, M35 and V40 complexed with (Δ) or without (□) G3P. (b) Protection offered by G3P to the Aβ fiber core residues at the end of 744 h. The normalized intensities of fAβ (black lines) and fAβ:G3P (red lines) are shown at the end of 744 h of incubation. 2511 G3P remodels amyloid fibers open closed (a) F-pili binding F-pili binding Closed Open (c) 1500 Response (R.U.) Alexa488-M13 pelleted (%) (b) 75 50 25 1250 1000 750 500 250 0 0 WT IIHY AAA 0 200 400 600 800 1000 1200 Time (s) Absorbance at 450nm (d) 0.75 0.50 0.25 0.00 0 3.125 12.5 25 TolA [μM] Fig. 7. Effect of g3pN1N2 conformation for fAβ binding. (a) A schematic ribbon diagram showing the N1 (golden) and N2 (violet) domains of G3P in the closed and open conformations. The F-pili binding region on g3pN2 is marked and the area shaded. On the inner side of the two domains (“inverted horseshoe” conformation), smaller and larger shaded regions represent the TolA-C binding sites in the closed and the open states, respectively. (b) Binding competition study with the closed g3p carrying IIHY phage variant and open g3p carrying AAA phage variant, respectively. Fluorescently labeled phage was allowed to bind fAβ pre-incubated with wild type (wt), IIHY or AAA phage at 37 °C or 50 °C. (c) SPR studies measuring binding of wt-G3P (black) and IF1-G3P (red) to fAβ. (d) Binding competition study showing a concentration-dependent inhibition of Ig-G3P binding to fAβ in the presence of soluble TolA-C. Ig-G3P (3 nM) was pre-incubated with increasing concentrations of soluble TolA-C prior to incubation with fAβ (1.6 μM). This estimated TolA-C IC50 for blocking Ig-G3P binding to fAβ is ~ 15 μM. Discussion Our studies demonstrate amyloid interaction activities of the filamentous phage minor capsid protein g3p, which correlate with the structure function of g3p described for E. coli infection [34,36,37]. Filamentous phage infection of E. coli requires g3p, a protein displayed as 3–5 copies at the filament tip. g3p mediates the earliest infection steps, including recognition of both primary (F-pilus) and secondary (TolA) receptors required for infection. The two outer domains of g3p define two alternative functional states represented in Fig. 7a. In the unbound or “closed” state, the two N-terminal domains of g3p adopt an inverted horseshoe-like conformation with a hydrophobic groove running between the two lobes [23]. Binding 2512 of the g3pN2 domain to the host cell F-pilus triggers g3p N1–N2 interdomain dissociation that exposes the hydrophobic groove (“open” state). Dissociation of the N1 domain in turn triggers a cis–trans isomerization of proline 213 located in the N2–N3 interdomain hinge region and breaks or destabilizes a number of hydrogen bonds at the interface of the two N-terminal domains [32]. These two events allow N1 to bind to a co-receptor, the C-terminal domain of a periplasmic protein, TolA, to trigger infection. Crystal structure of the N1-TolA complex shows that the association between the g3pN1 and TolA-C domains involves extensive hydrogen bonding [23,37], similar to β strand–β strand interactions described for several naturally self-templating and de novo synthesized amyloids [38,39]. Based on binding and g3p mutant analyses, we hypothesize that g3p utilizes TolA-C binding sequences (including the g3p β4 strand) to recognize and bind the canonical amyloid fold represented in Aβ fibers. This hypothesis was initially based on the temperature dependency of M13:fAβ binding and dissociation, which is predicted by the temperature dependency of g3p–TolA interactions. The hypothesis is further supported by direct evidence that g3p is both necessary and sufficient for amyloid targeting activities, since destruction of phage-associated g3p drastically reduces M13 binding and remodeling of fAβ and that G3P alone can mediate amyloid recognition. Finally, TolA-C binding competition experiments show that TolA-C and fAβ may interact with shared binding sites on g3p. H/D exchange studies on G3P:fAβ complex clearly indicate binding to two separate stretches of hydrophobic residues, 17–25 (17-LVFFAEDVG-25) and 31– 40 (31-IIGLMVGGVV-40) are strongly affected by binding. These stretches form the core β strands in Aβ fibers [40]. Since most amyloids, including fAβ, assemble in a nucleation-dependent manner [1,2,41], only a relatively low concentration of G3P molecules (1–10%), compared to the concentration of nucleating monomers, would be required to cap fAβ seeds in a spontaneous assembly reaction. This is consistent with the relative concentrations of G3P (20–100 nM) and Aβ (2500 nM) necessary for effective assembly inhibition (Table 3). The biphasic H/D exchange protection of early (b 50 h) C-terminal region and late (N 200 h) Aβ middle residues by G3P together with evidence for binding stoichiometry of 1:30::G3P:Aβ suggest that G3P binds more than one region on the fiber: (a) β strands that form edges of the growing amyloid fiber and (b) exposed hydrophobic patches along the C-terminal regions of the fiber. Preliminary mutational analysis suggests that, in addition to the TolA binding g3p β4 strand, other hydrophobic or reactive edge β strands, such as the β10 strand on the N2 domain (that anneals and protects the TolA-C binding sequence), might also participate in amyloid recognition. Deletion of the N2 domain or altering interdomain G3P remodels amyloid fibers interactions leads to less efficient binding and/or remodeling (data not shown). Multiple binding events to different regions of the amyloid might explain both the unusually potent binding for fAβ and the conformational rearrangement we measure by ThT binding, CA filter retention and loss of fibrillar fAβ by X-ray diffraction methods (Fig. S2b and c). Whether or not the same types of binding interactions are found between g3p and tau and/or α-synuclein aggregates has not been determined, the common role for hydrophobic edge-to-edge association for amyloid assembly indicates an analogous mechanism [28,39]. Further NMR studies of the amyloid: G3P complex should help to identify precisely the g3p residues involved in binding to amyloids. The ability to bind and engage a larger conformational motif that spans ~ 30 monomers of Aβ peptide similar to LCPs is a unique feature of G3P. If the repeating twist length in an amyloid fiber is between 11.5 and 14.5 nm or about 25–30 Aβ monomers stacked along the fiber axis, then each G3P molecule would wrap around an entire helical twist [42]. The flexible interdomain linker that allows the two domains to move apart by almost 20 nm during E. coli infection [43] could permit engaging both the middle region and the C-terminal sequences along the fiber. Interestingly, LCP molecules, such as polythiophene acetic acid and PTMI, have been predicted to bind and alter the structures of amyloids in a similar fashion [44]. The Ig-G3P we created shares several features with previously described antibodies that have been reported to bind, remodel or block assembly of various misfolded proteins. For example, both polyclonal and monoclonal IgM antibodies that recognize conformational epitopes on amyloid fibers have been characterized [45,46], and other antibodies have been shown to selectively interact with fibrillar [47] and non-fibrillar oligomers [48–50]. However, to our knowledge, no amyloid targeting agent that is as potent and as broadly acting as Ig-G3P has been described. These activities include sequence-independent recognition with high specificity for misfolded amyloid aggregates over native or unfolded proteins, including polymers and hydrophobic proteins, high binding selectivity of aggregates over monomer subunits and the ability to inhibit fiber assembly at an G3P:Aβ ratio (1:30) that implicates oligomer interactions. In fact, we have performed binding assays comparing Aβ monomer, oligomer and fiber binding that show evidence for oligomer recognition (Fig. S7). Evidence for general amyloid recognition is based on M13 and Ig-G3P binding data for amyloid fibers of Aβ, α-synuclein and tau proteins. Ig-G3P also recognizes other misfolded protein assemblies, including fibers of aggregated transthyretin, superoxide dismutase 1, the NAC peptide and huntingtin polyQ, all implicated in neurodegenerative diseases (data not shown). The NMR H/D exchange studies suggest that binding between g3p and β strand 2513 G3P remodels amyloid fibers edges could explain both fiber and oligomer recognition activities. Finally, Ig-G3P binding causes structural alteration of the fibers similar to that described for N-terminal antibodies [51,52]. Interestingly, g3p appears to recognize C-terminal and middle regions of the Aβ sequence, similar to anti-Aβ middle region-specific or C-terminal region-specific MAbs, respectively. However, these MAbs, unlike the N-terminal anti-Aβ MAbs, cannot directly mediate fiber disaggregation [53]. Naturally occurring autoantibodies, which recognize C-terminal and mid-region Aβ sequences, have been described, but these antibodies interact only with oligomer assemblies and have poor recognition of amyloid fibers [54]. Taken together, g3p appears to represent a unique type of amyloid targeting motif, which we call GAIM, which has the potential to serve as the basis for the creation of novel therapeutics for neurodegenerative diseases. Materials and Methods Low-retention microcentrifuge tubes (Fisher Scientific 02-681-320) were used for all experiments to ensure that there are no artifacts related to protein binding to the tubes. All the phage, Ig fusion and recombinant protein preparations were tested to ensure that there was no protease contamination. Binding and remodeling reactions with or without protease inhibitor cocktails (complete mini protease inhibitor tablets 11 836 170 001; Roche Diagnostics, USA) showed identical binding and remodeling activities. All buffer samples contained 0.1% NaN3 as a preservative. Bacteriophage preparation Bacteriophage M13 wild type was obtained from Prof. J. Messing (Rutgers University, New Brunswick, NJ). The nucleotide sequence is deposited with GenBank (accession number JX412914.1). Phage fd and IF1 variants and a G3P expression plasmid were gifts from Prof. Franz X. Schmid (Universität Bayreuth, Germany) [22]. Small-scale (b 1 l) bacteriophage M13, fd mutant and IF1 stocks were produced as previously described [55] and modified [10] using polyethylene glycol to concentrate clarified infection supernatants and with additional purification using cesium chloride (CsCl) buoyant density equilibrium centrifugation after the polyethylene glycol step [56], followed by dialysis of the isolated M13 band against phosphate-buffered saline (PBS) to remove CsCl. M13 phage were grown by infecting either of two F-pilus expressing E. coli strains, TG-1 [supE Δ(lac-proAB) hsdR4/F′ traD36 proAB + lacIq (lacZ)ΔM15] [57,58] (Zymo Research, Irvine, CA) or JM109 [supE Δ(lac-proAB) hsdR17 recA1/F′ traD36 proA + prob + lacIq (lacZ)ΔM15] [59], obtained from the American Type Culture Collection (ATCC#53323) and from J. Messing, Rutgers University, New Brunswick, NJ. Bacteriophage preparations following equilibrium centrifugation and dialysis reproducibly produced ~ 1 × 10 14 phage particles per milliliter determined by optical density (OD269 nm/OD320 nm) using the following equation to calculate phage particle titer: [(OD269 nm − OD320 nm) × 6 × 10 16]/6.4 × 10 3 = phage/ml. † M13 stocks were measured for plaque-forming units (PFU) using the TG-1 host, as previously described [60]. The PFU titer assay typically produces a lower titer (0.1–0.3) compared to the phage particle titer determined by OD due to variability of the PFU method. Expression, purification and assembly of amyloid fibers Full-length tau-441 and tau-MTBR expression, purification and assembly Human full-length tau (2N/4R isoform, MAPT NM_005910 from Origene) and the microtubule binding region (residues 243–375) were cloned into the bacterial expression vector pET21a (+ ) with a C- terminal myc tag. Recombinant fulllength tau protein was expressed in Rosetta™ 2(DE3) pLysS competent cells (Novagen) and purified as described by Margittai and Langen [61]. Purified protein was assembled into fibers at 37 °C for 7 days by dissolving 10 μM tau monomers in 50 mM ammonium acetate buffer (pH 7.0) containing 5 mM DTT. Aggregation was initiated by adding 2.5 μM low molecular weight heparin. Recombinant tau-MTBR protein was expressed and purified similar to full-length tau with minor changes: bacterial pellets containing tau-MTBR were resuspended in lysis buffer [50 mM 4-morpholineethanesulfonic acid (pH 6.5), 250 U/ml benzonase nuclease, 1 mg/ml lysozyme and 1 complete protease inhibitor tablet] and incubated on ice for 2 h and homogenized by sonication. Following ammonium sulfate precipitation, pellets were resuspended in 50 mM 4-morpholineethanesulfonic acid (pH 6.5), 10 mM NaCl, 0.5 mM EDTA and 2 mM DTT, at 50 °C for 30–60 min and filtered through a 0.22-μm filter before being applied onto a HiTrap SP XL column (GE Healthcare Life Sciences, Piscataway, NJ). tau-MTBR was eluted with 0.1–0.5 M NaCl and dialyzed into 100 mM sodium acetate buffer (pH 7). tau-MTBR fibers were assembled by dissolving 40 μM tau-MTBR in 0.1 M sodium acetate buffer, 2 mM DTT and 40 μM low-molecular-weight heparin. α-Synuclein purification and assembly Full-length α-synuclein (NM_000345; Origene) was cloned into a pET21a (+) vector and expressed as described by Hoyer et al. [62]. The purified protein was aggregated using 2 mM SDS as described by Ahmad et al. [63]. The fibers were thoroughly dialyzed using a 30-kDa membrane (Slidelyzer, Pierce) to remove trace amounts of detergent and lower-molecular-weight α-synuclein species. Purification and assembly of yeast prion domain NM of Sup35 The first 253 residues of the yeast translation termination protein Sup35 also referred to as NM region of Sup35 was PCR amplified from a genomic DNA preparation and cloned into a pET21a (+) vector. Cysteine mutants were prepared as described in Tessier and Lindquist [64]. The wild type and the cysteine variant proteins were expressed using Rosetta™ 2(DE3) pLysS competent cells (Novagen) and purified as described in Tessier and Lindquist [64]. The urea-methanol pellet containing NM protein was resuspended and centrifuged at 13,000 rpm for 2514 30 min on a bench top centrifuge. The pellet containing NM monomers was dissolved in PBS (5 μM final concentration) and allowed to assemble at room temperature by placing it on a rotator. Aβ fiber assembly Aβ (Bachem H-1368) was dissolved in hexafluoroisopropanol (HFIP), sonicated and incubated at room temperature for 2–18 h until a clear solution developed. The peptide solution was dried under vacuum for 2–3 h. The peptide was aggregated as described by LeVine [20]and Stine et al. [65]. Aβ peptide aggregation using DMSO/HCl:Aβ monomer was dissolved (200 μg) in 37.5 μl DMSO, sonicated for 1 min and diluted with 520 μl of 10 mM HCl solution (final Aβ concentration, 80 μM). Samples were incubated for 3 days at 37 °C (without shaking). Aβ peptide aggregation DMSO/ PBS:Aβ monomer (100 μg) was dissolved in 18 μl DMSO, sonicated for 1 min and diluted with 382 μl of PBS containing 0.1% NaN3 (final Aβ concentration, 55 μM). Samples were incubated for 3 days at 37 °C (without shaking). Fiber quality for each of the abovementioned amyloid proteins was monitored by transmission electron microscopy and by performing dye binding studies with ThT and the luminescent conjugated probe polythiophene acetic acid [26]. SPR binding studies Aβ fiber binding studies SPR experiments were performed on a BIACORE2000 or T200 instruments using CM3 sensor chips. Reagents and sensor chips were procured from GE Health Sciences. CM3 sensor chips was activated by injecting a 1:1 ratio of 0.4 M EDC and 0.1 M NHS solutions for 7 min at a flow rate of 20 μl/min. Sonicated fAβ (1 min in water bath sonicator) diluted 1:1 (40–50 μM final concentration) in 10 mM sodium acetate buffer (pH 5.0) and immobilized at 20 μl/min for 7 min. The remaining activated surface groups were blocked with 1 M ethanolamine (pH 8.0) for 7 min; in these studies, the fibril immobilization densities varied from 3000 to 5000 RU. Aβ monomer solution (1 mg/ml in DMSO) was immobilized using a similar protocol. M13, G3P and Ig-G3P were used at 25 °C or pre-incubated at 37 °C for 3 h, followed by 1 h cool down to 25 °C before injection. Binding was performed at 25 °C at a flow rate of 3 or 10 μl/min for M13 and G3P, Ig-G3P, respectively. Binding parameters were calculated using BIAevaluation software on the BIACORE2000. Tau fiber binding studies ftau (10 μM) in HBS-EDTA (10 mM Hepes, 150 mM NaCl and 3 mM EDTA) at pH 8 was allowed to react with EDC/ NHS activated CM3 sensor chip surface for 7 min at 5 μl/min followed by ethanolamine blocking. Surface density was 1000 RU under these conditions. Ig-G3P (1 μM) in PBS was allowed to bind at a flow rate of 5 μl/min for 10 min at 25 °C. PBS was used as the running buffer. AS fiber binding studies α-Synuclein fibers were diluted in 10 mM acetate buffer to final concentration of 25 μM and the solution pH was G3P remodels amyloid fibers adjusted to pH 3.5. Immobilization was performed at flow rate of 10 μl/min on EDC/NHS activated CM3 sensor chip followed by ethanolamine blocking. Surface density was 3500 RU under these conditions. Ig-G3P (2 μM) in HBS was allowed to bind at a flow rate of 5 μl/min for 10 min at 25 °C. HBS was used as the running buffer. M13 Alexa488 labeling Alexa488 labeling of M13 was carried out by adding 1.6 μl of Alexa488-succinimidyl ester [10 mg/ml stock or 15.5 mM of Alexa488-NHS (A20100; Life Technologies)] to 1 ml of 1 × 10 14 phage. The samples were incubated in the dark for an hour with mild vortexing at regular intervals. The labeling reaction was stopped by adding 50 μl of 0.5 M glycine-NaOH buffer (pH 8.5) and the M13 solution was extensively dialyzed (10 kDa membrane; Slidelyzer, Pierce). The labeling efficiency was calculated by estimating the concentrations of bound dye and the g8p coat protein (by ELISA trap phage quantitation assay). Quantitative solution binding assay We pre-incubated 10 μM fAβ with or without Alexa488labeled phage (1 × 10 10 to 5 × 10 13/ml). The mixtures incubated at 37 °C for 3 h were centrifuged at 7500g for 10 min. The pelleted fluorescent phage was washed in PBS and the concentration was estimated by measuring the fluorescence at 520 nm (excitation, 485 nm). ThT fluorescence We pre-incubated 2.5 μM fAβ with different concentrations of M13 (1 × 10 10 to 1 × 10 12/ml) at 37 °C for 3 days. We added 5-fold excess of ThT to each sample at the end of the incubation and recorded the fluorescence spectra using a Sapphire II Tecan plate reader. All samples were excited at 450 nm. CA filter retardation assay This assay is used to monitor the destabilization or remodeling of amyloid fibers into non-amyloidogenic or soluble aggregates. The assays were primarily adapted from Wanker and Chang and Kuret [18,19]. We pre-incubated 2.5 μM preparations of amyloid fibers (fAβ, ftau, fAS and fNM) with different concentrations of M13 at 37 °C for 3 days. At the end of the incubation, fibers with and without phage were diluted and spotted on CA membranes on vacuum blots. The membranes were extensively washed with PBS and probed with specific primary antibodies for 1 h. HRP-conjugated secondary Ab was used to quantitate the fibrillar aggregates retained on the membrane. Preparation of ArgC phage The endoproteinase ArgC (Clostripain, Worthington, catalog #P09870) was used to selectively remove the g3p domains N1 and N2 from M13. ArgC specifically hydrolyzes the amide bonds at the carboxylic side of the residue arginine in a polypeptide. ArgC was activated by adding CaCl2 and DTT (final concentration CaCl2 = 7.5 mM, DTT = 5 mM). To 0.5 ml of M13 (1 × 10 14 phage/ml), we added 30 μl of sequencing grade ArgC (1 mg/ml) and incubated the 2515 G3P remodels amyloid fibers solution at 37 °C for 2 days. At the end of the incubation, the enzyme was inactivated by adding 50 mM EDTA. The phage was then purified over a Superdex 200 size-exclusion chromatography column and concentrated. The removal of the N1N2 domain was verified by SDS-PAGE, Western blotting and peptide mapping analysis before use in disaggregation assays. In addition, mass spectrometry was used to analyze the products of proteolysis, which confirmed specific cleavage of g3p (data not shown). Cloning, expression and purification of human Ig fusion proteins The DNA sequence corresponding to the N1N2 domain of g3p protein was cloned into the pFUSE-hIgG1-Fc2 (Invivogen) vector. FreeStyle MAX 293 Expression System (Invitrogen) was used for large-scale transient transfection and protein expression of the fusion constructs. Transfected FreeStyle 293F cells are grown for 7 days before the medium is harvested by centrifugation. The fusion constructs were purified on HiTrap Protein A Fast Flow Columns (GE Health Sciences) and eluted in 0.1 M glycine (pH 3.0) as per manufacturer's instructions. This protein was neutralized with 1 M Tris–HCl (pH 8.0) and dialyzed against PBS prior to use in binding and disaggregation assays. The Ig fusions were characterized for protein quality and activity (Fig. S8a–d). Assembly inhibition studies Aβ, tau and α-synuclein assembly inhibition studies were carried out to calculate the amyloid aggregation inhibitory concentrations (IC 50). Phage (1 × 10 13 to 1 × 10 9 particles/ml), G3P (2000–0.2 nM) or Ig-G3P (500–0.2 nM) were serially diluted (10-fold dilutions) into fiber assembly reactions of Aβ (2.5 μM), tau (2 μM) or α-synuclein (10 μM), respectively, and incubated at 37 °C for 7 days. Small aliquots of each reaction mix were removed at the end of 7 days to perform ThT binding assays. The inhibition concentrations were calculated based on reduced ThT fluorescence compared to control assembly reactions. Aggregation of tau and α-synuclein were initiated by adding 0.5 μM heparin (low molecular weight) or by agitation (850 rpm), respectively. Detergent solubility of amyloid–M13 complexes Sup35NM fibers and α-synuclein fibers were fluorescently labeled as per the manufacturer's protocols and a mixture of unlabeled monomer and 10% labeled monomer was assembled into fibers. The fibers were co-incubated with M13 at indicated concentrations for 1–3 days. Samples were transferred into varying concentrations of and incubated at 37 °C for 6 h (Sup35NM) or 1 h (α-synuclein) centrifugation at 30,000 rpm for 30 min. Released fluorescence in supernatants was measured with a Sapphire II Tecan plate reader. Ig-G3P binding ELISA for non-amyloid proteins Proteins were immobilized on Maxisorb plates (ThermoFisher) 225 ng/well in PBS containing 0.05% sodium azide at 37 °C for ~ 16 h before being blocked in Superblock (ThermoFisher) for 1 h at room temperature. Non-amyloid proteins were from Sigma (BSA, A3059; gelatin, G9391; casein, C3400) and prepared as previously described [45]; elastin (E1625), bovine collagen (C9879), human collagen (C5483) and Aβ fibers were tip-sonicated for 5 × 30 s before immobilization on plates. Ig-G3P was added in PBS-Tween (0.05%) at the indicated concentrations and incubated at 37 °C for 1 h. All washes were performed within PBS-Tween (0.05%). After incubation with human specific Fc-HRP antibody at 1:10,000 in PBS-Tween (0.05%) and 10% superblock for 1 h at room temperature, the signal was developed with TMB solution (Sigma). The reaction was stopped by the addition of 2 N HCl and absorbance at 450 nM was read in a Sapphire II Tecan plate reader. Ig-G3P binding ELISA for α-synuclein and Aβ We immobilized 1.6 μM fiber or monomer in PBS containing 0.05% sodium azide on Maxisorb plates (ThermoFisher) pre-coated with 1% BSA solution. Ig-G3P was added in PBS-Tween (0.05%) at the indicated concentrations and incubated at 37 °C for 1 h. All washes were performed within PBS-Tween (0.05%). After incubation with human specific Fc-HRP antibody at 1:2500 in PBS-Tween (0.05%), 1% milk for 1 h at room temperature, the signal was developed with TMB solution (Sigma). The reaction was stopped by the addition of 2 N HCl and absorbance at 450 nM was measured with a Sapphire II Tecan plate reader. CD analysis G3P, Ig-G3P and their complexes with Aβ were buffer exchanged into 20 mm potassium phosphate buffer (pH 7.4) using an Amicon ultrafiltration setup (3 kDa). The ellipticity at 205 nm was measured using a 1-mm path length cuvette. Temperature melting data were recorded by placing the cuvettes on a peltier heated sample chamber (Chirascan, Applied Photophysics). NMR studies Materials for NMR studies Unlabeled Aβ (cat. A-1002-2, lot. 5181142T) and HFIP-treated 15N Aβ (cat. A-1102-2, lot. 12091142N) were from r-peptide LLC, Georgia, USA. Peptides were treated with HFIP prior to assembly into fibers. HFIP treatment involved dissolution of Aβ in 100% HFIP at 1 mg/ml and overnight fume hood evaporation at 25 °C to 1/8th original volume, followed by lyophilization. Standard reagents were purchased from Fisher Scientific, Melford and Sigma. In solutions where the majority of water was deuterated, pD was determined using the formula pD = 0.929 × pH(app) + 0.42, where pH(app) is the apparent uncorrected pH meter reading [66]. G3P preparation for NMR studies G3P was buffer exchanged at 4 °C from purification buffer into d.H2O (0.01% NaN3) with 3 rounds of diafiltration, the protein concentration determined by UV spectroscopy (ε = 43,235) and then lyophilized. One-dimensional (1D) 2516 H NMR indicated complete refolding of G3P in PBS buffer (pH 7.4) (10 mM NaHPO4, 2.0 mM KH2PO4, 2.7 mM KCl, 137 mM NaCl and 1.0 mM EDTA) following lyophilization. Fiber preparation for NMR studies For each time point, two 1-mg HFIP-treated 15N Aβ vials were each dissolved in 1 ml of 10 mM HCl (pH 4.0), immediately sonicated for 10 min and incubated at 37 °C for 4 days. Upon completion, two 5-μl batch representative samples were removed for electron microscopy analysis that revealed uniform fiber morphology with periodic twists and little lateral association. The Aβ fiber preparations (1 mg/ml) were centrifuged individually at 179,000g (RCFMax) for 30 min at 4 °C to pellet mature fibers. The pellets were washed twice by addition of 1 ml of dH2O (0.01% NaN3) and centrifugation repeated. The supernatant was immediately removed and the pellet was dried with filter paper strips and lyophilized in the individual centrifuge tube. H/D exchange of Aβ fibers On ice, lyophilized G3P was resuspended to 25 μM in anhydrous D2O PBS buffer, pD 7.4, (10 mM NaHPO4, 2.0 mM KH2PO 4, 2.7 mM KCl, 137 mM NaCl and 1.0 mM EDTA) and 1.0 ml was added to each Aβ lyophilized fiber pellet in the two centrifuge tubes. The two pellets were resuspended, aliquoted into a single Eppendorf tube and incubated at 37 °C for the allotted H/D exchange time. fAβ fiber preparations in 500 μl of 95% DMSO/5% D 2 O were measured at a mean concentration of 300 μM by integration of NMR aliphatic peak intensity against TSP; thus, it was determined that the Aβ:G3P stoichiometric ratio was 30:1, which equates to 75:25 μM in the 2-ml H/D exchange solution. Following H/D exchange, we centrifuged the preparation at 13,000g (RCFMax), for 15 min at 4 °C. On ice, the supernatant was immediately removed, the pellet was washed by 2.0 ml addition of D2O at pD 4.0 (20 μM HCl) [pD 4.0 to minimize H/D exchange] and the centrifugation was repeated. The pellet was dried using filter paper strips and the exchange was quenched by freezing in liquid nitrogen, followed by lyophilization. The fiber preparation dead-time was on average 42 min from the start of centrifugation to flash freezing. The procedure was repeated identically for the Aβ fiber control with omission of G3P. NMR experiments were performed on a Bruker 600-MHz Avance cryoprobe equipped NMR spectrometer at 25 °C. For NMR analysis, the H/D exchanged fibers were dissolved in 500 μl of 95% DMSO-d6 and 5% D2O (adjusted to pD 4.0 with microliter additions of D2-dichloroacetic acid) vortexed for 1 min, briefly centrifuged and pipetted into an NMR tube that was quickly hand centrifuged. A series of 6-min TROSY experiments over 70 min were acquired immediately following sample equilibration, shimming, pulse optimization and performance of a 1D proton NMR spectrum. The dead-time between DMSO dissolution and commencement of the TROSY data collection had a typical value of 14 min. G3P-Aβ NMR titration 1D H NMR experiments were performed at 37 °C on a sample containing 300 μM Aβ in 500 μl D2O PBS buffer (10 mM Na2HPO4, 2.0 mM G3P remodels amyloid fibers KH2PO4, 2.7 mM KCl, 137 mM NaCl and 0.01% NaN3, pD 7.4). An initial Aβ fiber spectrum was obtained and the fiber solution was then mixed with lyophilized G3P (30 μM final concentration) and left to stand at 25 °C for 15 min to ensure resuspension before an NMR spectrum was obtained. A control 1D H NMR spectrum of lyophilized and re-dissolved 30 μM G3P sample from the same preparation used above was collected. X-ray fiber diffraction Sample preparation for diffraction studies The aggregated form of Aβ42 was prepared by incubating the lyophilized peptide (2.5 mg) in 10 mM HCl solution for 24 h, after which the pH was adjusted to 7.4. The aggregation of Aβ42 was monitored by ThT assay and transmission electron microscopy (see above). The aggregated Aβ42 was mixed with M13 in phosphate buffer solution (PBS) and incubated for 0 and 48 h at 37 °C. Samples were then dialyzed at 4 °C against water to remove any residual salt and DMSO and were then lyophilized. The lyophilized samples for X-ray diffraction included M13 alone (2.5 × 10 13 phage), aggregated Aβ42 alone (400 μg), aggregated Aβ42 (400 μg) + M13 (2.5 × 10 13) at incubation time t = 0 and aggregated Aβ42 (400 μg) + M13 (2.5 × 10 13) at t = 48 h. Each lyophilized sample was packed into a thin walled glass capillary of 0.7 mm in diameter. X-ray diffraction measurements Measurements of X-ray diffraction patterns from the peptides were conducted at room temperature using the Oxford Diffraction Xcalibur PX Ultra System (Oxford Diffraction Ltd., Concord, MA) located in the laboratory of Dr. Andrew Bohm (Department of Biochemistry, Tufts University, Boston, MA). The CuK α X-ray beam with wavelength of 1.542 Å was generated at 45 kV/40 mA using an Enhance Ultra, which is a sealed tube-based system incorporating confocal multilayer optics. The X-ray beam was monochromated and the K β component was removed by means of the double bounce within the confocal optic. The X-ray beam was focused to 0.3 mm × 0.3 mm (full width at half-maximum width at detector position). A two-dimensional Onyx CCD detector (Oxford Diffraction Inc., Concord, MA) with the pixel size of 121 μm was placed 85 mm from the sample position, covering the scattering range for Bragg spacings 1.8– 54 Å. The sample-to-detector distance was calibrated using a spherical ylid crystal (molecular formula, C10H10SO4) or a cubic alum crystal according to the information given by manufacturer. Exposure time was 150 s. The observed intensity of powder diffraction was measured using the angular averaging the chosen image area using FIT2D. Linear combination of Aβ and M13 intensities Series of diffraction patterns from the lyophilized samples containing Aβ42 and M13 were observed as a function of incubation time (t) for t = 0 and t = 48 h. The observed intensity as a function of the radial component of 2517 G3P remodels amyloid fibers spherical coordinates was decomposed by two basis sets that were the normalized intensities of lyophilized Aβ42 (IA) and lyophilized M13 (IM). The total intensity is given by I T ðR Þ ¼ k A I A ðR Þ þ k M I M ðR Þ; where k A þ k M ¼ 1: Note that the intensities contained both peaks and background and were normalized in the range from R = 0.0460 (1/Å) to 0.4484 (1/Å) so that the area under the curve was 1. The scale factors refer to the weight content. The best-fit curve was derived when the relative error between the observed and calculated curves (R-factor) was a minimum, where X X R¼ jI T −I obs j= jI obs j Acknowledgements We are grateful to Prof. J. Messing (Rutgers University, New Brunswick, USA) for providing wildtype M13 and to Prof. Franz X. Schmid (Universität Bayreuth, Germany) for providing fd variants and the G3P construct and for valuable discussions. We appreciate the editorial assistance of Ms. Roxanne Bales. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jmb.2014.04.015. Received 27 January 2014; Received in revised form 21 March 2014; Accepted 14 April 2014 Available online 22 April 2014 Keywords: amyloid; amyloid remodeling; gene 3 protein; Ig fusion R.K. and H.T. contributed equally to this work. † http://www.biosci.missouri.edu/smithgp/ phagedisplaywebsite/phagedisplaywebsiteindex.html. Abbreviations used: AD, Alzheimer's disease; BSA, bovine serum albumin; MAb, monoclonal antibody; g3p, gene 3 protein; GAIM, general amyloid interaction motif; LCP, luminescent conjugated polythiophene; PTMI, polythiophenes methyl imidazole; PBS, phosphate-buffered saline; ThT, thioflavin T; CA, cellulose acetate; TROSY, transverse relaxation optimized spectroscopy; H/D, hydrogen/deuterium; RIPA, radioimmunoprecipitation assay; PFU, plaque-forming units; SPR, surface plasmon resonance; SA, streptavidin. References [1] Chiti F, Dobson CM. Protein misfolding, functional amyloid, and human disease. Annu Rev Biochem 2006;75:333–66. [2] Eichner T, Kalverda AP, Thompson GS, Homans SW, Radford SE. Conformational conversion during amyloid formation at atomic resolution. Mol Cell 2011;41:161–72. [3] Weiner HL, Frenkel D. Immunology and immunotherapy of Alzheimer's disease. Nat Rev Immunol 2006;6:404–16. [4] Karran E, Mercken M, De Strooper B. The amyloid cascade hypothesis for Alzheimer's disease: an appraisal for the development of therapeutics. Nat Rev Drug Discov 2011;10:698–712. [5] Cheng B, Gong H, Xiao H, Petersen RB, Zheng L, Huang K. Inhibiting toxic aggregation of amyloidogenic proteins: a therapeutic strategy for protein misfolding diseases. Biochim Biophys Acta 2013;1830:4860–71. [6] Mangialasche F, Solomon A, Winblad B, Mecocci P, Kivipelto M. Alzheimer's disease: clinical trials and drug development. Lancet Neurol 2010;9:702–16. [7] Skovronsky DM, Lee VM, Trojanowski JQ. Neurodegenerative diseases: new concepts of pathogenesis and their therapeutic implications. Annu Rev Pathol 2006;1:151–70. [8] Clinton LK, Blurton-Jones M, Myczek K, Trojanowski JQ, LaFerla FM. Synergistic interactions between Abeta, tau, and alpha-synuclein: acceleration of neuropathology and cognitive decline. J Neurosci 2010;30:7281–9. [9] Gomperts SN, Locascio JJ, Marquie M, Santarlasci AL, Rentz DM, Maye J, et al. Brain amyloid and cognition in Lewy body diseases. Mov Disord 2012;27:965–73. [10] Dimant H, Sharon N, Solomon B. Modulation effect of filamentous phage on alpha-synuclein aggregation. Biochem Biophys Res Commun 2009;383:491–6. [11] Dimant H, Solomon B. Filamentous phages reduce alphasynuclein oligomerization in the membrane fraction of SHSY5Y cells. Neurodegener Dis 2010;7:203–5. [12] Colletier JP, Laganowsky A, Landau M, Zhao M, Soriaga AB, Goldschmidt L, et al. Molecular basis for amyloid-beta polymorphism. Proc Natl Acad Sci U S A 2011;108:16938–43. [13] Sawaya MR, Sambashivan S, Nelson R, Ivanova MI, Sievers SA, Apostol MI, et al. Atomic structures of amyloid cross-beta spines reveal varied steric zippers. Nature 2007;447:453–7. [14] Han H, Weinreb PH, Lansbury PT. The core Alzheimer's peptide NAC forms amyloid fibrils which seed and are seeded by beta-amyloid: is NAC a common trigger or target in neurodegenerative disease? Chem Biol 1995;2:163–9. [15] Ramakrishnan M, Kandimalla KK, Wengenack TM, Howell KG, Poduslo JF. Surface plasmon resonance binding kinetics of Alzheimer's disease amyloid beta peptidecapturing and plaque-binding monoclonal antibodies. Biochemistry 2009;48:10405–15. [16] Demattos RB, Lu J, Tang Y, Racke MM, Delong CA, Tzaferis JA, et al. A plaque-specific antibody clears existing beta-amyloid plaques in Alzheimer's disease mice. Neuron 2012;76:908–20. [17] LeVine H. Thioflavine T interaction with synthetic Alzheimer's disease beta-amyloid peptides: detection of amyloid aggregation in solution. Protein Sci 1993;2:404–10. [18] Chang E, Kuret J. Detection and quantification of tau aggregation using a membrane filter assay. Anal Biochem 2008;373:330–6. [19] Wanker EE, Scherzinger E, Heiser V, Sittler A, Eickhoff H, Lehrach H. Membrane filter assay for detection of amyloidlike polyglutamine-containing protein aggregates. Methods Enzymol 1999;309:375–86. 2518 [20] LeVine H. Quantification of beta-sheet amyloid fibril structures with thioflavin T. Methods Enzymol 1999;309:274–84. [21] LeVine H, Scholten JD. Screening for pharmacologic inhibitors of amyloid fibril formation. Methods Enzymol 1999;309:467–76. [22] Martin A, Schmid FX. Evolutionary stabilization of the gene3-protein of phage fd reveals the principles that govern the thermodynamic stability of two-domain proteins. J Mol Biol 2003;328:863–75. [23] Lubkowski J, Hennecke F, Pluckthun A, Wlodawer A. The structural basis of phage display elucidated by the crystal structure of the N-terminal domains of g3p. Nat Struct Biol 1998;5:140–7. [24] Martin A, Schmid FX. A proline switch controls folding and domain interactions in the gene-3-protein of the filamentous phage fd. J Mol Biol 2003;331:1131–40. [25] Nilsson KP, Herland A, Hammarstrom P, Inganas O. Conjugated polyelectrolytes: conformation-sensitive optical probes for detection of amyloid fibril formation. Biochemistry 2005;44:3718–24. [26] Nilsson KP, Aslund A, Berg I, Nystrom S, Konradsson P, Herland A, et al. Imaging distinct conformational states of amyloid-beta fibrils in Alzheimer's disease using novel luminescent probes. ACS Chem Biol 2007;2:553–60. [27] Sanchez L, Madurga S, Pukala T, Vilaseca M, Lopez-Iglesias C, Robinson CV, et al. Abeta40 and Abeta42 amyloid fibrils exhibit distinct molecular recycling properties. J Am Chem Soc 2011;133:6505–8. [28] Richardson JS, Richardson DC. Natural beta-sheet proteins use negative design to avoid edge-to-edge aggregation. Proc Natl Acad Sci U S A 2002;99:2754–9. [29] Laganowsky A, Liu C, Sawaya MR, Whitelegge JP, Park J, Zhao M, et al. Atomic view of a toxic amyloid small oligomer. Science 2012;335:1228–31. [30] Lin A, Jimenez J, Derr J, Vera P, Manapat ML, Esvelt KM, et al. Inhibition of bacterial conjugation by phage M13 and its protein g3p: quantitative analysis and model. PLoS One 2011;6: e19991. [31] Jakob RP, Schmid FX. Energetic coupling between nativestate prolyl isomerization and conformational protein folding. J Mol Biol 2008;377:1560–75. [32] Weininger U, Jakob RP, Eckert B, Schweimer K, Schmid FX, Balbach J. A remote prolyl isomerization controls domain assembly via a hydrogen bonding network. Proc Natl Acad Sci U S A 2009;106:12335–40. [33] Deng LW, Perham RN. Delineating the site of interaction on the pIII protein of filamentous bacteriophage fd with the F-pilus of Escherichia coli. J Mol Biol 2002;319:603–14. [34] Eckert B, Schmid FX. A conformational unfolding reaction activates phage fd for the infection of Escherichia coli. J Mol Biol 2007;373:452–61. [35] Hoffmann-Thoms S, Weininger U, Eckert B, Jakob RP, Koch JR, Balbach J, et al. Initiation of phage infection by partial unfolding and prolyl isomerization. J Biol Chem 2013;288:12979–91. [36] Marvin DA. Filamentous phage structure, infection and assembly. Curr Opin Struct Biol 1998;8:150–8. [37] Riechmann L, Holliger P. The C-terminal domain of TolA is the coreceptor for filamentous phage infection of E. coli. Cell 1997;90:351–60. [38] Eichner T, Radford SE. A diversity of assembly mechanisms of a generic amyloid fold. Mol Cell 2011;43:8–18. [39] West MW, Wang W, Patterson J, Mancias JD, Beasley JR, Hecht MH. De novo amyloid proteins from designed combinatorial libraries. Proc Natl Acad Sci U S A 1999;96:11211–6. G3P remodels amyloid fibers [40] Luhrs T, Ritter C, Adrian M, Riek-Loher D, Bohrmann B, Dobeli H, et al. 3D structure of Alzheimer's amyloid-beta(142) fibrils. Proc Natl Acad Sci U S A 2005;102:17342–7. [41] Serio TR, Cashikar AG, Kowal AS, Sawicki GJ, Moslehi JJ, Serpell L, et al. Nucleated conformational conversion and the replication of conformational information by a prion determinant. Science 2000;289:1317–21. [42] Sunde M, Serpell LC, Bartlam M, Fraser PE, Pepys MB, Blake CC. Common core structure of amyloid fibrils by synchrotron X-ray diffraction. J Mol Biol 1997;273:729–39. [43] Holliger P, Riechmann L, Williams RL. Crystal structure of the two N-terminal domains of g3p from filamentous phage fd at 1.9 Å: evidence for conformational lability. J Mol Biol 1999;288:649–57. [44] Margalith I, Suter C, Ballmer B, Schwarz P, Tiberi C, Sonati T, et al. Polythiophenes inhibit prion propagation by stabilizing prion protein (PrP) aggregates. J Biol Chem 2012;287:18872–87. [45] O'Nuallain B, Wetzel R. Conformational Abs recognizing a generic amyloid fibril epitope. Proc Natl Acad Sci U S A 2002;99:1485–90. [46] Habicht G, Haupt C, Friedrich RP, Hortschansky P, Sachse C, Meinhardt J, et al. Directed selection of a conformational antibody domain that prevents mature amyloid fibril formation by stabilizing Abeta protofibrils. Proc Natl Acad Sci U S A 2007;104:19232–7. [47] Kayed R, Head E, Sarsoza F, Saing T, Cotman CW, Necula M, et al. Fibril specific, conformation dependent antibodies recognize a generic epitope common to amyloid fibrils and fibrillar oligomers that is absent in prefibrillar oligomers. Mol Neurodegener 2007;2:18. [48] Kayed R, Canto I, Breydo L, Rasool S, Lukacsovich T, Wu J, et al. Conformation dependent monoclonal antibodies distinguish different replicating strains or conformers of prefibrillar Abeta oligomers. Mol Neurodegener 2010;5:57. [49] Glabe CG. Conformation-dependent antibodies target diseases of protein misfolding. Trends Biochem Sci 2004;29:542–7. [50] Glabe CG. Structural classification of toxic amyloid oligomers. J Biol Chem 2008;283:29639–43. [51] Mamikonyan G, Necula M, Mkrtichyan M, Ghochikyan A, Petrushina I, Movsesyan N, et al. Anti-A beta 1-11 antibody binds to different beta-amyloid species, inhibits fibril formation, and disaggregates preformed fibrils but not the most toxic oligomers. J Biol Chem 2007;282:22376–86. [52] Solomon B, Koppel R, Frankel D, Hanan-Aharon E. Disaggregation of Alzheimer beta-amyloid by site-directed mAb. Proc Natl Acad Sci U S A 1997;94:4109–12. [53] Dodart JC, Bales KR, Gannon KS, Greene SJ, DeMattos RB, Mathis C, et al. Immunization reverses memory deficits without reducing brain Abeta burden in Alzheimer's disease model. Nat Neurosci 2002;5:452–7. [54] Dodel R, Balakrishnan K, Keyvani K, Deuster O, Neff F, Andrei-Selmer LC, et al. Naturally occurring autoantibodies against beta-amyloid: investigating their role in transgenic animal and in vitro models of Alzheimer's disease. J Neurosci 2011;31:5847–54. [55] Wickner W. Asymmetric orientation of a phage coat protein in cytoplasmic membrane of Escherichia coli. Proc Natl Acad Sci U S A 1975;72:4749–53. [56] Sambrook J, Russell DW. Screening bacterial colonies by hybridization: large numbers. CSH Protoc 2006;2006. http:// dx.doi.org/10.1101/pdb.prot3937. [57] Gibson JF, Poole RK, Hughes MN, Rees JF. Filamentous growth of Escherichia coli K12 elicited by dimeric, mixed-valence complexes of ruthenium. Arch Microbiol 1984;139:265–71. G3P remodels amyloid fibers [58] Barthelmebs L, Divies C, Cavin JF. Knockout of the pcoumarate decarboxylase gene from Lactobacillus plantarum reveals the existence of two other inducible enzymatic activities involved in phenolic acid metabolism. Appl Environ Microbiol 2000;66:3368–75. [59] Yanisch-Perron C, Vieira J, Messing J. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 1985;33:103–19. [60] Sambrook J, Russell DW. Growing bacteriophage m13 in liquid culture. CSH Protoc 2006;2006. http://dx.doi.org/10.1101/pdb. prot3994. [61] Margittai M, Langen R. Spin labeling analysis of amyloids and other protein aggregates. Methods Enzymol 2006;413:122–39. 2519 [62] Hoyer W, Antony T, Cherny D, Heim G, Jovin TM, Subramaniam V. Dependence of alpha-synuclein aggregate morphology on solution conditions. J Mol Biol 2002;322:383–93. [63] Ahmad MF, Ramakrishna T, Raman B, Rao Ch M. Fibrillogenic and non-fibrillogenic ensembles of SDS-bound human alpha-synuclein. J Mol Biol 2006;364:1061–72. [64] Tessier PM, Lindquist S. Prion recognition elements govern nucleation, strain specificity and species barriers. Nature 2007;447:556–61. [65] Stine WB, Dahlgren KN, Krafft GA, LaDu MJ. In vitro characterization of conditions for amyloid-beta peptide oligomerization and fibrillogenesis. J Biol Chem 2003;278:11612–22. [66] Krezel A, Bal W. A formula for correlating pKa values determined in D2O and H2O. J Inorg Biochem 2004;98:161–6.
© Copyright 2025 Paperzz