FAMILIAL DANISH DEMENTIA: CO-EXISTENCE OF DANISH AND ALZHEIMER AMYLOID SUBUNITS (ADan AND Aβ) IN THE ABSENCE

OF COMPACT PLAQUES* Yasushi Tomidokoro1, Tammaryn Lashley5, Agueda Rostagno1, Thomas A. Neubert3,4,

Marie Bojsen-Møller6, Hans Braendgaard7, Gordon Plant8, Janice Holton5, Blas Frangione1,2, Tamas Révész5, and Jorge Ghiso1,2

From the Departments of 1Pathology, 2Psychiatry, and 3Pharmacology and the 4Skirball Institute for Biomolecular Medicine, New York University School of Medicine, New York, NY 10016; 5Department of Neuropathology and Brain Bank, Institute of Neurology, London WC1N3BG, UK; Departments of 6Neuropathology and 7Neurology, Ǻrhus University Hospital, Ǻrhus DK8000, Denmark; 8The National Hospital for Neurology and Neurosurgery, London WC1N3BG, UK.

Running title: Biochemical analysis of ADan and Aβ in FDD

Address correspondence to: Jorge Ghiso, Departments of Pathology and Psychiatry, New York University School of Medicine, 550 First Avenue (TH-432), New York, New York 10016, Tel. 212 263-7997; Fax. 212 263-6751; E-Mail: ghisoj01@popmail.med.nyu.edu Familial Danish Dementia (FDD) is an early onset autosomal dominant neurodegenerative disorder linked to a genetic defect in the BRI2 gene and clinically characterized by dementia and ataxia. Cerebral amyloid and pre-amyloid deposits of two unrelated molecules (ADan and Aβ), absence of compact plaques, and neurofibrillary degeneration indistinguishable to that observed in Alzheimer’s disease (AD) are the main neuropathological features of the disease. Biochemical analysis of extracted amyloid and pre-amyloid species indicates that as the solubility of the deposits decrease, the heterogeneity and complexity of the extracted peptides exponentially increase. Non-fibrillar deposits were mainly composed of intact ADan1-34 and its N-terminally modified (pyroglutamate) counterpart together with Aβ1-42 and Aβ4-42 in ~1:1 mixture. The post-translational modification, glutamate to pyroglutamate, was not present in soluble circulating ADan. In the amyloid fractions, ADan was heavily oligomerized, highly heterogeneous at the N- and C-terminus and, when intact, its N-terminus was post-translationally modified (pyroglutamate) while Aβ was mainly Aβ4-42. In all cases, the presence of AβX-40 was negligible, a surprising finding in view of the prevalence of Aβ40 in vascular deposits observed in sporadic and familial AD, Down’s syndrome and normal

aging. Whether the presence of the two amyloid subunits is imperative for the disease phenotype or just reflects a conformational mimicry remains to be elucidated; nonetheless a specific interaction between ADan oligomers and Aβ molecules was demonstrated in vitro by ligand blot analysis using synthetic peptides. The absence of compact plaques in the presence of extensive neurofibrillar degeneration strongly suggests that compact plaques, fundamental lesions for the diagnosis of AD, are not essential for the mechanism of dementia. Familial Danish Dementia (FDD), originally described by Strömgren et al. in 1970 as heredopathia opthalmo-oto-encephalica, is an early onset autosomal dominant neurodegenerative disorder characterized by cataracts, deafness, progressive ataxia and dementia (1,2). Retinal neovascularizations eventually resulting in vitreous hemorrhage and neovascular glaucoma may also be present (3). Cataracts occurring around 20 years of age seem to be the first manifestation of the disease. Hearing loss develops ten to twenty years later and cerebellar ataxia occurs, in general, shortly after the age of 40. Paranoid psychosis usually appears after the age of 50 evolving to cognitive impairment and dementia in the majority of the cases. Most patients die in their fifth or sixth decade of life.

1

JBC Papers in Press. Published on August 9, 2005 as Manuscript M504038200

Copyright 2005 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

Neuropathologically, FDD shares several features with Alzheimer’s disease (AD), among them widespread cerebrovascular amyloidosis, parenchymal amyloid and pre-amyloid lesions and neurofibrillary degeneration (4,5). Vascular and perivascular amyloid, parenchymal pre-amyloid lesions, as well as scarce neuritic plaques in the hippocampus are mainly composed of ADan peptides (4). ADan deposits co-localize with dystrophic neurites surrounding the plaques and with neurofibrillary tangles immunoreactive with hyperphosphorylated tau antibodies. As judged by their paired helical filaments ultrastructure, their reactivity in immunohistochemical analysis and their hyperphosphorylated tau pattern in western blot analysis, these tangles are strikingly similar (if not identical) to those present in AD (5). Surprisingly, detailed anti-Aβ immunohisto- chemical survey of different brain areas from all FDD available autopsy cases [three family members from different generations including the case analyzed in the present study (5)] unequivocally identified Aβ co-deposited with ADan mainly in vascular and perivascular amyloid lesions although in a smaller scale the co-deposition was also observed in parenchymal pre-amyloid deposits (5). The fact that Aβ co-deposition was a constant feature of all three FDD cases available argues against a coincidental non-specific interaction between two hydrophobic peptides. FDD patients are carriers of a genetic defect in the coding region of the BRI2 gene located in the long arm of chromosome 13. The wild type BRI2 gene (also known as ITM2B) encodes a 266-residues type II single-spanning transmembrane protein (BRI2) (6-8) which under normal conditions is proteolytically processed by a furin-like protease that produces a single cleavage between Arg243 and Glu244, releasing a C-terminal 23-residues peptide (9). In patients affected with FDD, a decamer duplication insertion (TTTAATTTGT) between codons 265 and 266 in the BRI2 gene 3’ (just one codon before the normal stop codon 267), produces a frame-shift that eliminates the stop signal and generates a longer-than-normal precursor, namely ADanPP. Furin-like proteolytic processing of ADanPP at the same Arg243-Glu244 peptide bond results in the release of a 34 amino acids-long C-terminal peptide ADan which deposits in the form of amyloid fibrils in different

brain regions of FDD patients, particularly in limbic structures (4). The ten-nucleotide duplication insertion observed in FDD patients is not the only BRI2 genetic defect that is linked to a neurodegenerative disorder. A single point mutation at the stop codon of the same gene was previously found to be associated with familial British dementia (FBD) (6). This early-onset disorder, clinically characterized by progressive dementia, cerebellar ataxia and spastic paraparesis, was first described by Worster-Drought et al. in 1933 as a familial presenile dementia with spastic paralysis (10-13). The neuropathology of FBD is remarkably similar to that of FDD and AD, particularly in regard to the presence of neurofibrillary tangles ultra-structurally composed of classical paired helical filaments (14,15). As in FDD, the stop to Arg point mutation in FBD also eliminates the BRI2 stop codon and generates a longer-than-normal precursor ABriPP. FBD amyloid deposits in the form of cerebral amyloid angiopathy, pre-amyloid lesions and amyloid plaques are composed of ABri, an amyloid subunit that is also generated by furin proteolytic processing and shares 100% identity with ADan in its first 22 amino acids but that is completely different in its 12 C-terminal residues (6). Interestingly and in spite of the similarities between FDD and FBD, detailed immuno-histochemical studies never found Aβ co-localized with ABri deposits (15,16). We describe herein the isolation and biochemical characterization of ADan and Aβ species deposited in both, vessels and parenchymal FDD lesions. Non-fibrillar pre-amyloid lesions were strikingly less complex than their fibrillar counterparts, poorly oligomerized, only partially post-translationally modified and minimally proteolytically degraded. Peptides ending at residue 42 constituted the main Aβ species in both, amyloid and pre-amyloid deposits. Surprisingly, and against the dogma that Aβ peptides ending at residue 40 are always the main components of cerebrovascular lesions in Aβ-related cerebral amyloidosis, AβX-42 species were the predominant Aβ components in FDD deposits. Of note, a specific interaction between synthetic ADan and AβX-42 peptides was demonstrated by ligand blot analysis, pointing to this peptide-peptide interaction as the potential mechanism for their co-localization in vivo.

2

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

EXPERIMENTAL PROCEDURES

Materials - Mouse monoclonals 4G8 (anti-Aβ17-24) and 6E10 (anti-Aβ1-17) were purchased from Signet Laboratories (Dedham, MA); mouse monoclonal 6F/3D (anti-Aβ8-17) was from Dako Corp. (Carpinteria, CA); mouse monoclonal AT8 (anti-tau phosphorylated Ser202/Thr205) was from Innogenetics (Gent, Belgium). Rabbit polyclonals anti-Aβ40 and anti-Aβ42 antibodies were obtained from Biosource International (Camarillo, CA). Rabbit polyclonal antibodies anti-ADan (Ab5282, raised against ADan C-terminal amino acids 22-34) and anti-ABri (Ab338, raised against ABri C-terminal amino acids 22-34) were prepared in our laboratory (4,6) and the IgG fractions further purified using Gamma-bind G Sepharose (Amersham Bioscience, Piscataway, NJ) using the manufacturer’s standard protocol. Paramagnetic beads coated with anti-rabbit and anti-mouse IgG (Dynabeads M-280) were from Dynal Biotech (Brown Deer, WI). Micro reverse-phase chromatography tips (Zip-Tip C4) were purchased from Millipore (Billerica, MA). SDS-removal reagent SDS-OUT was from Pierce (Rockford, IL). Chemicals were from Sigma (St. Louis, MO). Complete protease inhibitor cocktail was purchased from Roche (Indianapolis, IN). Protein content was measured by the Bicinchoninic/BCA Protein assay (Pierce) employing bovine serum albumin as standard. Synthetic peptides Aβ1-42 (DAEFRHDSGYEV HHQKLVFFAEDVGSNKGAIIGLMVGGVVIA), Aβ4-42 (FRHDSGYEVHHQKLVFFAEDVGSN KGAIIGLMVGGVVIA), ADan1-34 (pEASNCFA IRHFENKFAVETLICFNLFLNSQEKHY) and ABri1-34 (pEASNCFAIRHFENKFAVETLICSR TVKKNIIEEN), the last two featuring N-terminal pyroglutamate and internal disulphide bonds between cysteine residues at positions 5 and 22, were synthesized at the W. M. Keck Facility at Yale University using N-tert-butyloxycarbonyl chemistry, purified by reverse phase-high performance liquid chromatography, their molecular mass corroborated by Matrix-Assisted Laser Desorption / Ionization Time-of-Flight (MALDI-TOF) mass spectrometry, and their concentrations assessed by amino acid analysis.

Clinical data - Frozen brain tissue was available from a male patient who developed blurred vision, nystagmus, positive Romberg’s sign and mild memory loss at the age of 25. He underwent bilateral vitrectomy few years later. Transient ischemic attacks occurred at the age of 31. After the development of double vision, ataxia and dementia with threatening behavior, he died at the age of 43 from bronchopneumonia (5). In this patient, the TTTAATTTGT decamer insertion at codon 265 of the BRI2 gene, characteristic of FDD, was previously confirmed and reported (4) and its ApoE genotype, tested as described below, was found to be ε4/ε3. Immunohistochemistry - Immunohistochemical investigations were carried out in three previously reported cases of FDD (5) using 7 µm thick sections of formalin fixed paraffin embedded tissue samples. The antibodies used included anti-ADan polyclonal 5282 (1:2,000), anti-Aβ monoclonal 6F/3D (1:100), and the anti-tau monoclonal AT8 (1:600) recognizing phosphorylated Ser202/Thr205. For confocal microscopy, 20 µm tissue sections were prepared for double staining with the 5282 and 6F/3D antibodies. Tissue sections were pre-treated with 99% formic acid (FA) and pressure cooker in citrate buffer. Sections were initially incubated with the 5282 antibody followed by the appropriate secondary antibody and ABC complex (Dako). Antibody 5282 binding was visualized by applying the tetramethylrhodamine signal amplification kit (PerkinElmer Life and Analytical Sciences, Boston, MA). This step was followed by incubating the sections with the 6F/3D antibody overnight, treating with the appropriate secondary antibody and the ABC complex. Antibody binding was visualized with the fluorescein signal amplification kit (PerkinElmer Life and Analytical Sciences). A Leica TCS SP confocal microscope running TCS NT control software or a Leica DMRE upright microscope using a 3 channel scan head and Argon/Krypton laser were used for the confocal studies. Sequential tissue fractionation - Leptomeningeal vessels and frontal cortex were used for the biochemical studies described below. Frontal cortex was further dissected into gray and white matter while large vessels were separately

3

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

analyzed. The extraction strategy took advantage of the differential solubility properties of pre-amyloid [usually poorly soluble in 10 mM phosphate buffer, pH 7.4 containing 2.7 mM KCl and 137 mM NaCl (PBS) but soluble in 2% sodium dodecyl sulphate (SDS)] and amyloid structures (soluble in 70-98% FA). Samples (typically 2.5 g of gray matter) were homogenized in 12.5 ml of cold PBS containing 2X protease inhibitor cocktail (Complete) using a Douncer glass homogenizer immersed on ice. After homogenization, small vessels were collected by filtration through a 70 µm nylon mesh, combined with the dissected leptomeningeal vessels (typical yield: ~250 mg of vessels per 2.5 g of gray matter) and further extracted as a vessels fraction (12.5 ml of PBS-2X protease inhibitors / 250 mg vessels) while the filtrates were subjected to ultracentrifugation in a XL100K ultracentrifuge (Beckman Coulter, Fullerton, CA) using a Beckman 70.1 Ti rotor at 112,000 g for 1 hour at 4°C. The resultant supernatants were analyzed as PBS-extracted fractions and the pellets were re-suspended in 20 mM Tris, pH 7.4 containing 2% SDS (BioRad, Hercules, CA) centrifuged as above although raising the temperature to 10°C to avoid SDS crystallization, and the resultant supernatants analyzed as SDS-extracted (pre-amyloid - rich) fractions. Finally, the Congo-red positive SDS-insoluble material enriched in amyloid fibrils was further extracted in 70% (v/v) FA in water, centrifuged at 14,000 rpm using a 5417R microcentrifuge (Eppendorf, Westbury, NY) and the supernatants analyzed as FA-extracted (amyloid) fractions. Immunoprecipitation experiments - a) For tissue-extracted ADan or Aβ peptides: Fifty microliters of paramagnetic beads coated with either goat anti-rabbit IgG or anti- mouse IgG (Dynabeads M-280, Dynal) were allowed to interact with 6 µg of purified IgG from antiserum 5282 or a combination of 3 µg each of 4G8 and 6E10, for the immunoprecipitation of ADan and Aβ respectively. After washing three times in PBS, beads were blocked in 0.1% (w/v) bovine serum albumin in PBS. Antibody-coated beads were incubated with either 20 µl (~4% of the total fraction) of FA-extracted amyloid fraction previously neutralized in ~800 µl of 0.5 M Tris-base pH 11 or 20% of the total volumes of each,

PBS-soluble or SDS-extracted fractions. In the case of the SDS-extracted fractions, SDS was mostly removed prior to immunoprecipitation with SDS-OUT sodium dodecyl sulfate precipitation reagent employing supplier’s specifications. Elution of the bound material from the beads was performed by different methods, according to the technique used for the studies that followed. For MALDI-TOF mass spectrometry analysis, beads were subsequently washed three times with PBS, three times with water, and the bound peptides eluted in 5 µl of a 4:4:1 mixture of isopropyl alcohol-water-formic acid (17). For western blot analysis, beads were suspended in 20 µl of Tris-Tricine SDS sample buffer (BioRad) containing 10% (v/v) β-mercaptoethanol (Sigma) and directly applied onto the 16% Tris-Tricine gels for SDS-PAGE analysis (17). In some cases, 5 µl of 1M dithiothreitol (Sigma) was used for sample reduction instead of β-mercaptoethanol. To verify the existence of peptide-peptide interactions between Aβ and ADan, immuno-precipitation experiments using either anti-ADan or anti-Aβ antibodies were carried out in SDS-extracts from cerebral vessels using the same immunoprecipitation protocol described above. Samples immunoprecipitated with the mixture of 4G8 plus 6E10 anti-Aβ antibodies were probed in western blot analysis with anti-ADan antibody (Ab5282) and conversely, samples immuno-precipitated with anti-ADan antibody were probed for Aβ with either a mixture of 4G8 plus 6E10, anti-Aβ40 or anti-Aβ42 antibodies. Immuno-reactivity was detected by chemiluminescence, as described below. b) For circulating sADan: Plasma from two carriers of the FDD genetic defect (44 and 24 years-old) as well as from 10 non-carriers were analyzed by immunoprecipitation. Fifty microliters of goat anti-rabbit IgG-coated paramagnetic beads were allowed to interact with antibody 5282, as described (17). After washing and blocking as before, beads were incubated with 1 ml of a 1:1 dilution of plasma in RIPA buffer (1% Triton X-100 in 50 mM Tris, pH 8.0 containing 150 mM NaCl, 0.5% cholic acid, 0.1% SDS, 5 mM EDTA and Complete protease inhibitor). Bound peptides were eluted for MALDI-TOF mass spectrometry or western blot analysis as described above.

4

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

MALDI-TOF analysis - Before mass spectrometry analysis, extracted peptides from PBS-, SDS-, and FA-fractions were further purified by immunoprecipitation, as described above. Due to the abundance of ADan in FA fractions, immunoprecipitation was not necessary for its identification by mass spectrometry. In some cases, FA-extracts were alternatively purified by micro reverse-phase chromatography using Zip-TipC4 according to the manufacturer’s protocol utilizing 90% (v/v) acetonitrile - 0.1% (v/v) trifluoroacetic acid in water for elution. MALDI-TOF mass spectrometry analysis was performed at the New York University Protein Analysis Facility. Samples were analyzed with 10 mg/ml α-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 0.1% trifluoroacetic acid (Sigma-Aldrich) matrix (feasible for <10,000 Da mass range) on a Micromass TofSpec-2E MALDI-TOF mass spectrometer in linear mode using standard instrument settings (17). Under these experimental acidic conditions, which also renders non-covalent binding unobservable (18), mostly monomeric species of ADan and Aβ were detected. Internal and / or external calibration was carried out using human adrenocorticotropic hormone peptide 18-39 (average mass = 2,465.68 Da) and insulin (average mass = 5,733.49 Da) as standards. FindPept tool from the ExPASy Proteomics server (http://us.expasy.org/tools) was used to assign the experimental mass values to specific peptide sequences and to search for the presence of post-translational modifications. Western Blot Analysis - Samples from each of the extracts, either before or after immuno-precipitation, were separated on a 16% Tris-Tricine SDS-PAGE and electrotransfered for 45 min at 400 mA onto polyvinylidene difluoride membranes (Immobilon-P, Millipore) using 10 mM 3-cyclohexylamino-1-propanesulfonic acid (Sigma) buffer, pH 11 containing 10% (v/v) methanol. After blocking in 5% non-fat milk in PBS containing 0.1% Tween20 (Sigma), the membranes were immunoreacted with the following antibodies: anti-ADan 5282 (2 µg/ml), anti-Aβ 4G8 (1 µg/ml), anti-Aβ 6E10 (1 µg/ml), anti-Aβ40 (0.05 µg/ml) or anti-Aβ42 (0.05 µg/ml) followed by either anti-rabbit or anti-mouse horseradish peroxidase-labeled F(ab’)2 (Amersham). Signals were developed with Super

Signal (Pierce) and exposed to Hyperfilm ECL (Amersham). To enhance Aβ immunoreactivity, membranes after transfer were boiled in PBS for 5 min (19) before incubation with the specific anti-Aβ antibodies. When necessary for re-probing, membranes were incubated in Restore Western Blot Stripping Buffer (Pierce) for 15 min at room temperature, re-blocked and probed with another antibody. Ligand Blot Analysis – Five micrograms each of synthetic ADan1-34 and ABri1-34 peptides were separated on a 16% Tris-Tricine SDS-PAGE and electrotransfered onto Immobilon-P membranes as described above. After blocking with 5% non-fat milk in PBS, membranes were separately incubated at 37°C for 3 hours with synthetic Aβ1-42 and Aβ4-42 peptides (1 µg/ml in PBS/milk). Bound Aβ was assessed by overnight incubation at 4°C with a mixture of 4G8 and 6E10 (0.5 µg/ml each in PBS/milk) followed by 1 hour incubation at room temperature with anti-mouse horseradish peroxidase-labeled F(ab’)2 and chemiluminescence detected as above. Fifty nanograms of both ADan and ABri peptides were also subjected to Western blot analysis using a mixture of Ab5282 (2 µg/ml) and Ab338 (1 µg/ml) and used as controls. Amino Acid Sequence Analysis - N-terminal amino acid sequence analysis of isolated amyloid species extracted in FA was carried out by automatic Edman degradation on a 494 Procise Protein Sequencer (Applied Biosystems). Aliquotes from FA-solubilized amyloid fractions were separated on 16% Tris-Tricine SDS-PAGE, transfered onto Immmobilon-P as described above and visualized with 0.125% (w/v) Coomassie Blue R-250 in 50% (v/v) methanol. Relevant components were excised and N-terminal sequenced. ApoE genotyping - Genomic DNA was extracted from brain tissue using DNeasy Tissue kit (Qiagen, Valencia, CA) and the ApoE genotyping was performed as described (20), with minor modifications. Briefly, 244 bp polymerase chain reaction products were generated using reverse primer F4 (5’-ACAGAATTCGCCCCGGCCTGG TACAC-3’) and forward primer F6 (5’-TAAGC TTGGCACGGCTGCCAAGGA-3’), digested for 4 hours at 37°C with 20 units of HhaI (Promega, Madison, WI), subjected to electrophoresis on

5

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

non-denaturing 10% polyacrylamide gels and visualized under ultraviolet light after staining with ethidium bromide.

RESULTS FDD main histopathological features - Figure 1 illustrates the extent of ADan deposition (Figure 1A) and associated tau pathology (Figure 1B) in patients with FDD. As shown in panel A, ADan forms loose, ill-defined parenchymal diffuse (pre-amyloid) plaques (double arrow) and deposits in blood vessel walls (arrow) in FDD. The parenchymal lesions are usually Thioflavin S / Congo red-negative and ultrastructurally appear as mostly amorphous (non-fibrillar) electron-dense material with sparse fibrils, while the vascular deposits are of amyloid nature. The presence of Aβ, sometimes co-localizing with ADan deposits, was previously localized immunohistochemically in three affected family members belonging to different generations (5) and is further illustrated below. Soluble ADan is present in the blood of carriers of the FDD genetic defect - Western blot analysis of immunoprecipitated plasma samples obtained from 2 carriers of the FDD genetic defect revealed the presence of an ~4 kDa component (Figure 2A) immunoreactive with Ab5282 (specific for the 12 C-terminal amino acid residues of ADan, a fragment not existent in the normal population). Soluble ADan was consistently monomeric in SDS-PAGE; in mass spectrometry, its m/z experimental value rendered a single peak of 4,063.4 (Figure 2A) in good agreement with the expected mass of 4063.6 corresponding to the full-length peptide bearing glutamate at the N-terminus and a single disulphide-bond between cysteine residues 5 and 22. Soluble ADan was not detected in plasma samples from non-carriers of the disease (n=10) using identical experimental conditions (not shown). Intact and N-terminally post-translationally modified ADan are predominant components of pre-amyloid and amyloid deposits - FDD parenchymal samples and cerebral vessels were sequentially extracted in PBS, 2% SDS and 70% FA and analyzed for the presence of ADan species by mass spectrometry and western blot. In some

cases, particularly in samples extracted with FA which were more abundant, we also conducted N-terminal amino acid sequence analysis. As expected, the vast majority of ADan was recovered from the amyloid extracts; however, using a combination of immunoprecipitation and western blot / mass spectrometry, it was also possible to detect ADan species in all other fractions. Compared with the plasma specimens there were noticeable differences in terms of degree of aggregation and N-terminal post-translational modifications. Figure 2B illustrates ADan species in PBS-extracts. PBS-soluble ADan was mainly monomeric with some dimers barely visible in both parenchymal and vascular fractions, as indicated by the western blot. Mass spectrometry revealed that these fractions were composed of a mixture of two ADan species differing in 18 mass units 4045.2 for parenchyma / 4045.3 for vessels (expected 4045.6) and 4063.5 for parenchyma / 4062.7 for vessels (expected 4063.6), a difference equivalent to the mass of one molecule of water and likely corresponding to ADan-pE and ADan-E species in a ~60:40 ratio (see below). Since parenchymal and vessels fractions were representative of the same amount of original tissue, it is apparent by the western blot analysis that the PBS-soluble ADan species in FDD are more abundant in vessels than in brain parenchyma. Similar results were obtained with the SDS-extracted fractions, although as judged by the monomers:dimers ratio, samples had less tendency to remain monomeric. As indicated in Figure 2C, ADan was predominantly dimeric in parenchymal extracts (ratio monomers:dimers ~1:3) while in vascular extracts the monomeric form was still predominant (ratio monomers:dimers ~5:1). By mass spectrometry analysis, both extracts contained the same two ADan species seen in the PBS-extracts and differing in 18 units of mass in a similar 60:40 ratio. The experimental masses, 4043.6 for parenchyma /4045.6 for vessels and 4062.0 for parenchyma / 4063.6 for vessels were in good aggreement with the expected masses of the post-translationally modified ADan-pE (4045.6) and the wild-type ADan-E (4063.6), with a single disulphide-bond between cysteine residues 5 and 22. Analysis of the FA-extracts revealed more complexity in terms of composition and degree of

6

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

aggregation than the other fractions. The interpretation of the mass spectrometry analysis was also more challenging due to the additional presence of mono-, di-, and tri-formylated species, likely on lysine and/or serine residues (21), artifactually generated during the extraction procedure. As illustrated in Figure 3, both parenchymal and vascular lesions were composed of a heterogeneous mixture of post-translationally modified and N- and C-terminally truncated ADan species. The predominant components were ADan1-33pE, ADan1-28pE and ADan1-34pE although the presence of ADan3-28, ADan1-27pE, ADan1-30pE, ADan3-33, ADan3-34, ADan 1-33E and ADan 1-34E was consistently demonstrated (See Figure 3A, mass spectrometry spectra and Table). This ample heterogeneity was also reflected in the extensive ADan aggregation observed in SDS-PAGE when compared with PBS- and SDS-extracts. Monomers, dimers, trimers and larger aggregates were clearly highlighted in parenchymal FA-extracts. The aggregation pattern was even more prominent in FA-extracts obtained from vessels (Figure 3B) in which the bulk of amyloid deposition occurs in FDD patients. The abundance of FA-soluble ADan species compared with their water-soluble counterparts was not only reflected in the intensity of the immunoreactivity of both parenchymal and vascular extracts but in the amount of protein necessary to visualize ADan species (1,500-fold and 5,000-fold more protein was required for the PBS- and SDS-extracts of parenchyma and vessels, respectively, than for the FA-fractions). The N-terminal heterogeneity of ADan peptides in FA-extracted fractions was confirmed by N-terminal sequence analysis in both parenchymal and vascular extracts. The minor sequences retrieved, SNXFAIR… and EASNXF…, corresponded to ADan starting at positions 3 and 1 respectively, in ~3:1 ratio. Yield calculations revealed that these sequences represented about 10-15% of the material loaded into the sequencer cartridge, indicating that 85-90% of the material did not undergo Edman degradation and consistent with the presence of cyclic N-terminal pyroglutamate. These results are in agreement with the relevance of the N-terminal pyroglutamate ADan species identified by mass spectrometry (Figure 3) and with previous published data in leptomeningeal vessels (4).

Aβ42 colocalizes with ADan parenchymal and vascular lesions - All parenchymal and vascular extracts previously analyzed for ADan (PBS-, SDS- and FA-extracts) were also evaluated for the presence of Aβ species via immunoprecipitation/ mass spectrometry and immunoprecipitation/ western blot using four different anti-Aβ antibodies: 4G8 (anti-Aβ17-24), 6E10 (anti-Aβ1-17), and the C-terminal specific anti-Aβ40 and anti-Aβ42. As illustrated in Figure 4 (panels A and B), Aβ species were detectable in PBS-extracts from parenchyma and vessels fractions. Parenchymal extracts were immunoreactive with antibodies 4G8, 6E10 and anti-Aβ42 but not with anti-Aβ40. Monomers and dimers of AβX-42 were observed with an estimated monomers:dimers ratio of ~1:2 (Panel A). Although low signal was retrieved in the mass spectrometry there was a peak with a mass suspected to correspond to the fragment Aβ4-34 (experimental 3473.5; expected 3472.9); no clear peaks of AβX-42 were detected (panel A and Table). Vascular extracts, on the other hand, rendered a more defined picture both at the western blot and mass spectrometry levels. As indicated in panel B, Aβ species were immunoreactive with all antibodies, including anti-Aβ40. Several differences were spotted when compared with the parenchymal counterparts: i) the signals for monomers and dimers were stronger, ii) Aβ40 immunoreactivity (although not strong) was seen with dimeric species, a signal not detected in parenchymal fractions, and iii) the monomeric component detected by 6E10 and anti-Aβ42 was clearly composed of two bands of very close molecular mass. As judged by mass spectrometry analysis, the major Aβ species in the sample were Aβ1-42, Aβ4-42, and Aβ1-40, (Panel B and Table), being the latter a minor component. Interestingly, the immunoprecipitation procedure using a mixture of 4G8 + 6E10 antibodies prior to the MALDI-TOF analysis also retrieved full-length ADan (both ADan-pE and ADan-E), suggestive of a protein-protein interaction between both peptides (see below). The identification of Aβ4-42 and Aβ1-42 differing in 315 units of mass strongly argues that these species are the components of the double band with the electrophoretic mobility of the monomeric species observed in the 4 kDa molecular mass of the western blot with 6E10 and anti-Aβ42 (Panel B).

7

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

When the SDS-extracts were analyzed by the same procedures, the results obtained were similar to those in the PBS-extracts although i) under the same protein load the immunoreactive signals were stronger suggesting a relative enrichment on Aβ species, and ii) the dimeric forms were more relevant. Parenchymal extracts were immunoreactive with 4G8, 6E10 and anti-Aβ42, but not with anti-Aβ40 (Panel C). Monomeric species, specially those immuno-reactive with 6E10 seemed to be composed of more than one band in western blot, very similar to the pattern shown in panel B. Mass spectrometry indicated that this extracts mainly contained Aβ1-42 and its N-terminally truncated derivative Aβ4-42 whereas Aβ1-40 was a minor component (Panel C and Table). The double peaks obtained in this sample originate in the presence of Aβ peptides bearing oxidized Met35 (Met sulfoxide; +16 mass units), most probably a technical artifact due to the use of formic acid to process this sample for mass spectrometry. In vitro pre-treatment of synthetic Aβ40 and Aβ42 with formic acid produced similar effect (not shown). Aβ species in vessel extracts were immunoreactive with all the antibodies tested (Panel D). As curiously observed with the PBS-fractions (Panel B), i) Aβ40 immunoreactivity was also limited to dimeric species, and ii) the monomeric component detected by 6E10 and anti-Aβ42 was also composed of two bands of very close molecular mass (as shown in panels B and C). In general terms, the dimeric species were more prominent than in the PBS-fractions. Mass spectrometry identified Aβ1-42 and Aβ4-42 as the major components whereas Aβ1-40 was a minor constituent of the extract. As illustrated above with the PBS-fractions, immunoprecipitation with the anti-Aβ antibodies also retrieved a minor amount of full-length ADan (Panel D). Formic acid retrieved mainly polymerized Aβ species in both, parenchymal and vascular lesions. As illustrated in Figure 5A, parenchymal extracts were immunoreactive with 4G8, 6E10 and anti-Aβ42 antibodies but were not recognized by anti-Aβ40. Aβ species were mainly dimeric although small proportion of higher oligomers were visible with both, 4G8 and anti-Aβ42. When the same sample was analyzed by mass spectrometry, the most prominent signal corresponded to an experimental m/z of 4200.0, consistent with the theoretical protonated mass of 4199.8 for Aβ4-42.

Both, Aβ1-42 and Aβ1-40 were also present in the sample, although Aβ1-40 was a minute contributor to the composition of the FA-extract, data consistent with the lack of immunoreactive signal in the western blots. Two additional peaks differing in +16 and +28 mass units with Aβ4-42 were also observed. Both experimental masses matched the expected values for Aβ4-42 (Met sulfoxide) and Aβ4-42 (formylated). Since samples were exposed to FA during the extraction procedure and prior to the mass spectrometry analysis, these Aβ derivatives may well represent undesirable secondary products of the speciment treatment rather than actual post-translationally modified components present in the lesions. Alternatively, these oxidized Aβ42 species may represent existent in vivo neurotoxic subproducts of oxidative stress, as previously proposed (22). Vascular FA-extracts showed a similar pattern than the parenchymal counterpart although Aβ species seemed to be present in a slightly higher degree of oligomerization. As indicated in Figure 5B, the vessels extracts were immunoreactive with all anti-Aβ antibodies tested; 4G8 and anti-Aβ42 reacted with dimers and larger oligomers whereas anti-Aβ40 was bearily immunoreactive with the dimeric species. Mass spectrometry analysis of the same fractions revealed that Aβ1-42 and its N-terminal truncated derivative Aβ4-42 were the main Aβ species with experimental masses matching their theoretical values whereas minor Aβ4-42 oxidized and formylated derivatives were also detected (Figure 5B and Table). As observed with ADan, the abundance of FA-soluble Aβ species compared with their water-soluble counterparts was reflected in the intensity of the immunoreactivity of both parenchymal and vascular extracts and in the amount of protein required to visualize them (1,500-fold and 5,000-fold more protein for the PBS- and SDS-extracts of parenchyma and vessels, respectively, than for the FA-fractions). Confocal microscopy demonstrates co-localization but not complete overlap between ADan and Aβ staining of parenchymal lesions and blood vessels (Figure 6, panels A-C), suggesting co-deposition on many areas. In order to further test whether the partial co-localization of Aβ and ADan was related at least in part to a specific peptide-peptide interaction we conducted co-immuno-precipitation experiments with tissue extracts and ligand-blot

8

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

analysis with synthetic peptides. As illustrated in Figure 6D for vascular SDS-extracts, a combination of anti-Aβ antibodies 4G8 and 6E10 retrieved a mixture of monomeric and dimeric ADan immunoreactive with Ab 5282 (left panel). However, in the reverse experiment with the same vascular SDS-extracts, anti-ADan antibodies retrieved dimeric Aβ that was reactive with the mixture 4G8 plus 6E10, with anti-Aβ42, but not with anti-Aβ40 antibodies (right panel). Similar results were obtained with PBS-extracts (data not shown). FA-extracts, on the contrary, rendered negative results in spite of the high concentration of both ADan and Aβ, likely indicating that the interaction is susceptible to dissociation at low pH as previously seen in other protein-protein interactions, i.e. Aβ-apolipoprotein J (23). To verify the specificity of the ADan-Aβ interaction, ligand blot analysis using synthetic ADan and Aβ homologues and synthetic ABri as a control was conducted in vitro. As indicated in Figure 7, specific binding of Aβ4-42 and Aβ1-42 (the main Aβ species found in FDD lesions) to ADan oligomers (particularly dimers) was clearly evident whereas no interaction was observed with ADan monomers in spite of their abundance. In contrast to ADan, neither monomers nor oligomers of ABri peptides show binding interaction with either Aβ4-42 or Aβ1-42.

DISCUSSION

Immunohistochemical analysis of FDD cases indicate that vascular amyloid and parenchymal pre-amyloid deposits of ADan and Aβ co-exist in the absence of compact plaques. The ADan/Aβ composition of both lesions were analyzed by western blot, mass spectromery and amino acid sequence analysis after tissue dissection and separation of vessels from parenchymal specimens. Accordingly, water-based solutions in the absence and presence of low concentrations of the anionic detergent SDS were used to extract mainly non-fibrillar components whereas formic acid extraction was necessary to solubilize and analyze fibrillar deposits. Our biochemical data (Figures 2-5) show that FDD non-fibrillar and fibrillar lesions are heterogeneous mixtures of various ADan and Aβ species. The peptide composition, degree of oligomerization, predominant amyloid subunits, extent of post-translational modifications and

magnitude of N- and C-terminal proteolytic degradations reflect the complexity of the lesions. In general terms, non-fibrillar deposits appeared less complex than the fibrillar counterparts, perhaps reflecting early intermediate states of the fibrillization process. The degree of oligomerization was limited only to dimeric forms whereas the diversity of the ADan and Aβ peptides was also highly restricted. ADan species were limited to the non-degraded full-length ADan1-34, although in contrast to what is seen in plasma, ~60% of these molecules were found to be post-translationally modified at their N-terminus. Cyclic pyroglutamate was previously observed in other brain amyloids, i.e. ABri in FBD (6) and in truncated forms of the Alzheimer’s Aβ peptide (24-26) as well as in some hormones and neuropeptides including neurotensin, thyrotropin-, and gonadotropin-releasing hormones in which their biological activities largely depend on the existence of the N-terminal pyroglutamate (27). Although the final pyroglutamate product is the same, neuropeptides and amyloids differ in the amino acid that serves as a substrate for the post-translational modification. In many neuropeptides and peptide hormones the N-terminal post-translational modification occurs at a glutamine residue and the cyclization involves the nucleophilic attack of the α-amino group on the amidated carboxyl group and the release of NH3 catalyzed by glutaminyl cyclase at neutral pH (28,29). In contrast, few examples are known for the post-translational modification from glutamic acid to pyroglutamate, which involve the loss of a molecule of water instead of deamination. Among them are bovine and ovine β-lipotropin and joining peptide derived from the proteolytic processing of their common precursor pro-opiomelanocortin (30,31), ABri amyloid in FBD (6), and the Aβ N-terminal truncated derivatives Aβ3-XpE and Aβ11-XpE in AD (24,25,32). The process has been demonstrated in Aplysia neurons (33) and is not exclusive to the brain, since a similar conversion has been identified in peripheral organs affected with amyloid deposition in FBD (17). Whether the same glutaminyl cyclase is responsible for the dehydration process in all these cases remains known. Since this N-terminal modification was not detected in the soluble circulating ABri or ADan species, it is conceivable

9

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

that cyclization occurs in situ at the sites of deposition. The Aβ peptides found in the non-fibrillar deposits were also restricted in diversity. Aβ1-42 and the N-terminally truncated Aβ4-42 were the major species in an ~1:1 ratio whereas Aβ1-40 and Aβ derivatives ending at position 34 were minor component of the extracts (Figure 8). Although predominant species ending at position 42 were expected in the parenchymal extracts, their abundance in vascular components was surprising. Limited published studies indicate that Aβ-related parenchymal pre-amyloid lesions are enriched in species ending at position 42. The α-secretase fragment Aβ17-42, in particular, was previously identified as the main component of pre-amyloid lesions in AD (34), Down’s syndrome (35) and aged dogs (36), being Aβ1-42 and Aβ4-42 minor components in the last 2 cases. In FDD extracts, Aβ17-42 was not detected and although it could be argued that the present extraction conditions were milder than those previously reported (2% SDS vs. 15% SDS) thereby precluding the retrieval of Aβ17-42, this fragment was also undetected in the subsequent formic acid extracts. Although not through biochemical studies, the existence of other Aβ species were also previously identified in pre-amyloid lesions based on classical immunohisto-chemical stainings and their immunoreactivity with specific antibodies. Accordingly, Aβ1-42 and Aβ3-42pE were reported to be major components in AD, Down’s syndrome and normal aging pre-amyloid lesions whereas antibodies recognizing position 17 were only weakly positive (37). Anti-Aβ42 immunoreactivity was also found prevalent in parenchymal pre-amyloid deposits in the Dutch variant of familial AD (38). ADan and Aβ species extracted with formic acid from amyloid deposits were more heterogeneous. ADan was fully post-translationally modified at the N-terminus (pyroglutamate), partially N- and C- terminally degraded at positions 3, 28 and 33, and heavily oligomerized (Figures 3 and 8), whereas Aβ, still mainly composed of Aβ1-42 and Aβ4-42 (with negligible Aβ1-40), showed more N-terminal degradation (ratio Aβ4-42:Aβ1-42 approximately 9:1) and higher degree of oligomerization (Figures 5 and 8). Whether extensive proteolytic degradation and heavier oligomerization reflect critical necessary steps in the process of ADan amyloid formation or simply

a futile cellular effort to end and clear the formation of fibrillar deposits is still under investigation. In the case of Aβ, the N-terminal degraded Aβ4-42 is known to have a faster aggregation kinetics than the intact Aβ1-42 (39). The fact that the ADan proteolytic fragments were not detected in blood and that circulating soluble ADan peptides differ from the deposited homologues in that they consistently lack the chemically irreversible pyroglutamate modification at the N-terminus is a clear indication that the circulating species do not represent a clearance mechanism from the cerebral deposits. In this sense, we previously reported similar findings for the soluble and deposited ABri species in FBD (17). As summarized in Figure 8, the proteolytic degradation of ADan and Aβ reflects the activity of more than a single enzyme. A potential candidate for the C-terminal de-tyrosination of ADan is cathepsin A (carboxypeptidase A) (40) which is widely distributed in lysosomes and has broad specificity, releasing the C-terminal amino acid at optimum pH of 4.5-6.0. The cleavage at ADan Ala2-Ser3 peptide bond seems to be catalyzed by a dipeptidyl-peptidase, in a similar fashion than deposited ABri amyloid in FBD (6). It is interesting to note that no proteolytic processing of ADan molecules was detected between cysteine residues 5 and 22, suggesting a protective effect of the disulphide bond as reported for other proteins (41-43). Degradation of Aβ at peptide-bond Glu3-Phe4 may result from either neprilysin or insulysin (insulin-degrading enzyme) degradation, as previously reported in Aβ-related disorders (44,45). However, these enzymes are not exclusive for the Glu3-Phe4 peptide bond processing and are known to cleave Aβ at several other sites, generating degradation fragments that were not detected in FDD extracts. Another probable enzyme to consider is tripeptidyl-peptidase I, a lysosomal protease known to cleave the Glu3-Phe4 peptide bond in vitro and a known component of amyloid plaques (46,47). It is important to note that the same proteolytic activity also results from the use of bacterial collagenase I (EC 3.4.24.3; http://us.expasy.org/enzyme). Using synthetic Aβ40 and Aβ42 as substrates, this enzyme was able to efficiently cleave both peptides at Glu3-Phe4 and Leu34-Met35 bonds in in vitro

10

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

experiments (unpublished observations). Although collagenase I has been commonly used in amyloid purification protocols to increase the yield of the extracted fibrils (48-52), we purpously avoided its use in our FDD isolation protocol in order to prevent anwanted artifacts. In sporadic and familial AD as well as in other Aβ-related disorders, amyloid deposits in cerebral vessels are primarily constituted of fibrillar forms of Aβ40 although the presence of Aβ42 in detectable amounts is considered a common finding in these lesions [reviewed in (53)]. Similarly, in the various transgenic models of Aβ deposition, Aβ40 is the main component of vascular deposits (54-56) including the recently published APP-Dutch mouse (57). The fact that the main Aβ components of the vascular FDD lesions were AβX-42 species was somehow an unexpected finding of still unknown connotations for the disease pathogenesis. The importance of Aβ42 for the formation of vascular deposits is also highlighted by the recent report on the chimeric BRI-Aβ42 mice, which selectively overproduce Aβ42 but not Aβ40, and the demonstration of extensive Aβ42 cerebral amyloid angiopathy whereas the BRI-Aβ40 chimeric mice overproducing Aβ40 did not show overt amyloid pathology, questioning the importance of Aβ40 in the initiation of vascular amyloid deposits. (58). It is also not clear whether the co-localization of ADan with Aβ reflects important aspects of the mechanism of amyloidogenesis or simply a conformational mimicry, by which one molecule induces another to adopt a similar structural conformation favoring oligomerization. In this sense, dimers (not monomers) of Aβ molecules were retrieved in co-immunoprecipitation experiments from tissue extracts and Aβ1-42 and Aβ4-42 bound to ADan dimers and higher oligomers (not monomers) in ligand blot analysis,

suggesting conformational specificity. It is important to note that Aβ was neither detected in parenchymal or vascular ABri deposits in FBD (15,16) nor identified as an Aβ ligand in the ligand-blot analysis shown in Figure 7 in spite of being ABri and ADan molecules identical in their 22 N-terminal residues. Mounting evidence indicate that plaque burden correlates poorly with degree of dementia and that soluble oligomers rather than the final amyloid fibrils are the harmful Aβ species (59-62). This viewpoint is supported by the two hereditary disorders FBD and FDD (collectively known as Chromosome 13 Dementias) and by the Iowa variant of familial AD associated with the deposition of the mutant AβD23N (63). In spite of the structural differences among ABri, ADan and Aβ, absence or limited number of compact plaques and extensive neurofibrillar degeneration with tangles identical to those found in Alzheimer’s brains are main features of these disorders. Therefore, not only different amyloid peptides are associated with similar neuropathological changes but these disorders question the importance of compact plaques in the mechanism of neuronal toxicity. Compact plaques are absent in FDD, in the Iowa variant of familial AD, and in certain brain areas in FBD whereas all disorders feature parenchymal pre-amyloid lesions as well as vascular and perivascular deposits and yet, the final clinical outcome of these diseases is AD-like dementia. Thus, if amyloid is of paramount importance in the development of dementia, it is certainly neither exclusive for Aβ nor dependent on the presence of compact plaques. We propose that these disorders are suitable models to study early steps in peptide oligomerization/fibrillization and the role of pre-amyloid and vascular amyloid in the process of neurodegeneration.

REFERENCES 1. Strömgren, E., Dalby, A., Dalby, M. A., and Ranheim, B. (1970) Acta Neurol. Scand. 46, 261-

262 2. Strömgren, E. (1981) Heredopathia ophthalmo-oto-encephalica. Handbook of Clinical

Neurology (Vinken, P. J., and Bruyn, G. W., Eds.), 42, North Holland Publishing Co., Amsterdam

3. Bek, T. (2000) British J. Ophtalmol. 84, 1298-1302

11

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

4. Vidal, R., Ghiso, J., Revesz, T., Rostagno, A., Kim, E., Holton, J., Bek, T., Bojsen-Moller, M., Braendgaard, H., Plant, G., and Frangione, B. (2000) Proc. Natl. Acad. Sci. USA 97, 4920-4925

5. Holton, J., Lashley, T., Ghiso, J., Braendgaard, H., Vidal, R., Guerin, C., Gibb, G., Hanger, D. P., Rostagno, A., Anderton, B., Strand, C., Ayling, H., Plant, G., Frangione, B., Bojsen-Moller, M., and Revesz, T. (2002) J. Neuropathol. Exp. Neurol. 61, 254-267

6. Vidal, R., Frangione, B., Rostagno, A., Mead, S., Revesz, T., Plant, G., and Ghiso, J. (1999) Nature 399, 776-781

7. Pittois, K., Deleersnijder, W., and Merregaert, J. (1998) Gene 217, 141-149 8. Rostagno, A., Tomidokoro, Y., Lashley, T., Ng, D., Plant, G., Holton, J., Frangione, B., Revesz,

T., Ghiso, J. (2005) Cell Mol Life Sci., [Epub ahead of print] 9. Kim, S.-H., Wang, R., Gordon, D. J., Bass, J., Steiner, D., Lynn, D. G., Thinakaran, G., Meredith,

S., and Sisodia, S. S. (1999) Nature Neurosc. 2, 984-988 10. Worster-Drought, C., Hill, T., and McMenemey, W. (1933) J. Neurol. Psychopathol 14, 27-34 11. Worster-Drought, C., Greenfield, J. G., and McMenemey, W. (1940) Brain 63, 237-254 12. Plant, G., Revesz, T., Barnard, R., Harding, A., and Gautier-Smith, P. (1990) Brain 113, 721-747 13. Mead, S., James Galton, M., Revesz, T., Doshi, R. B., Harwood, G., Pan, E. L., Ghiso, J.,

Frangione, B., and Plant, G. (2000) Brain 123, 975 986 14. Holton, J., Ghiso, J., Lashley, T., Rostagno, A., Guerin, C., Gibb, G., Houlden, H., Ayling, H.,

Martinian, L., Anderton, B., Wood, N., Vidal, R., Plant, G., Frangione, B., and Revesz, T. (2001) Am. J. Pathology 158, 515-526

15. Revesz, T., Holton, J. L., Doshi, B., Anderton, B. H., Scaravilli, F., and Plant, G. (1999) Acta Neuropathol. 97, 170-176

16. Ghiso, J., Plant, G., Revesz, T., Wisniewski, T., and Frangione, B. (1995) J. Neurol. Sci. 129, 74-75

17. Ghiso, J., Holton, J., Miravalle, L., Calero, M., Lashley, T., Vidal, R., Houlden, H., Wood, N., Neubert, T. A., Rostagno, A., Plant, G., Revesz, T., and Frangione, B. (2001) J. Biol. Chem. 276, 43909-43914

18. Woods, A. S., Buchsbaum, J. C., Worrall, T. S., Berg, J. M., and Cotter, R. J. (1995) Anal. Chem. 67, 4462-4465

19. Ida, N., Hartmann, T., Pantel, J., Schroder, J., Zerfass, R., Forstl, H., Sandbrink, R., Masters, C. L., and Beyreuther, K. (1996) J. Biol. Chem. 271, 22908-22914

20. Hixson, J. E., and Vernier, D. T. (1990) J. Lipid Res. 31, 545-548 21. Duewel, H. S., and Honek, J. F. (1998) J. Protein Chem. 17, 337-350 22. Butterfield, D. A. (2002) Free Radical Res. 36, 1307-1313 23. Matsubara, E., Frangione, B., and Ghiso, J. (1995) J. Biol. Chem. 270, 7563-7567 24. Mori, H., Takio, k., Ogawara, M., and Selkoe, D. (1992) J. Biol. Chem. 267, 17082-17086 25. Saido, T., Yamao-Harigaya, W., Iwatsubo, T., and Kawashima, S. (1996) Neuroscience Letters

13, 173-176 26. Tekirian, T., Saido, T., Markesbery, W., Russell, M., Wekstein, D., Patel, E., and Geddes, J.

(1998) J. Neuropathol. Exp. Neurol. 57, 76-94 27. Sykes, P. A., Watson, S. J., Temple, J. S., and Bateman, R. C. J. (1999) FEBS Lett. 455, 159-161 28. Busby, W. H., Quackenbush, G. E., Humm, J., Youngblood, W. W., and Kizer, J. S. (1987) J.

Biol. Chem. 262, 8532-8536 29. Fischer, W. H., and Spiess, J. (1987) Proc Natl Acad Sci U S A 84, 3628-3632 30. Bateman, A., Solomon, S., and Bennett, H. P. J. (1990) J. Biol. Chem. 265, 22130-22136 31. Yamashiro, D., and Li, C. H. (1976) Biochim. Byophys. Acta 451, 124-132 32. Saido, T., Iwatsubo, T., Mann, D. M., Shimada, H., Ihara, Y., and Kawashima, S. (1995) Neuron

14, 457-466 33. Garden, R. W., Moroz, T. P., Gleeson, J. M., Floyd, P. D., Li, L., Rubakhin, S. S., and Sweedler,

J. V. (1999) J. Neurochem. 72, 676-681

12

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

34. Gowing, E., Roher, A. E., Woods, A. S., Cotter, R. J., Chaney, M., Little, S. P., and Ball, M. J. (1994) J. Biol. Chem. 269, 10987-10990

35. Lalowski, M., Golabek, A., Lemere, C. A., Selkoe, D. J., Wisniewski, H. M., Beavis, R. C., Frangione, B., and Wisniewski, T. (1996) J. Biol. Chem. 271, 33623-33631

36. Wisniewski, T., Lalowski, M., Bobik, M., Russell, M., Strosznajder, J., and Frangione, B. (1996) Biochem J. 313, 575-580

37. Iwatsubo, T., Saido, T., Mann, D. M., Lee, V., and Trojanowski, J. Q. (1996) Am. J. Pathol. 149, 1823-1830

38. Castaño, E. M., Prelli, F., Soto, C., Beavis, R., Matsubara, E., Shoji, M., and Frangione, B. (1996) J. Biol. Chem. 271, 32185-32191

39. Pike, C. J., Overman, M. J., and Cotman, C. W. (1995) J. Biol. Chem. 270, 23895-23898 40. Hiraiwa, M. (1999) Cell. Molec. Life Sci. 56, 894-907 41. Wallis, T. P., Pitt, J. J., and Gorman, J. J. (2001) Protein Sci. 10, 22511-22271 42. van der Burg, H. A., Westerink, N., Francoijs, K. J., Roth, R., Woestenenk, E., Boeren, S., de Wit,

P. J. G. M., Joosten, M. H. A. J., and Vervoort, J. (2003) J. Biol. Chem. 278, 27340-27346 43. Novokhatny, V. V., and Ingham, K. C. (1994) J. Mol. Biol. 238, 833-844 44. Howell, S., Nalbantoglu, J., and Crine, P. (1995) Peptides 16, 647-652 45. Morelli, L., Llovera, R., Gonzalez, S. A., Affranchino, J. L., Prelli, F., Frangione, B., Ghiso, J.,

and Castaño, E. M. (2003) J. Biol. Chem. 278, 23221-23226 46. Junaid, M. A., Wu, G., and Pullarkat, R. K. (2000) J. Neurochem. 74, 287-294 47. Liao, L., Cheng, D., Wang, J., Duong, D. M., Losik, T. G., Gearing, M., Rees, H. D., Lah, J. J.,

Levey, A. I., and Peng, J. (2004) J. Biol. Chem. 279, 37061-37068 48. Cohen, A. S., and Calkins, E. (1964) J. Cell Biol. 21, 481-486 49. Glenner, G. G., and Wong, C. W. (1984) Biochem. Biophys. Res. Comm. 120, 885-890 50. Glenner, G. G., and Wong, C. W. (1984) Biochem. Biophys. Res. Comm. 122, 1131-1135 51. Prelli, F., Castaño, E. M., Glenner, G. G., and Frangione, B. (1988) J. Neurochem. 51, 648-651 52. Miller, D. L., Papayannopoulos, I. A., Styles, J., Bobin, S. A., Lin, Y. Y., Biemann, K., and Iqbal,

K. (1993) Arch. Biochem. Biophys. 301, 41-52 53. Ghiso, J., and Frangione, B. (2002) Adv. Drug Delivery Rev. 54, 1539-1551 54. Van Dorpe, J., Smeijers, L., Dewachter, I., Nuyens, D., Spittaels, K., Van der Haute, C., Mercken,

M., Moechars, D., Laenen, I., Kuiperi, C., Bruynseels, K., Tesseur, I., Loos, R., Vanderstichele, H., Checler, F., Sciot, R., and Van Leuven, F. (2000) Am. J. Pathol. 157, 1283-1298

55. Kuo, Y.-M., Kokjohn, T. A., Beach, T., Sue, L., Brune, D., Lopez, J., Kalback, W., Abramowski, D., Sturchler-Pierrat, C., Staufenbiel, M., and Roher, A. (2001) J Biol Chem 276, 12991-12998

56. Fryer, J. D., Taylor, W., DeMattos, R. B., Bales, K. R., Paul, S. M., Parsadanian, M., and Holtzman, D. M. (2003) J. Neurosci. 23, 7889-7896

57. Herzig, M. C., Winkler, D. T., Burgermeister, P., Pfeifer, M., Kohler, E., Schmidt, S. D., Danner, S., Abramowski, D., Stürchler-Pierrat, C., Bürki, K., van Duinen, S. G., Maat-Schieman, M. L. C., Staufenbiel, M., Mathews, P. M., and Jucker, M. (2004) Nature Neurosc. 7, 954-960

58. McGowan, E., Pickford, F., Kim, J., Onstead, L., Eriksen, J., Yu, C., Skipper, L., Murphy, M.P., Beard, J., Das, P., Jansen, K., DeLucia, M., Lin, W.-L., Dolios, G., Wang, R., Eckman, C.B., Dickson, D.W., Hutton, M., Hardy, J., Golde, T. (2005) Neuron 47, 191-199

59. Klein, W. L., Krafft, G. A., and Finch, C. (2001) Trends Neurosci. 24, 219-224 60. Lansbury, P. T. (1999) Proc. Natl. Acad. Sci. USA 96, 3342-3344 61. Small, D. H. (1998) Amyloid: Intl. J. Exp. Clin. Invest. 5, 301-304 62. Haass, C., and Steiner, H. (2001) Nat. Neurosci. 4, 859-860 63. Grabowski, T. J., Cho, H. S., Vonsattel, J. P. G., Rebeck, G. W., and Greenberg, S. M. (2001)

Ann. Neurol. 49, 697-705

13

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

FOOTNOTES * This work was supported in part by the National Institutes of Health Grants AG05891, AG08721, S10RR14662 and NS38777, by the Alzheimer Association and by the Alzheimer’s Disease Research program of the American Health Assistance Foundation. We thank Ms. Yun Lu of the NYU Protein Analysis Facility for MALDI-TOF mass spectrometry and Ms. Zhihong Zhao for the ApoE genotyping. The abbreviations used are: FDD, familial Danish Dementia; ADan, Danish-amyloid; ABri, British-amyloid; AD, Alzheimer’s disease; FBD, familial British dementia; Aβ, β-amyloid; MALDI-TOF, Matrix-Assisted Laser Desorption Ionization-Time-of-Flight; PBS, phosphate buffered saline; SDS, sodium dodecyl sulphate; FA, formic acid.

FIGURE LEGENDS

Figure 1. Immunohistochemical analysis of FDD lesions. Panel A: ADan deposited in rather ill-defined parenchymal plaques (double arrow) and blood vessels (arrow) in the hippocampus in an FDD case was detected with antibody 5282. Lesions in the frontal cortex are similar, although with less dramatic pre-amyloid load. Bar represents 120 µm. Panel B: Tau immunohistochemistry (AT8 antibody) reveals numerous neurofibrillary tangles (arrow), neuropil threads (arrowhead) and abnormal neurites (double arrow) formed around amyloid-laden blood vessels (asterisk). Bar represents 60 µm. Figure 2: Western blot and mass spectrometry analysis of plasma and non-fibrillar ADan deposits. ADan molecules were immunoprecipitated using Dynabeads coated with anti-ADan Ab5282 and subsequently eluted in either Tris-Tricine sample buffer containing dithiothreitol for the western blot analysis or water/isopropyl alcohol/formic acid (4:4:1) for mass spectrometry via MALDI-TOF, as described in Experimental Procedures. Panel A: soluble ADan immunoprecipitated from 0.5 ml of plasma. Panel B: PBS-extracted fractions from 500 mg of vessel-depleted frontal cortex and 50 mg of isolated microvessels and leptomeningeal vessels. Panel C: SDS-extracted fractions from 500 mg of vessel-depleted frontal cortex and 50 mg of isolated microvessels and leptomeningeal vessels. Theoretical average masses and experimental ADan molecular masses bearing a single disulphide-bond between cysteine residues 5 and 22 are shown in the bottom table. The y-axis of the mass spectra show relative peak intensity based on the maximum ion counts, which are displayed to the right of each spectrum. Notation for the western blot: Mo, monomers; Di, dimers. Figure 3: Western blot and mass spectrometry analysis of fibrillar ADan ADan molecules were extracted in 70% FA, lyophilized, re-dissolved in Tris-Tricine sample buffer containing dithiothreitol and directly analyzed by western blot. For MALDI-TOF analysis, ADan molecules were desalted by ZipTip micro-reverse phase column, eluted with 90% (v/v) acetonitrile-0.1% trifluoroacetic acid in water, as described in Experimental Procedures. Panel A: parenchymal FA-extracts from microvessel-depleated frontal cortex after PBS and SDS extractions. Western blots represent 1.5 mg of frontal cortex whereas mass spectrometry spectra represent 22.5 mg of frontal cortex. Panel B: vessel extracts in FA-isolated microvessels and leptomeningeal vessels after PBS and SDS extractions. Western blots represent 0.5 mg of frontal cortex whereas mass spectrometry spectra represent 22.5 mg of frontal cortex. Peaks corresponding in mass to formylated species are indicated by asterisks (*, single

14

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

formylation; **, double formylation; ***, triple formylation). Theoretical average masses and experimental ADan molecular masses bearing a single disulphide-bond between cysteine residues 5 and 22, are shown in the bottom table. Notation for the western blots: arrowheads indicate electrophoretic mobility of ADan monomers, dimers and trimers. Figure 4: Western blot and mass spectrometry analysis of non-fibrillar Aβ Aβ species were immunoprecipitated with Dynabeads coated with a mixture of 4G8 and 6E10. Bound peptides were eluted and analyzed either by western blot using a panel of anti-Aβ antibodies (4G8, 6E10, anti-Aβ40, anti-Aβ42) or by MALDI-TOF mass spectrometry as described in Experimental Procedures. PBS-extracted fractions from 500 mg of microvessel-depleted frontal cortex (Panel A) and 50 mg of isolated microvessels and leptomeningeal vessels equivalent to 500 mg of frontal cortex (Panel B). SDS-extracted fractions from 500 mg of microvessel-depleted frontal cortex (Panel C) and 50 mg of isolated microvessels and leptomeningeal vessels equivalent to 500 mg of frontal cortex (Panel D). Theoretical average masses and experimental Aβ molecular masses are shown in the bottom table. Notation for the western blot: Mo, monomers; Di, dimers. Peaks of mass spectra indicated by # were present in the negative controls (without brain extracts) and considered to be non-specific. Peaks notated by ¤ were also obtained in negative controls (using brain extracts but uncoated Dynabeads) and considered non-specific. Figure 5: Western blot and mass spectrometry analysis of fibrillar Aβ Aβ species were extracted in 70% FA, lyophilized, re-dissolved in Tris-Tricine sample buffer containing dithiothreitol and directly analyzed by western blot against 4G8, 6E10, anti-Aβ40 and anti-Aβ42. For MALDI-TOF analysis, Aβ molecules were immunoprecipitated with Dynabeads coated with a mixture of 4G8 and 6E10 antibodies, eluted with a mixture of water/isopropyl alcohol/formic acid (4:4:1) and analyzed by mass spectrometry as described in Experimental Procedures. Panel A: parenchymal FA-extracts from microvessel-depleated frontal cortex after PBS and SDS extractions. Western blots represent 1.5 mg of frontal cortex whereas mass spectrometry spectra represent 15 mg of frontal cortex. Panel B: vessel extracts in FA-isolated microvessels and leptomeningeal vessels after PBS and SDS extractions. Western blots represent 0.5 mg of frontal cortex whereas mass spectra represent 1.5 mg of vessels (equivalent to 15 mg of frontal cortex). Formylated species are indicated by asterisk; Ox represents Aβ species oxidized at Met35. Theoretical average masses and experimental Aβ molecular masses are shown in the bottom table. Notation for the western blots: arrowheads indicate electrophoretic mobility of Aβ monomers and dimers. Figure 6: Co-localization of ADan and Aβ might be related to a specific interaction between deposited peptides. Panels A-C: Confocal images of neocortical blood vessels (arrow) show deposition of both ADan (red) and Aβ (green), although the co-localization is not complete. Please note that in this particular area the parenchymal protein deposit is primarily ADan (double arrow). Confocal microscopy; objective x 20. Panel D: Pull-down experiments of ADan and Aβ in SDS-soluble fractions from isolated vessels in FDD. Presence of ADan was detected with Ab5282 after immunoprecipitation with a mixture of anti-Aβ antibodies 4G8 and 6E10 (left panel). Conversely, AβX-42 is detectable by the mixture 4G8/6E10 and by the C-terminus specific anti-Aβ42 antibody (right panel) after amyloid species were immunoprecipitated by Ab5282-coated paramagnetic beads. Notation for the western blot: Mo, monomers; Di, dimers. Figure 7: Ligand blot analysis Synthetic Aβ1-42 and Aβ4-42 peptides were used as ligands against synthetic ADan peptides resolved in 16% Tris-Tricine SDS-PAGE. Synthetic ABri peptides were used as controls. Bound Aβ molecules were detected by a mixture of 4G8/6E10 anti-Aβ antibodies whereas oligomerization of ADan and ABri

15

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

peptides was assessed with a mixture of anti-ADan and anti-ABri antibodies Ab5282/Ab338, respectively. Notation for the western blot: Mo, monomers; Di, dimers; Tri, trimers. Figure 8: ADan and Aβ species in cerebral cortex of FDD brain The figure summarizes the ADan and Aβ major and minor components of fibrillar and non-fibrillar deposits in FDD. Major proteolytic processing sites are indicated in large arrows whereas small arrows with discontinuous lines indicate minor cleavage sites. pE: pyroglutamate.

16

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

Figure 1

A B

*

*

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

ADan1-34pE ADan1-34E(-) 4063.4

Parenchyma 4045.2 4063.5Vessels 4045.3 4062.7

Parenchyma 4043.6 4062.0Vessels 4045.6 4063.6

4045.6 4063.6

Observed m/z

Expected Mass [M+H]

Bra

in

PlasmaADan species

SDS

PBS

B

C Di

Mo

Figure 2

PBS

SDS

Mo

APlasma

Bra

in E

xtra

cts

Mo

Di

Parenchyma

Vessels

Parenchyma

Vessels

max. ion count: 2850

100

rela

tive

inte

nsity

0m/z41004000

100

m/z

100

rela

tive

inte

nsity

0m/z41004000

100

rela

tive

inte

nsity

0m/z41004000

4043.6

3919.14062.0

4174.4

ADan1-34E

ADan1-34pE

0 0 . 4

3 9 6 7 . 5

4 0 6 3 . 6

4 1 4 1

100

rela

tive

inte

nsity

0m/z41004000

ADan1-34E

ADan1-34pE

ADan1-34pE

rela

tive

inte

nsity

041004000

901.1

4065.7

4

4063.5

4180

ADan1-34pE

ADan1-34EADan1-34E

max. ion count: 226 max. ion count: 1970

max. ion count: 2550max. ion count: 907

Parenchyma

Parenchyma

Vessels

Vessels

4046.20

3997.86

3955.00

4034.21

4143.36

4118.69

ADan1-34E

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

3 3 2 9 . 8 1

3 1 1 9 . 7 2

3 1 0 2 . 1 13 0 0 4 . 9 3

2 8 9 1 . 5 1

3 2 1 5 . 3 0

3 9 3 9 . 7 1

3 7 5 7 . 7 2

3 5 7 3 . 4 2

3 4 4 5 . 1 3

3 3 6 3 . 2 1 3 5 4 5 . 5 4 3 7 0 2 . 3 6

3 6 7 4 . 2 3

3 6 2 0 . 4 7

3 9 2 0 . 9 0

3 9 1 1 . 5 7

3 8 3 0 . 1 3

4 1 3 0 . 6 3

4 1 0 2 . 6 5

4 0 7 4 . 6 8

13 9 3 9 . 3 2

3 9 1 1 . 9 63 3 2 9 . 6 9

3 3 0 1 . 8 6

3 1 2 0 . 2 1

3 0 9 1 . 6 12 8 1 1 . 4 2

3 0 7 3 . 9 92 9 5 8 . 7 02 8 2 0 . 4 5

3 2 7 2 . 8 63 2 1 5 . 5 4

3 9 0 4 . 2 1

3 7 5 7 . 4 9

3 7 2 9 . 7 6

3 4 1 6 . 3 6

3 5 4 6 . 0 2

3 4 4 4 . 8 2

3 7 0 1 . 6 3

3 5 7 3 . 2 2

3 6 7 5 . 1 0

3 6 2 8 . 7 9

3 8 9 1 . 7 1

3 7 8 4 . 6 6

3 8 6 5 . 7 2

3 9 6 7 . 2 8

4 1 0 2 . 6 7

4 0 7 4 . 1 64 1 2 9 . 5 8

4 1 5 6 . 5 8

4 1 8 5 . 1 8

4 2 3 0 . 2 14 2 5 5 . 6 1

4 3 7 0 . 1 74 4 0 1 . 5 0 4 5 1 8 . 3 1

Parenchyma

Figure 3

3-28 1-27pE 1-28pE 1-30pE 3-333091.5 3158.8 3272.9 3489.2 3700.73091.3 3159.1 3272.7 3489.3 3701.33090.6 3158.7 3272.8 3488.0 3700.2

Observed m/z

Expected Mass [M+H]

ADan speciesParenchyma

Vessels

FA

4530

20.114.3

6.53.5

Vessels

4530

20.114.3

6.53.5

3000 3200 3400 3600 3800 4000 4200 4400

100

0

m/z

1-33

pE** 1-

33pE

***

3-28

*

1-34

pE**

*1-

34pE

**

3-34

**

3-33

**

1-28

pE**

1-28

pE*

1-30

pE**

*

100re

lativ

e in

tens

ity

03-

28* 1-

28pE

**

1-34

pE**

m/z

1-34

pE*

3-33

**

3000 3200 3400 3600 3800 4000 4200 44001-

28pE

*

3-33

*

1-33

pE*

1-30

pE**

1-34

pE**

*

1-33

pE**

*

* :mono-formylated, ** :di-formylated and *** :tri-formylated species

3-34 1-33pE 1-33E 1-34pE 1-34E3863.7 3883.3 3903.2 4045.6 4063.63864.8 3883.4 3900.4 4046.5 4063.73863.4 3882.4 3900.4 4045.6 4063.6

ADan speciesParenchyma

VesselsExpected Mass [M+H]

1-34

pE*

1-33

pE**

3-28

**

3-34

*

1-27

pE**

3-33

***

3-34

*

1-27

pE**

max. ion count: 1840

max. ion count: 39200

rela

tive

inte

nsity

B

A

3-33

***

3-28

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

3884.68

08.28

3697.93817.053429.66

2897.38 2995.423323.90

3108.90 3272.44

3471.77

3644.47

3854.67

3840.98

3787.34

4216.30

4014.65

3999.82

3909.64

4028.64

4200.204047.92

4799.174247.69

4396.34 4478.084697.52

4616.60

4936.73

Aβ species Aβ 4-34 Aβ 4-42 Aβ 1-40 Aβ 1-42Parenchyma 3473.5 (-) (-) (-)

Vessels (-) 4199.4 4329.3 4513.9Parenchyma (-) 4200.0 4331.1 4514.9

Vessels (-) 4198.1 4331.1 4515.63472.9 4199.8 4330.9 4515.1

Observed m/z

Expected Mass [M+H]

PBS

SDS

SDS

4200.0

3886.1

8.4 3787.83336.1

4468.7

4615.6

Aβ1

-42

Mo

Di

Parenchyma Vessels

4G86E10anti-42anti-40

Aβ4

-34

¤

4G86E10anti-42anti-40

4G86E10anti-42

anti-40

Mo

Di

4214.1

3883.4

3700.93443.83371.34065.9

4529.5

4344.9

3701.3

3273.33092.0

6.6

4529.9

4045.34215.4

4345.3

4711.7

#

AD

an1-

34pE

AD

an1-

34pE

A

C

B

D

Mo

Di

4G86E10anti-42anti-40

PBS

Parenchyma Vessels

100

0

rela

tive

inte

nsity

3000 3500 4000 4500 5000

¤

¤

Aβ4

-42

Aβ1

-40

Aβ1

-42

max. ion count: 6320

Aβ4

-42

Aβ1

-40

100

0

rela

tive

inte

nsity

3000 3500 4000 4500 5000

¤¤¤

## Aβ1

-40

Aβ1

-42

100

0

rela

tive

inte

nsity

3000 3500 4000 4500 5000

max. ion count: 6890max. ion count: 8620

¤

¤

max. ion count: 504

#

¤

Mo

DiAβ4

-42

Figure 4

m/z m/z

m/z m/z

¤¤ ¤

¤

¤

100

0

rela

tive

inte

nsity

3000 3500 4000 4500 5000

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

B VesselsAFA

Parenchyma

4530

20.114.3

6.53.5

Aβ species 4-42 4-42 Ox 4-42 * 1-40 1-42Parenchyma 4200.0 4215.9 4228.1 4330.0 4514.9

Vessels 4200.2 4216.3 4228.4 4330.3 4515.9Expected Mass [M+H] 4199.8 4215.8 4227.8 4330.9 4515.1

Observed m/z

4600

100

rela

tive

inte

nsity

04200

m/z4400

Aβ4

-42*

Aβ1

-40

Aβ1

-42

4G86E10

anti-42anti-40

4G86E10

anti-42anti-40

m/z

100

rela

tive

inte

nsity

04200 4400 4600

m/z

4530.8Aβ4

-42

Max. ion count: 4360 Max. ion count: 610Aβ4

-42*

Aβ4

-42

Ox

Aβ1

-42

¤¤

¤

Aβ4

-42

Ox

Figure 5

Aβ4

-42

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

Figure 6

a-ADan a-Aβ combined

ADan Aβ Combined

A CB

Mo Mo

Di

4G8 + 6E10anti-ADanAb5282

IP: anti-Aβ IP: anti-ADan

+ +--

a-Aβ42 a-Aβ40

+ +

D

Detection:

by guest on December 11, 2020 http://www.jbc.org/ Downloaded from

Figure 7

ABriABri

ABriADan

ADanADan

molecules onthe membrane:

Di

Mo

Tri

(-)ligands: Aβ1-42 Aβ4-42

Abs: 4G8 + 6E10 Ab5282 + Ab338

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

pEASNCFAIRHFENKFAVETLICFNLFLNSQEKHY

1 34DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA

424

ADan:

Aβ :

Figure 8

1 3 28 33 34

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

Revesz and Jorge GhisoBojsen-Moller, Hans Braendgaard, Gordon Plant, Janice Holton, Blas Frangione, Tamas Yasushi Tomidokoro, Tammaryn Lashley, Agueda Rostagno, Thomas Neubert, Marie

) in the absence of compact plaquesβ(ADan and AFamilial Danish dementia: Co-existence of Danish and Alzheimer amyloid subunits

published online August 9, 2005J. Biol. Chem. 

  10.1074/jbc.M504038200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from