The Caenorhabditis elegans Aβ1–42 Model of Alzheimer Disease Predominantly Expresses Aβ3–42
2009; Elsevier BV; Volume: 284; Issue: 34 Linguagem: Inglês
10.1074/jbc.c109.028514
ISSN1083-351X
AutoresGawain McColl, Blaine R. Roberts, Adam P. Gunn, Keyla Perez, Deborah J. Tew, Colin L. Masters, Kevin J. Barnham, Robert A. Cherny, Ashley I. Bush,
Tópico(s)Cholinesterase and Neurodegenerative Diseases
ResumoTransgenic expression of human amyloid β (Aβ) peptide in body wall muscle cells of Caenorhabditis elegans has been used to better understand aspects of Alzheimer disease (AD). In human aging and AD, Aβ undergoes post-translational changes including covalent modifications, truncations, and oligomerization. Amino truncated Aβ is increasingly recognized as potentially contributing to AD pathogenesis. Here we describe surface-enhanced laser desorption ionization-time of flight mass spectrometry mass spectrometry of Aβ peptide in established transgenic C. elegans lines. Surprisingly, the Aβ being expressed is not full-length 1–42 (amino acids) as expected but rather a 3–42 truncation product. In vitro analysis demonstrates that Aβ3–42 self-aggregates like Aβ1–42, but more rapidly, and forms fibrillar structures. Similarly, Aβ3–42 is also the more potent initiator of Aβ1–40 aggregation. Seeded aggregation via Aβ3–42 is further enhanced via co-incubation with the transition metal Cu(II). Although unexpected, the C. elegans model of Aβ expression can now be co-opted to study the proteotoxic effects and processing of Aβ3–42. Transgenic expression of human amyloid β (Aβ) peptide in body wall muscle cells of Caenorhabditis elegans has been used to better understand aspects of Alzheimer disease (AD). In human aging and AD, Aβ undergoes post-translational changes including covalent modifications, truncations, and oligomerization. Amino truncated Aβ is increasingly recognized as potentially contributing to AD pathogenesis. Here we describe surface-enhanced laser desorption ionization-time of flight mass spectrometry mass spectrometry of Aβ peptide in established transgenic C. elegans lines. Surprisingly, the Aβ being expressed is not full-length 1–42 (amino acids) as expected but rather a 3–42 truncation product. In vitro analysis demonstrates that Aβ3–42 self-aggregates like Aβ1–42, but more rapidly, and forms fibrillar structures. Similarly, Aβ3–42 is also the more potent initiator of Aβ1–40 aggregation. Seeded aggregation via Aβ3–42 is further enhanced via co-incubation with the transition metal Cu(II). Although unexpected, the C. elegans model of Aβ expression can now be co-opted to study the proteotoxic effects and processing of Aβ3–42. Numerous studies support a role for aggregating Aβ 3The abbreviations used are: Aβamyloid βADAlzheimer diseaseSELDI-TOF MSsurface-enhanced laser desorption ionization-time of flight mass spectrometryQCglutaminyl cyclaseMES2-(N-morpholino) ethanesulfonic acidBicineN,N-bis(2-hydroxyethyl)glycineThTthioflavin-TpEpyroglutamateEMelectron micrograph. 3The abbreviations used are: Aβamyloid βADAlzheimer diseaseSELDI-TOF MSsurface-enhanced laser desorption ionization-time of flight mass spectrometryQCglutaminyl cyclaseMES2-(N-morpholino) ethanesulfonic acidBicineN,N-bis(2-hydroxyethyl)glycineThTthioflavin-TpEpyroglutamateEMelectron micrograph. in mediating the toxicity that underlies AD (1Hardy J. Selkoe D.J. Science. 2002; 297: 353-356Crossref PubMed Scopus (10863) Google Scholar, 2Tanzi R.E. Bertram L. Cell. 2005; 120: 545-555Abstract Full Text Full Text PDF PubMed Scopus (1481) Google Scholar). However, several key questions remain central to understanding how AD and Aβ pathology are related. What is the connection between Aβ aggregation and toxicity? Is there a specific toxic Aβ conformation or species? How and why does aging impact on Aβ precipitation? Significant effort to address these questions has been invested in the use of vertebrate and simple invertebrate model organisms to simulate neurodegenerative diseases through transgenic expression of human Aβ (3Link C.D. Mech. Ageing Dev. 2001; 122: 1639-1649Crossref PubMed Scopus (83) Google Scholar). From these models, several novel insights into the proteotoxicity of Aβ have been gained (4Cohen E. Dillin A. Nat. Rev. Neurosci. 2008; 9: 759-767Crossref PubMed Scopus (219) Google Scholar, 5Kraemer B. Schellenberg G.D. Int. Rev. Neurobiol. 2007; 77: 219-246Crossref PubMed Scopus (8) Google Scholar, 6Link C.D. Genes Brain Behav. 2005; 4: 147-156Crossref PubMed Scopus (79) Google Scholar, 7Boyd-Kimball D. Poon H.F. Lynn B.C. Cai J. Pierce Jr., W.M. Klein J.B. Ferguson J. Link C.D. Butterfield D.A. Neurobiol. Aging. 2006; 27: 1239-1249Crossref PubMed Scopus (76) Google Scholar). amyloid β Alzheimer disease surface-enhanced laser desorption ionization-time of flight mass spectrometry glutaminyl cyclase 2-(N-morpholino) ethanesulfonic acid N,N-bis(2-hydroxyethyl)glycine thioflavin-T pyroglutamate electron micrograph. amyloid β Alzheimer disease surface-enhanced laser desorption ionization-time of flight mass spectrometry glutaminyl cyclase 2-(N-morpholino) ethanesulfonic acid N,N-bis(2-hydroxyethyl)glycine thioflavin-T pyroglutamate electron micrograph. Human Aβ (e.g. in brain, cerebrospinal fluid, or plasma) is not found as a single species but rather as diverse mixtures of various modified, truncated, and cross-linked forms (8Sergeant N. Bombois S. Ghestem A. Drobecq H. Kostanjevecki V. Missiaen C. Wattez A. David J.P. Vanmechelen E. Sergheraert C. Delacourte A. J. Neurochem. 2003; 85: 1581-1591Crossref PubMed Scopus (177) Google Scholar, 9Saido T.C. Yamao-Harigaya W. Iwatsubo T. Kawashima S. Neurosci. Lett. 1996; 215: 173-176Crossref PubMed Scopus (235) Google Scholar, 10Maler J.M. Klafki H.W. Paul S. Spitzer P. Groemer T.W. Henkel A.W. Esselmann H. Lewczuk P. Kornhuber J. Wiltfang J. Proteomics. 2007; 7: 3815-3820Crossref PubMed Scopus (47) Google Scholar). Specific truncations, covalent modifications, and cross-linked oligomers of Aβ have potentially important roles in determining Aβ-associated neurotoxicity. For example, N-terminal truncations of Aβ have increased abundance in AD, rapidly aggregate, and are neurotoxic (9Saido T.C. Yamao-Harigaya W. Iwatsubo T. Kawashima S. Neurosci. Lett. 1996; 215: 173-176Crossref PubMed Scopus (235) Google Scholar, 11Pike C.J. Overman M.J. Cotman C.W. J. Biol. Chem. 1995; 270: 23895-23898Abstract Full Text Full Text PDF PubMed Scopus (283) Google Scholar). Furthermore, the N-terminal glutamic acid residue of Aβ3–42 can be cyclized to pyroglutamate (Aβ3(pE)-42) (12Saido T.C. Iwatsubo T. Mann D.M. Shimada H. Ihara Y. Kawashima S. Neuron. 1995; 14: 457-466Abstract Full Text PDF PubMed Scopus (504) Google Scholar), which may be particularly important in AD pathogenesis (13Russo C. Violani E. Salis S. Venezia V. Dolcini V. Damonte G. Benatti U. D'Arrigo C. Patrone E. Carlo P. Schettini G. J. Neurochem. 2002; 82: 1480-1489Crossref PubMed Scopus (175) Google Scholar, 14Schilling S. Zeitschel U. Hoffmann T. Heiser U. Francke M. Kehlen A. Holzer M. Hutter-Paier B. Prokesch M. Windisch M. Jagla W. Schlenzig D. Lindner C. Rudolph T. Reuter G. Cynis H. Montag D. Demuth H.U. Rossner S. Nat. Med. 2008; 14: 1106-1111Crossref PubMed Scopus (289) Google Scholar). Aβ3(pE)-42 is a significant fraction of total Aβ in AD brain (15Piccini A. Russo C. Gliozzi A. Relini A. Vitali A. Borghi R. Giliberto L. Armirotti A. D'Arrigo C. Bachi A. Cattaneo A. Canale C. Torrassa S. Saido T.C. Markesbery W. Gambetti P. Tabaton M. J. Biol. Chem. 2005; 280: 34186-34192Abstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar), accounting for more than 50% of Aβ accumulated in plaques (16Harigaya Y. Saido T.C. Eckman C.B. Prada C.M. Shoji M. Younkin S.G. Biochem. Biophys. Res. Commun. 2000; 276: 422-427Crossref PubMed Scopus (165) Google Scholar). Aβ3(pE)-42 seeds Aβ aggregation (17Schilling S. Lauber T. Schaupp M. Manhart S. Scheel E. Böhm G. Demuth H.U. Biochemistry. 2006; 45: 12393-12399Crossref PubMed Scopus (215) Google Scholar), confers proteolytic resistance, and is neurotoxic (13Russo C. Violani E. Salis S. Venezia V. Dolcini V. Damonte G. Benatti U. D'Arrigo C. Patrone E. Carlo P. Schettini G. J. Neurochem. 2002; 82: 1480-1489Crossref PubMed Scopus (175) Google Scholar). Recently, glutaminyl cyclase (QC) has been proposed to catalyze, in vivo, pyroglutamate formation of Aβ3(pE)-40/42 (14Schilling S. Zeitschel U. Hoffmann T. Heiser U. Francke M. Kehlen A. Holzer M. Hutter-Paier B. Prokesch M. Windisch M. Jagla W. Schlenzig D. Lindner C. Rudolph T. Reuter G. Cynis H. Montag D. Demuth H.U. Rossner S. Nat. Med. 2008; 14: 1106-1111Crossref PubMed Scopus (289) Google Scholar, 18Schilling S. Appl T. Hoffmann T. Cynis H. Schulz K. Jagla W. Friedrich D. Wermann M. Buchholz M. Heiser U. von Hörsten S. Demuth H.U. J. Neurochem. 2008; 106: 1225-1236Crossref PubMed Scopus (63) Google Scholar). Aβ1–42 itself cannot be cyclized by QC to Aβ3(pE)-42 (19Shirotani K. Tsubuki S. Lee H.J. Maruyama K. Saido T.C. Neurosci. Lett. 2002; 327: 25-28Crossref PubMed Scopus (21) Google Scholar), unlike Aβ that commences with an N-terminal glutamic acid-residue (e.g. Aβ3–42 and Aβ11–42) (20Schilling S. Hoffmann T. Manhart S. Hoffmann M. Demuth H.U. FEBS Lett. 2004; 563: 191-196Crossref PubMed Scopus (139) Google Scholar). QC has broad expression in mammalian brain (21Pohl T. Zimmer M. Mugele K. Spiess J. Proc. Natl. Acad. Sci. U.S.A. 1991; 88: 10059-10063Crossref PubMed Scopus (91) Google Scholar, 22Sykes P.A. Watson S.J. Temple J.S. Bateman Jr., R.C. FEBS Lett. 1999; 455: 159-161Crossref PubMed Scopus (49) Google Scholar), and its inhibition attenuates accumulation of Aβ3(pE)-42 into plaques and improves cognition in a transgenic mouse model of AD that overexpresses human amyloid precursor protein (14Schilling S. Zeitschel U. Hoffmann T. Heiser U. Francke M. Kehlen A. Holzer M. Hutter-Paier B. Prokesch M. Windisch M. Jagla W. Schlenzig D. Lindner C. Rudolph T. Reuter G. Cynis H. Montag D. Demuth H.U. Rossner S. Nat. Med. 2008; 14: 1106-1111Crossref PubMed Scopus (289) Google Scholar). N-terminal truncations at position 3 have been reported in senile plaques (23Miller D.L. Papayannopoulos I.A. Styles J. Bobin S.A. Lin Y.Y. Biemann K. Iqbal K. Arch. Biochem. Biophys. 1993; 301: 41-52Crossref PubMed Scopus (419) Google Scholar, 24Tamaoka A. Odaka A. Ishibashi Y. Usami M. Sahara N. Suzuki N. Nukina N. Mizusawa H. Shoji S. Kanazawa I. J. Biol. Chem. 1994; 269: 32721-32724Abstract Full Text PDF PubMed Google Scholar); however, the process that generates Aβ3–42 is unknown. Currently there are no reported animal models of Aβ3–42 expression. Advances in surface-enhanced laser desorption ionization-time of flight mass spectrometry (SELDI-TOF MS) analysis now facilitate accurate identification of particular Aβ species. Using this technology, we examined well characterized C. elegans transgenic models of AD that develop amyloid aggregates (25Link C.D. Proc. Natl. Acad. Sci. U.S.A. 1995; 92: 9368-9372Crossref PubMed Scopus (508) Google Scholar, 26Fay D.S. Fluet A. Johnson C.J. Link C.D. J. Neurochem. 1998; 71: 1616-1625Crossref PubMed Scopus (143) Google Scholar) to see whether the human Aβ they express is post-translationally modified. The strains N2, wild type; CL2006, dvIs2(pCL12(unc-54:hu-Aβ 1–42) + pRF4) (25Link C.D. Proc. Natl. Acad. Sci. U.S.A. 1995; 92: 9368-9372Crossref PubMed Scopus (508) Google Scholar); CL2120, dvIs14(pCL12(unc-54:hu-Aβ 1–42) + mtl-2:gfp); and CL2122, dvIs15(mtl-2:gfp) (26Fay D.S. Fluet A. Johnson C.J. Link C.D. J. Neurochem. 1998; 71: 1616-1625Crossref PubMed Scopus (143) Google Scholar) were provided by the Caenorhabditis Genetics Center. All strains were cultured at 20 °C on 8P/22Na medium (27Bianchi L. Driscoll M. Wormbook. 2006; (The C. elegans Research Community, ed) doi/10.1895/wormbook.1.122.1Google Scholar), and then at the first day of adulthood (4 days old), were aged at 25 °C as indicated. Human Aβ1–40, Aβ1–42, and Aβ3–42 peptides were synthesized by the W. M. Keck Laboratory (Yale University, New Haven, CT). Peptides were dissolved in 60 mm NaOH at room temperature and then diluted to 1 mg/ml in distilled H2O and 10× PBS (PBS is defined as 50 mm sodium phosphate, 2.7 mm KCl, 137 mm NaCl, pH 7.4) at a volume ratio of 2:7:1. Preparations were sonicated for 10 min in a water bath and then centrifuged at 13,500 × g for 10 min at 4 °C; the supernatant was then filtered through 0.2-μm filters (Supor, PALL) and kept on ice for immediate use. Peptide concentration was determined by measuring the absorbance value at 214 nm and applying the molar extinction coefficient values of 91,460 m−1 cm−1 for Aβ1–40 and Aβ3–42 and 94,530 m−1 cm−1 for Aβ1–42. Molar extinction coefficients were determined using amino acid analysis (Australian Proteome Analysis Facility Ltd.) and UV spectrometry (Lambda 25 UV-visible, PerkinElmer Life Sciences). Synchronized 4-day-old C. elegans adults were filtered (to remove eggs and larvae) through 40-μm nylon mesh (BD Biosciences) prior to sonication in chilled TBS (100 mm Tris-Cl and 150 mm NaCl, pH 8.0). After homogenization with a probe sonicator, lysates were clarified by centrifugation (13,500 × g for 5 min), and then the supernatant was removed and then kept on ice for immediate use or stored at −20 °C for subsequent analysis. For synthetic Aβ1–42 and Aβ3–42 standards, 40 pmol was analyzed. SELDI-TOF MS was performed using our established protocol (28Adlard P.A. Cherny R.A. Finkelstein D.I. Gautier E. Robb E. Cortes M. Volitakis I. Liu X. Smith J.P. Perez K. Laughton K. Li Q.X. Charman S.A. Nicolazzo J.A. Wilkins S. Deleva K. Lynch T. Kok G. Ritchie C.W. Tanzi R.E. Cappai R. Masters C.L. Barnham K.J. Bush A.I. Neuron. 2008; 59: 43-55Abstract Full Text Full Text PDF PubMed Scopus (571) Google Scholar). Immunocapture used affinity-purified W0-2 (epitope: Aβ 5–8) (29Ida N. Hartmann T. Pantel J. Schröder J. Zerfass R. Förstl H. Sandbrink R. Masters C.L. Beyreuther K. J. Biol. Chem. 1996; 271: 22908-22914Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar) antibody coupled to ProteinChip PS10 arrays (Bio-Rad), as described (28Adlard P.A. Cherny R.A. Finkelstein D.I. Gautier E. Robb E. Cortes M. Volitakis I. Liu X. Smith J.P. Perez K. Laughton K. Li Q.X. Charman S.A. Nicolazzo J.A. Wilkins S. Deleva K. Lynch T. Kok G. Ritchie C.W. Tanzi R.E. Cappai R. Masters C.L. Barnham K.J. Bush A.I. Neuron. 2008; 59: 43-55Abstract Full Text Full Text PDF PubMed Scopus (571) Google Scholar). Insoluble Aβ species from C. elegans lysates (described above) were collected by centrifugation (100,000 × g for 20 min). Pellets were then incubated in 70% (v/v) formic acid overnight at room temperature. Aβ species were resolved based on hydrophobicity using bis/Bicine urea PAGE as reported previously (30Klafki H.W. Wiltfang J. Staufenbiel M. Anal. Biochem. 1996; 237: 24-29Crossref PubMed Scopus (98) Google Scholar). Synthetic Aβ peptide standards were prepared as described above, with 75 ng of each loaded per lane. Following electrophoresis, gels were equilibrated (three washes for 5 min at room temperature) in MES SDS buffer (50 mm Tris base, 50 mm MES, 1 mm EDTA, 0.01% SDS at pH 7.3) prior to semidry transfer (iBlot, Invitrogen) to 0.2-μm nitrocellulose membrane. Membranes were then boiled in PBS for 3 min by microwave. Affinity-purified primary antibodies 4G8 (epitope: Aβ 18–22, Signet) and 6E10 (epitope: Aβ 3–8, Signet) were used at 1 μg/ml dilution in plus 0.05% Tween 20 (Sigma-Aldrich). Chemiluminescence (Pierce) was quantitated using a LAS-3000 Image Analyzer (Fuji). SignalP 3.0 (31Bendtsen J.D. Nielsen H. von Heijne G. Brunak S. J. Mol. Biol. 2004; 340: 783-795Crossref PubMed Scopus (5622) Google Scholar) was used to predict signal peptide cleavage site(s) from the 60-amino acid peptide encoded by the pCL12 open reading frame (25Link C.D. Proc. Natl. Acad. Sci. U.S.A. 1995; 92: 9368-9372Crossref PubMed Scopus (508) Google Scholar). For analysis of self-aggregation, we followed a derived protocol (32Hortschansky P. Schroeckh V. Christopeit T. Zandomeneghi G. Fändrich M. Protein Sci. 2005; 14: 1753-1759Crossref PubMed Scopus (216) Google Scholar). Solutions of 5 μm Aβ peptide in PBS, pH 7.4, were incubated at 30 °C in a 96-well microtitre plate (Wallac) with 20 μm thioflavin-T (ThT, Sigma) at a volume of 150 μl for 20 h. Measurements of ThT binding to Aβ fibrils were obtained using a Flexstation 3 fluorescence spectrophotometer (MDS Analytical Technologies) and measuring fluorescence at 482 nm (excitation = 450 nm) with a 475 nm emission cut-off filter. Data points were collected in 5-min intervals via top reading, with each cycle consisting of 3 s of orbital shaking immediately followed by the fluorescence measurement. Plates were sealed with acetate adhesive seals (MP Biomedicals) to minimize evaporative loss. ThT binding was represented as the mean relative fluorescent units from n = 6 replicate wells following subtraction of the reaction vehicle background fluorescence. Mean lag time (tl) and aggregation rates (k) were determined as described previously (32Hortschansky P. Schroeckh V. Christopeit T. Zandomeneghi G. Fändrich M. Protein Sci. 2005; 14: 1753-1759Crossref PubMed Scopus (216) Google Scholar). Fitting curves were performed using Prism (version 5.01, GraphPad). tl error was calculated by the intersects of the 95% confidence interval curves and reported as 2× S.E. Seeded Aβ-aggregation was similarly analyzed. ThT binding of 4.5 μm Aβ1–40 ("substrate") aggregation with and without 0.5 μm Aβ1–42 or Aβ3–42 ("seed") was measured as above. In addition, differing stoichiometric ratios of Cu(II), presented as Cu(II)-glycine (1:6 metal/ligand molar ratio), were co-incubated with the reactions. Aβ peptides alone (60 μm) or in seeded reactions (54 μm Aβ1–40 + 6 μm Aβ1–42 or Aβ3–42 with or without equimolar Cu(II)-glycine) were incubated in PBS, pH 7.4 (with 0.1% w/v NaN3) at 37 °C for 3 days. Samples were prepared for transmission electron microscopy by adsorbing 5 μl of reaction onto a carbon-coated Formvar film mounted on 300 mesh gold grids (ProSciTech). Prior to adsorption, the grids were rendered hydrophilic by glow discharge in a reduced atmosphere of air for 10 s. After 120 s of adsorption, samples were blotted, washed twice with (18 megaohms) milli-Q water (Millipore), and negatively stained with 1.5% aqueous uranyl acetate (ProSciTech). Transmission electron microscopy was performed using a Tecnai G2 TF30 (FEI Co.) instrument operated at 200 kV. Images were acquired digitally with an UltraScan 1000 (2k × 2k) pixels CCD camera (Gatan). Comparison of seeded aggregation was based upon previous protocols (33Jarrett J.T. Berger E.P. Lansbury Jr., P.T. Biochemistry. 1993; 32: 4693-4697Crossref PubMed Scopus (1747) Google Scholar, 34Huang X. Atwood C.S. Moir R.D. Hartshorn M.A. Tanzi R.E. Bush A.I. J. Biol. Inorg. Chem. 2004; 9: 954-960Crossref PubMed Scopus (224) Google Scholar). Aβ1–40 alone (22 μm) or 20 μm Aβ1–40 with either 2 μm Aβ1–42 or 2 μm Aβ3–42 was co-incubated in sterile filtered PBS, pH 7.4, incubation at 37 °C, in the presence or absence of (10 μm) Cu(II)-glycine (1:6 metal/ligand molar ratio). Turbidometric measurements were then performed as described previously (33Jarrett J.T. Berger E.P. Lansbury Jr., P.T. Biochemistry. 1993; 32: 4693-4697Crossref PubMed Scopus (1747) Google Scholar, 34Huang X. Atwood C.S. Moir R.D. Hartshorn M.A. Tanzi R.E. Bush A.I. J. Biol. Inorg. Chem. 2004; 9: 954-960Crossref PubMed Scopus (224) Google Scholar) using clear flat-bottomed 96-well plates (Greiner). Absorbance at 400 nm, measured using a Powerwave spectrophotometer (BioTek), was taken immediately and then again every 24 h over 5 days from n = 4 replicate wells. The differences in absorbance (ΔA400 nm) were then compared. An automatic 5-s plate agitation was incorporated prior to the analysis to evenly suspend any aggregates. Incubation was performed in a humidified chamber to prevent evaporative loss. To determine whether Aβ is truncated or modified in the C. elegans model of Aβ toxicity, we performed SELDI-TOF MS analysis using W-02 (as the capture antibody) on C. elegans expressing Aβ1–42 (expected M+H+ 4515.1 Da). Unexpectedly, neither strain CL2006 nor strain CL2120 expressed any detectable Aβ1–42 (Fig. 1A). However, in both strains, we detected a major species that had an m/z consistent with Aβ3–42 (expected M+H+ 4328.9 Da, observed m/z of 4327.5 and 4327.4 in CL2006 and CL2120, respectively). Due to the inherent limitations in the accuracy of the SELDI-TOF technique and because the molecular mass of Aβ1–40 (4329.9 Da) is only 2 Da greater than that of Aβ3–42, we did not conclusively assign the species at m/z of 4327.5 and 4327.4 as Aβ3–42 (M+H+ 4328.9) on the basis of SELDI-TOF alone. No other species of Aβ were detected. To confirm whether the species present in the C. elegans was Aβ3–42, we used bis/Bicine urea gels to resolve the Aβ species based on hydrophobicity of the peptide (30Klafki H.W. Wiltfang J. Staufenbiel M. Anal. Biochem. 1996; 237: 24-29Crossref PubMed Scopus (98) Google Scholar). Analysis of synthetic Aβ1–40, Aβ1–42, Aβ3–42, and Aβ3(pE)-42 using this technique resulted in the resolution of the Aβ peptides (Fig. 1B) in an order consistent with the calculated grand average of hydropathy (GRAVY) (35Kyte J. Doolittle R.F. J. Mol. Biol. 1982; 157: 105-132Crossref PubMed Scopus (17098) Google Scholar) scores (Aβ1–40 0.057, Aβ1–42 0.205 Aβ3–42 0.258) calculated using ProtParam (36Gasteiger E. Hoogland C. Gattiker A. Duvaud S. Wilkins M.R. Appel R.D. Bairoch A. Protein Identification and Analysis Tools on the ExPASy Server.in: Walker J.M. The Proteomics Protocols Handbook. Humana Press, Totowa, NJ2005: 571-607Crossref Google Scholar) and previous reports of Aβ3(pE)-42 (37Casas C. Sergeant N. Itier J.M. Blanchard V. Wirths O. van der Kolk N. Vingtdeux V. van de Steeg E. Ret G. Canton T. Drobecq H. Clark A. Bonici B. Delacourte A. Benavides J. Schmitz C. Tremp G. Bayer T.A. Benoit P. Pradier L. Am. J. Pathol. 2004; 165: 1289-1300Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). Analysis of Aβ in formic acid-soluble fractions using bis/Bicine urea gel combined with immunodetection with 4G8 confirmed that the species observed in two C. elegans models of Aβ (strains CL2120 and CL2006) corresponds to Aβ3–42 and that Aβ1–42 was not present (Fig. 1C). The background strain CL2122 possessed no Aβ immunoreactivity (Fig. 1C). Aging populations to 8 days, previously shown to exacerbate amyloid formation (38Florez-McClure M.L. Hohsfield L.A. Fonte G. Bealor M.T. Link C.D. Autophagy. 2007; 3: 569-580Crossref PubMed Scopus (95) Google Scholar), did not detectably alter Aβ any further. Equivalent results were observed in the TBS-soluble fractions probed with 4G8 (supplemental Fig. 1) or 6E10 (data not shown). When originally constructed, the Aβ transgene (used in both CL2006 and CL2120 strains) possessed an 18-amino acid synthetic signal peptide N-terminal to Aβ (25Link C.D. Proc. Natl. Acad. Sci. U.S.A. 1995; 92: 9368-9372Crossref PubMed Scopus (508) Google Scholar), which, it was proposed, would be cleaved to yield Aβ1–42 (Fig. 1D). We therefore re-examined how the translated protein in these transgenic C. elegans might be processed. Using SignalP 3.0 (31Bendtsen J.D. Nielsen H. von Heijne G. Brunak S. J. Mol. Biol. 2004; 340: 783-795Crossref PubMed Scopus (5622) Google Scholar), a neural networking predictor of eukaryotic secretory signal peptides, we determined the most likely position of cleavage to be between residues 20 and 21 (Fig. 1D). Cleavage at this site results in the generation of Aβ3–42. The amyloidogenic properties of Aβ3–42 are not well characterized. Therefore, we examined synthetic peptides in vitro using ThT, a diagnostic fluorescent dye that specifically binds β-sheets, such as those in amyloid oligomers. When bound, ThT undergoes a characteristic blue shift of its excitation spectrum. We found that following incubation in PBS, pH 7.4, for 20 h at 30 °C, both Aβ1–42 and Aβ3–42 increased ThT fluorescence (Fig. 2A). Kinetic parameters for each curve were then determined, showing that Aβ3–42 has a shorter lag phase and faster rate of aggregation, where for Aβ1–42, tl = 8.41 ± 0.15 (±2 S.E.) h and k = 0.30 ± 0.01 (±S.E.) h−1, and for Aβ3–42, tl = 6.25 ± 0.15 h and k = 0.52 ± 0.02 h−1. To examine the quaternary structure(s) formed by aggregation, we examined sample morphology via electron micrograph (EM) (Fig. 2, B–D). Consistent with our ThT data, and as has been reported previously, Aβ1–40 did not form fibrillar structures at pH 7.4; consequently, no aggregation kinetics were determined for Aβ1–40. Aβ1–42 formed long fibrils of length in excess of 200 nm. In contrast, Aβ3–42 formed fewer long fibrils but a greater number shorter fibrils and protofibrils (less than 100–200 nm in length) with an increased tendency to clump together. These data indicate that under these in vitro conditions, Aβ3–42, like Aβ1–42, self-aggregates and forms amyloid fibrils. We then examined whether Aβ3–42 could substitute for Aβ1–42 in the classical assay of nucleated precipitation ("seeding") of Aβ1–40 (39Jarrett J.T. Lansbury Jr., P.T. Cell. 1993; 73: 1055-1058Abstract Full Text PDF PubMed Scopus (1909) Google Scholar). We found that Aβ3–42, when compared with Aβ1–42, is the more aggressive initiator of Aβ1–40 aggregation (Fig. 2E). Kinetic parameters were determined, such that for Aβ1–42-seeded aggregation, tl = 12.21 ± 0.54 h and k = 0.46 ± 0.03 h−1, and for Aβ3–42-seeded aggregation, tl = 10.01 ± 0.67 h and k = 0.25 ± 0.01 h−1. EM analysis of aggregated Aβ1–40, seeded with either Aβ1–42 or Aβ3–42 (Fig. 2, F–H), revealed that both seeded reactions produced similar morphology of protofibril and short fibrillar structures (less than 200 nm in length), which appeared crooked or flexible. We have previously reported that Cu(II) accelerates, in vitro, Aβ1–42-seeded aggregation of Aβ1–40 (34Huang X. Atwood C.S. Moir R.D. Hartshorn M.A. Tanzi R.E. Bush A.I. J. Biol. Inorg. Chem. 2004; 9: 954-960Crossref PubMed Scopus (224) Google Scholar). To examine interactions between Cu(II) and Aβ3–42-seeded aggregation, we replicated this experiment using both ThT and turbidometric analyses (Fig. 3). We observed an Aβ:Cu(II) ratio-dependent suppression of Aβ3–42-seeded mature fibril formation (Fig. 3, C and D). However, Cu(II) caused an increase in amorphous Aβ aggregation (Fig. 3E) as observed by light scatter. These data are consistent with the effects of Cu(II) on Aβ1–42 aggregation reported previously (40Smith D.P. Ciccotosto G.D. Tew D.J. Fodero-Tavoletti M.T. Johanssen T. Masters C.L. Barnham K.J. Cappai R. Biochemistry. 2007; 46: 2881-2891Crossref PubMed Scopus (162) Google Scholar) and with transition metals increasing the α-helical content of Aβ (41Huang X. Atwood C.S. Moir R.D. Hartshorn M.A. Vonsattel J.P. Tanzi R.E. Bush A.I. J. Biol. Chem. 1997; 272: 26464-26470Abstract Full Text Full Text PDF PubMed Scopus (310) Google Scholar). The accumulation of Aβ truncations and modifications and their respective contribution to the pathogenesis of Aβ-mediated toxicity have not been studied comprehensively in simple model systems. Therefore, we initiated a study to delineate the Aβ variants that accumulate in established transgenic C. elegans models expressing Aβ. Surprisingly, we discovered that only Aβ3–42 was expressed rather than the predicted Aβ1–42. We found concordant data using both SELDI-TOF MS, as well as immunoblot analysis of samples resolved by hydrophobicity. Neither technique, nor the antibodies used, are expected to bias detection of specific Aβ truncations. The detection of a single Aβ species, rather than a range of variants, suggests simple post-translational processing. Both the original Aβ-expressing strain, CL2006 (25Link C.D. Proc. Natl. Acad. Sci. U.S.A. 1995; 92: 9368-9372Crossref PubMed Scopus (508) Google Scholar), and a derived strain, CL2120 (26Fay D.S. Fluet A. Johnson C.J. Link C.D. J. Neurochem. 1998; 71: 1616-1625Crossref PubMed Scopus (143) Google Scholar), expressed only Aβ3–42. The likely cause is aberrant cleavage of the N-terminal signal peptide incorporated in the C. elegans transgene (Fig. 1D). Although signal peptide cleavage (including that in C. elegans) is not fully understood, prediction accuracy of cleavage sites has improved greatly (42Emanuelsson O. Brunak S. von Heijne G. Nielsen H. Nat. Protoc. 2007; 2: 953-971Crossref PubMed Scopus (2610) Google Scholar), especially over the 14 years since the Aβ-expressing C. elegans model was first reported (25Link C.D. Proc. Natl. Acad. Sci. U.S.A. 1995; 92: 9368-9372Crossref PubMed Scopus (508) Google Scholar). We examined the accumulated Aβ in young (4-day-old) and aged (8-day-old) post-reproductive C. elegans adults and observed no additional truncations, covalent modifications (including pyroglutamate), or cross-linking. This model, therefore, represents a relatively clean system in which further Aβ modifications could be studied via additional transgenics. The C. elegans genome encodes a potential QC ortholog, H27A22.1 (data not shown). However, it is not known whether an analogous enzyme activity is found in the same cell type that expresses the transgenic Aβ (body wall muscle cells). The H27A22.1 promoter does not appear active in body wall muscles (43Hunt-Newbury R. Viveiros R. Johnsen R. Mah A. Anastas D. Fang L. Halfnight E. Lee D. Lin J. Lorch A. McKay S. Okada H.M. Pan J. Schulz A.K. Tu D. Wong K. Zhao Z. Alexeyenko A. Burglin T. Sonnhammer E. Schnabel R. Jones S.J. Marra M.A. Baillie D.L. Moerman D.G. PLoS. Biol. 2007; 5: e237Crossref PubMed Scopus (290) Google Scholar). Therefore, additional genetic manipulations may be useful to examine interactions between QC and Aβ3–42 in C. elegans. Proteotoxicity of the (cytoplasmic) expressed Aβ3–42 in this C. elegans model has been clearly demonstrated (6Link C.D. Genes Brain Behav. 2005; 4: 147-156Crossref PubMed Scopus (79) Google Scholar, 26Fay D.S. Fluet A. Johnson C.J. Link C.D. J. Neurochem. 1998; 71: 1616-1625Crossref PubMed Scopus (143) Google Scholar, 38Florez-McClure M.L. Hohsfield L.A. Fonte G. Bealor M.T. Link C.D. Autophagy. 2007; 3: 569-580Crossref PubMed Scopus (95) Google Scholar, 44Cohen E. Bieschke J. Perciavalle R.M. Kelly J.W. Dillin A. Science. 2006; 313: 1604-1610Crossref PubMed Scopus (692) Google Scholar). The expressed Aβ3–42 aggregates and forms amyloid in vivo. We observe that in vitro, Aβ3–42 forms fibrils. In addition, we found that Aβ3–42-seeded aggregation of Aβ1–40 substrate is more rapidly initiated when compared with Aβ1–42-seeded aggregation. Furthermore, Aβ3–42-seeded aggregation is exacerbated in the presence of Cu(II). These results are important because of the increasing evidence that Cu(II) is constitutively enriched in the synapse and is dysregulated in AD, where it pools in the vicinity of the synapse and within plaques (45Bush A.I. J. Alzheimers. Dis. 2008; 15: 223-240Crossref PubMed Scopus (236) Google Scholar). Our observation that Cu(II) promotes in vitro Aβ3–42 aggregation (Fig. 3, C and E) is consistent with the observation that Aβ amyloid formation is induced in C. elegans (CL2120, the same strain as used in our study) by exposing animals to Cu(II) and rescued by treatment with clioquinol (46Minniti A.N. Rebolledo D.L. Grez P.M. Fadic R. Aldunate R. Volitakis I. Cherny R.A. Opazo C. Masters C. Bush A.I. Inestrosa N.C. Mol. Neurodegener. 2009; 4: 2Crossref PubMed Scopus (36) Google Scholar). Clioquinol binds Cu(II), abolishes the redox activity of Aβ:Cu(II) complexes, rescues amyloid neuropathology, and improves cognition in amyloid precursor protein transgenic mice (28Adlard P.A. Cherny R.A. Finkelstein D.I. Gautier E. Robb E. Cortes M. Volitakis I. Liu X. Smith J.P. Perez K. Laughton K. Li Q.X. Charman S.A. Nicolazzo J.A. Wilkins S. Deleva K. Lynch T. Kok G. Ritchie C.W. Tanzi R.E. Cappai R. Masters C.L. Barnham K.J. Bush A.I. Neuron. 2008; 59: 43-55Abstract Full Text Full Text PDF PubMed Scopus (571) Google Scholar). Our data suggest that Aβ3–42 exhibits similar interactions with metal ions that have been characterized previously for Aβ1–42 (47Atwood C.S. Moir R.D. Huang X. Scarpa R.C. Bacarra N.M. Romano D.M. Hartshorn M.A. Tanzi R.E. Bush A.I. J. Biol. Chem. 1998; 273: 12817-12826Abstract Full Text Full Text PDF PubMed Scopus (936) Google Scholar, 48Atwood C.S. Scarpa R.C. Huang X. Moir R.D. Jones W.D. Fairlie D.P. Tanzi R.E. Bush A.I. J. Neurochem. 2000; 75: 1219-1233Crossref PubMed Scopus (573) Google Scholar). This is consistent with residues 3–9 of Aβ mediating a transition to β-sheet and insoluble aggregates upon Cu(II) binding (49Miura T. Mitani S. Takanashi C. Mochizuki N. J. Inorg. Biochem. 2004; 98: 10-14Crossref PubMed Scopus (20) Google Scholar). Although we examined the effects of Cu(II), it should be noted that oligomerization of Aβ may be induced by other metals, such as synaptic zinc (50Deshpande A. Kawai H. Metherate R. Glabe C.G. Busciglio J. J. Neurosci. 2009; 29: 4004-4015Crossref PubMed Scopus (192) Google Scholar). In summary, we find that by serendipity a C. elegans model of cytoplasmic Aβ3–42 expression and toxicity has been generated that has value as a model for AD despite not expressing Aβ1–42. In AD, Aβ3–42 can be modified to Aβ3(pE)-42 that may represent a key species in Aβ-derived neurotoxicity. This C. elegans model can now be exploited to provide further insight into the biology of this truncated Aβ in AD pathogenesis. We thank Christopher D. Link (University of Colorado), Simon James (MHRI), and members of the Masters, Barnham (University of Melbourne), Cherny, and Bush laboratories (MHRI) for helpful discussions and critical reading of this manuscript. In addition, we acknowledge the technical expertise of the staff of the University of Melbourne Bio21 EM Unit for their contribution to this work. All nematode strains were provided by the Caenorhabditis Genetics Center funded by the U. S. National Institutes of Health National Center for Research Resources. Download .zip (.16 MB) Help with zip files
Referência(s)