| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 44, 36883-36894, November 4, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
) IN THE ABSENCE OF COMPACT PLAQUES*
¶¶¶¶1
From the
Departments of Pathology, ¶¶Psychiatry, and ¶Pharmacology and the ||Skirball Institute for Biomolecular Medicine, New York University School of Medicine, New York, New York 10016, the Department of Neuropathology and Brain Bank, Institute of Neurology, London WC1N3BG, United Kingdom, the Departments of **Neuropathology and Neurology, Århus University Hospital, Århus DK8000, Denmark, and The National Hospital for Neurology and Neurosurgery, London WC1N3BG, United Kingdom
Received for publication, April 13, 2005 , and in revised form, August 5, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESDISCUSSIONREFERENCES |
|---|
-amyloid (A)), the absence of compact plaques, and neurofibrillary degeneration indistinguishable from that observed in Alzheimer disease (AD) are the main neuropathological features of the disease. Biochemical analysis of extracted amyloid and preamyloid species indicates that as the solubility of the deposits decreases, the heterogeneity and complexity of the extracted peptides exponentially increase. Nonfibrillar deposits were mainly composed of intact ADan-(1-34) and its N-terminally modified (pyroglutamate) counterpart together with A-(1-42) and A-(4-42) 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 and highly heterogeneous at the N and C terminus, and, when intact, its N terminus was post-translationally modified (pyroglutamate), whereas A was mainly A-(4-42). In all cases, the presence of A-(X-40) was negligible, a surprising finding in view of the prevalence of A40 in vascular deposits observed in sporadic and familial AD, Down 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 A molecules was demonstrated in vitro by ligand blot analysis using synthetic peptides. The absence of compact plaques in the presence of extensive neuro fibrillar degeneration strongly suggests that compact plaques, fundamental lesions for the diagnosis of AD, are not essential for the mechanism of dementia. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESDISCUSSIONREFERENCES |
|---|
Neuropathologically, FDD shares several features with Alzheimer disease (AD), among them widespread cerebrovascular amyloidosis, parenchymal amyloid and preamyloid lesions, and neurofibrillary degeneration (4, 5). Vascular and perivascular amyloid, parenchymal preamyloid lesions, and 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 filament 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, a detailed anti-A immunohistochemical survey of different brain areas from all available FDD autopsy cases (three family members from different generations, including the case analyzed in the present study (5)) unequivocally identified A co-deposited with ADan mainly in vascular and perivascular amyloid lesions, although on a smaller scale the co-deposition was also observed in parenchymal preamyloid deposits (5). The fact that A co-deposition was a constant feature of all three FDD cases available argues against a coincidental nonspecific 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-residue 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-residue 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 frameshift 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 acid-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 10-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. (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 ultrastructurally 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, preamyloid 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 despite the similarities between FDD and FBD, detailed immunohistochemical studies never found A co-localized with ABri deposits (15, 16).
We describe herein the isolation and biochemical characterization of ADan and A species deposited in both vessels and parenchymal FDD lesions. Nonfibrillar preamyloid 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 A species in both amyloid and preamyloid deposits. Surprisingly, and against the dogma that A peptides ending at residue 40 are always the main components of cerebrovascular lesions in A-related cerebral amyloidosis, A-(X-42) species were the predominant A components in FDD deposits. Of note, a specific interaction between synthetic ADan and A-(X-42) peptides was demonstrated by ligand blot analysis, pointing to this peptide-peptide interaction as the potential mechanism for their co-localization in vivo.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESDISCUSSIONREFERENCES |
|---|
-(17-24)) and 6E10 (anti-A-(1-17)) were purchased from Signet Laboratories (Dedham, MA); mouse monoclonal 6F/3D (anti-A-(8-17)) was from Dako Corp. (Carpinteria, CA); mouse monoclonal AT8 (anti-tau phosphorylated Ser202/Thr205) was from Innogenetics (Gent, Belgium). Rabbit polyclonal anti-A40 and anti-A42 antibodies were obtained from BIOSOURCE International (Camarillo, CA). Rabbit polyclonal antibodies anti-ADan (Ab 5282, raised against ADan C-terminal amino acids 22-34) and anti-ABri (Ab 338, raised against ABri C-terminal amino acids 22-34) were prepared in our laboratory (4, 6), and the IgG fractions were further purified using Gamma-bind G-Sepharose (Amersham Biosciences) 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 Corp. (Billerica, MA). SDS removal reagent SDS-OUT was from Pierce. Chemicals were from Sigma. Complete protease inhibitor mixture was purchased from Roche Applied Science. Protein content was measured by the BCA protein assay (Pierce) employing bovine serum albumin as a standard. Synthetic peptides A-(1-42) (DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA), A-(4-42) (FRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA), ADan-(1-34) (pEASNCFAIRHFENKFAVETLICFNLFLNSQEKHY; where pE represents pyroglutamate), and ABri-(1-34) (pEASNCFAIRHFENKFAVETLICSRTVKKNIIEEN), the last two featuring N-terminal pyroglutamate and internal disulfide 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 and purified by reverse-phase high performance liquid chromatography, their molecular mass was corroborated by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry, and their concentrations were 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 a 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 his ApoE genotype, tested as described below, was found to be 
4/3.
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-A 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 pretreated with 99% formic acid (FA) and a 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 Sciences). This step was followed by incubating the sections with the 6F/3D antibody overnight and treating with the appropriate secondary antibody and the ABC complex. Antibody binding was visualized with the fluorescein signal amplification kit (PerkinElmer Life Sciences). A Leica TCS SP confocal microscope running TCS NT control software or a Leica DMRE upright microscope using a three-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, whereas large vessels were separately analyzed. The extraction strategy took advantage of the differential solubility properties of preamyloid (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% 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 mixture (Complete) using a Dounce 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/2.5 g of gray matter) and further extracted as a vessel fraction (12.5 ml of PBS-2x protease inhibitors/250 mg of vessels), whereas the filtrates were subjected to ultracentrifugation in a XL100K ultracentrifuge (Beckman Coulter) using a Beckman 70.1 Ti rotor at 112,000 x g for 1 h at 4°C. The resultant supernatants were analyzed as PBS-extracted fractions, and the pellets were resuspended in 20 mM Tris, pH 7.4, containing 2% SDS (Bio-Rad) centrifuged as above although raising the temperature to 10 °C to avoid SDS crystallization, and the resultant supernatants were analyzed as SDS-extracted (preamyloid-rich) fractions. Finally, the Congo red-positive SDS-insoluble material enriched in amyloid fibrils was further extracted in 70% (v/v) FA in water and centrifuged at 14,000 rpm using a 5417R microcentrifuge (Eppendorf, Westbury, NY), and the supernatants were analyzed as FA-extracted (amyloid) fractions.
Immunoprecipitation Experiments
Tissue-extracted ADan or A 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 A, 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 µ l (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 volume of each PBS-soluble or SDS-extracted fraction. In the case of the SDS-extracted fractions, SDS was mostly removed prior to immunoprecipitation with SDS-OUT SDS precipitation reagent employing the 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 and three times with water, and the bound peptides were 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 (Bio-Rad) containing 10% (v/v) -mercaptoethanol (Sigma) and directly applied onto the 16% Tris-Tricine gels for SDS-PAGE analysis (17). In some cases, 5 µl of 1 M dithiothreitol (Sigma) was used for sample reduction instead of -mercaptoethanol.
To verify the existence of peptide-peptide interactions between A and ADan, immunoprecipitation experiments using either anti-ADan or anti-A 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-A antibodies were probed in Western blot analysis with anti-ADan antibody (Ab 5282), and conversely, samples immunoprecipitated with anti-ADan antibody were probed for A with either a mixture of 4G8 plus 6E10, anti-A40, or anti-A42 antibodies. Immunoreactivity was detected by chemiluminescence, as described below.
Circulating Soluble ADan—Plasma from two carriers of the FDD genetic defect (44 and 24 years old) as well as from 10 noncarriers was 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 radioimmuno-precipitation assay 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.
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 
-cyano-4-hydroxycinnamic acid in a 50% acetonitrile and 0.1% trifluoroacetic acid (Sigma) matrix (feasible for <10,000-Da mass range) on a Micromass Tof-Spec-2E MALDI-TOF mass spectrometer in linear mode using standard instrument settings (17). Under these experimental acidic conditions, which also render noncovalent binding unobservable (18), mostly monomeric species of ADan and A 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 (available on the World Wide Web at 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 immunoprecipitation, were separated on a 16% Tris-Tricine SDS-PAGE and electrotransferred for 45 min at 400 mA onto polyvinylidene difluoride membranes (Immobilon-P; Millipore Corp.) using 10 mM 3-cyclohexylamino-1-propanesulfonic acid (Sigma) buffer, pH 11 containing 10% (v/v) methanol. After blocking in 5% nonfat milk in PBS containing 0.1% Tween 20 (Sigma), the membranes were immunoreacted with the antibodies anti-ADan 5282 (2 µg/ml), anti-A 4G8 (1 µg/ml), anti-A 6E10 (1 µg/ml), anti-A40 (0.05 µg/ml), or anti-A42 (0.05 µg/ml), followed by either anti-rabbit or anti-mouse horseradish peroxidase-labeled F(ab')2 (Amersham Biosciences). Signals were developed with Super Signal (Pierce) and exposed to Hyperfilm ECL (Amersham Biosciences). To enhance A immunoreactivity, membranes after transfer were boiled in PBS for 5 min (19) before incubation with the specific anti-A antibodies. When necessary for reprobing, membranes were incubated in Restore Western blot stripping buffer (Pierce) for 15 min at room temperature, reblocked, and probed with another antibody.
Ligand Blot Analysis
Five micrograms each of synthetic ADan-(1-34) and ABri-(1-34) peptides were separated on a 16% Tris-Tricine SDS-PAGE and electrotransferred onto Immobilon-P membranes as described above. After blocking with 5% nonfat milk in PBS, membranes were separately incubated at 37 °C for 3 h with synthetic A-(1-42) and A-(4-42) peptides (1 µg/ml in PBS/milk). Bound A 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 a 1-h incubation at room temperature with anti-mouse horseradish peroxidase-labeled F(ab')2, and chemiluminescence was detected as above. Fifty nanograms of both ADan and ABri peptides were also subjected to Western blot analysis using a mixture of Ab 5282 (2 µg/ml) and Ab 338 (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). Aliquots from FA-solubilized amyloid fractions were separated on 16% Tris-Tricine SDS-PAGE, transferred 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-terminally sequenced.
|
RESULTS
FDD Main Histopathological Features—Fig. 1 illustrates the extent of ADan deposition (Fig. 1A) and associated tau pathology (Fig. 1B) in patients with FDD. As shown in Fig. 1A, ADan forms loose, ill defined parenchymal diffuse (preamyloid) 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 (nonfibrillar) electron-dense material with sparse fibrils, whereas the vascular deposits are of amyloid nature. The presence of A, 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 two carriers of the FDD genetic defect revealed the presence of an 4-kDa component (Fig. 2A) immunoreactive with Ab 5282 (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 (Fig. 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 disulfide bond between cysteine residues 5 and 22. Soluble ADan was not detected in plasma samples from noncarriers of the disease (n = 10) using identical experimental conditions (not shown).
Intact and N-terminally Post-translationally Modified ADan Are Predominant Components of Preamyloid 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. Fig. 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 probably 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 monomer/dimer ratio, samples had less tendency to remain monomeric. As indicated in Fig. 2C, ADan was predominantly dimeric in parenchymal extracts (ratio of monomers/dimers 1:3), whereas in vascular extracts, the monomeric form was still predominant (ratio of 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 agreement with the expected masses of the post-translationally modified ADan-pE (4045.6) and the wild-type ADan-E (4063.6), with a single disulfide bond between cysteine residues 5 and 22.
Analysis of the FA extracts revealed more complexity in terms of composition and degree of 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 triformylated species, probably on lysine and/or serine residues (21), artifactually generated during the extraction procedure. As illustrated in Fig. 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 ADan-(1-33)pE, ADan-(1-28)pE, and ADan-(1-34)pE, although the presence of ADan-(3-28), ADan-(1-27)pE, ADan-(1-30)pE, ADan-(3-33), ADan-(3-34), ADan-(1-33)E, and ADan-(1-34)E was consistently demonstrated (See Fig. 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 (Fig. 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- 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, SNX-FAIR...and EASNXF..., corresponded to ADan starting at positions 3 and 1, respectively, in a 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 (Fig. 3) and with previous published data in leptomeningeal vessels (4).
|
42 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 A species via immunoprecipitation/mass spectrometry and immunoprecipitation/Western blot using four different anti-A antibodies: 4G8 (anti-A-(17-24)), 6E10 (anti-A-(1-17)), and the C-terminus-specific anti-A40 and anti-A42. As illustrated in Fig. 4 (A and B), A species were detectable in PBS extracts from parenchyma and vessels fractions. Parenchymal extracts were immunoreactive with antibodies 4G8, 6E10, and anti-A42 but not with anti-A40. Monomers and dimers of A-(X-42) were observed with an estimated monomer/dimer ratio of 1:2 (Fig. 4A). Although a low signal was retrieved in the mass spectrometry, there was a peak with a mass suspected to correspond to the fragment A-(4-34) (experimental 3473.5; expected 3472.9); no clear peaks of A-(X-42) were detected (Fig. 4, 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 Fig. 4B, A species were immunoreactive with all antibodies, including anti-A40. Several differences were spotted when compared with the parenchymal counterparts: (i) the signals for monomers and dimers were stronger, (ii) A40 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-A42 was clearly composed of two bands of very close molecular mass. As judged by mass spectrometry analysis, the major A species in the sample were A-(1-42), A-(4-42), and A-(1-40) (Fig. 4, B and table), the last being 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 A-(4-42) and A-(1-42), 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-A42 (Fig. 4B).
|
species, and (ii) the dimeric forms were more relevant. Parenchymal extracts were immunoreactive with 4G8, 6E10, and anti-A42 but not with anti-A40 (Fig. 4C). Monomeric species, especially those immunoreactive with 6E10, seemed to be composed of more than one band in Western blot, very similar to the pattern shown in Fig. 4B. Mass spectrometry indicated that this extracts mainly contained A-(1-42) and its N-terminally truncated derivative A-(4-42), whereas A-(1-40) was a minor component (Fig. 4, C and table). The double peaks obtained in this sample originate in the presence of A 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 pretreatment of synthetic A40 and A42 with formic acid produced similar effect (not shown). A species in vessel extracts were immunoreactive with all of the antibodies tested (Fig. 4D). As curiously observed with the PBS-fractions (Fig. 4B), (i) A40 immunoreactivity was also limited to dimeric species, and (ii) the monomeric component detected by 6E10 and anti-A42 was also composed of two bands of very close molecular mass (as shown in Fig. 4, B and C). In general terms, the dimeric species were more prominent than in the PBS fractions. Mass spectrometry identified A-(1-42) and A-(4-42) as the major components whereas A-(1-40) was a minor constituent of the extract. As illustrated above with the PBS fractions, immunoprecipitation with the anti-A antibodies also retrieved a minor amount of full-length ADan (Fig. 4D).
|
|
species in both parenchymal and vascular lesions. As illustrated in Fig. 5A, parenchymal extracts were immunoreactive with 4G8, 6E10, and anti-A42 antibodies but were not recognized by anti-A40. A species were mainly dimeric, although a small proportion of higher oligomers was visible with both 4G8 and anti-A42. 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 A-(4-42). Both, A-(1-42) and A-(1-40) were also present in the sample, although A-(1-40) 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 A-(4-42) were also observed. Both experimental masses matched the expected values for A-(4-42) (Met sulfoxide) and A-(4-42) (formylated). Since samples were exposed to FA during the extraction procedure and prior to the mass spectrometry analysis, these A derivatives may well represent undesirable secondary products of the specimen treatment rather than actual post-translationally modified components present in the lesions. Alternatively, these oxidized A42 species may represent existent in vivo neurotoxic subproducts of oxidative stress, as previously proposed (22).
|
|
species seemed to be present in a slightly higher degree of oligomerization. As indicated in Fig. 5B, the vessel extracts were immunoreactive with all anti-A antibodies tested; 4G8 and anti-A42 reacted with dimers and larger oligomers, whereas anti-A40 was barely immunoreactive with the dimeric species. Mass spectrometry analysis of the same fractions revealed that A-(1-42) and its N-terminal truncated derivative A-(4-42) were the main A species with experimental masses matching their theoretical values, whereas minor A-(4-42) oxidized and formylated derivatives were also detected (Fig. 5, B and table). As observed with ADan, the abundance of FA-soluble A 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- 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 A staining of parenchymal lesions and blood vessels (Fig. 6, A-C), suggesting co-deposition on many areas. In order to further test whether the partial co-localization of A and ADan was related at least in part to a specific peptide-peptide interaction, we conducted co-immunoprecipitation experiments with tissue extracts and ligand blot analysis with synthetic peptides. As illustrated in Fig. 6D for vascular SDS extracts, a combination of anti-A antibodies 4G8 and 6E10 retrieved a mixture of monomeric and dimeric ADan immunoreactive with Ab 5282 (Fig. 6D, left). However, in the reverse experiment with the same vascular SDS-extracts, anti-ADan antibodies retrieved dimeric A that was reactive with the mixture 4G8 plus 6E10, with anti-A42 but not with anti-A40 antibodies (Fig. 6D, right). Similar results were obtained with PBS extracts (data not shown). FA extracts, on the contrary, rendered negative results despite the high concentration of both ADan and A, probably indicating that the interaction is susceptible to dissociation at low pH as previously seen in other protein-protein interactions (i.e. A-apolipoprotein J) (23).
To verify the specificity of the ADan-A interaction, ligand blot analysis using synthetic ADan and A homologues and synthetic ABri as a control was conducted in vitro. As indicated in Fig. 7, specific binding of A-(4-42) and A-(1-42) (the main A species found in FDD lesions) to ADan oligomers (particularly dimers) was clearly evident, whereas no interaction was observed with ADan monomers despite their abundance. In contrast to ADan, neither monomers nor oligomers of ABri peptides show binding interaction with either A-(4-42) or A-(1-42).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESDISCUSSIONREFERENCES |
|---|
co-exist in the absence of compact plaques. The ADan/A compositions of both lesions were analyzed by Western blot, mass spectrometry, 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 nonfibrillar components, whereas formic acid extraction was necessary to solubilize and analyze fibrillar deposits. Our biochemical data (Figs. 2, 3, 4, 5) show that FDD nonfibrillar and fibrillar lesions are heterogeneous mixtures of various ADan and A 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.
|
peptides was also highly restricted. ADan species were limited to the nondegraded full-length ADan-(1-34), 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 A peptide (24-26) as well as in some hormones and neuropeptides, including neurotensin and 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 -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 deamidation. Among them are bovine and ovine -lipotropin and joining peptide derived from the proteolytic processing of their common precursor pro-opiomelanocortin (30, 31), ABri amyloid in FBD (6), and the A N-terminal truncated derivatives A-(3-X)pE and A-(11-X)pE 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 unknown. Since this N-terminal modification was not detected in the soluble circulating ABri or ADan species, it is conceivable that cyclization occurs in situ at the sites of deposition.
The A peptides found in the nonfibrillar deposits were also restricted in diversity. A-(1-42) and the N-terminally truncated A-(4-42) were the major species in an 1:1 ratio, whereas A-(1-40) and A derivatives ending at position 34 were minor component of the extracts (Fig. 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 A-related parenchymal preamyloid lesions are enriched in species ending at position 42. The -secretase fragment A-(17-42), in particular, was previously identified as the main component of preamyloid lesions in AD (34), Down syndrome (35), and aged dogs (36), A-(1-42) and A-(4-42) being minor components in the last two cases. In FDD extracts, A-(17-42) was not detected, and although it could be argued that the present extraction conditions were milder than those previously reported (2% SDS versus 15% SDS), thereby precluding the retrieval of A-(17-42), this fragment was also undetected in the subsequent formic acid extracts. Although not through biochemical studies, the existence of other A species was also previously identified in preamyloid lesions based on classical immunohistochemical stainings and their immunoreactivity with specific antibodies. Accordingly, A-(1-42) and A-(3-42)pE were reported to be major components in AD, Down syndrome, and normal aging preamyloid lesions, whereas antibodies recognizing position 17 were only weakly positive (37). Anti-A42 immunoreactivity was also found prevalent in parenchymal preamyloid deposits in the Dutch variant of familial AD (38).
ADan and A 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 (Figs. 3 and 8), whereas A, still mainly composed of A-(1-42) and A-(4-42) (with negligible A-(1-40)), showed more N-terminal degradation (ratio of A-(4-42)/A-(1-42) 9:1) and a higher degree of oligomerization (Figs. 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 A, the N-terminal degraded A-(4-42) is known to have a faster aggregation kinetics than the intact A-(1-42) (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 Fig. 8, the proteolytic degradation of ADan and A reflects the activity of more than a single enzyme. A potential candidate for the C-terminal detyrosination 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 the ADan Ala2-Ser3 peptide bond seems to be catalyzed by a dipeptidyl-peptidase, in a similar fashion as 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 disulfide bond, as reported for other proteins (41-43).
Degradation of A at peptide bond Glu3-Phe4 may result from either neprilysin or insulysin (insulin-degrading enzyme) degradation, as previously reported in A-related disorders (44, 45). However, these enzymes are not exclusive for the Glu3-Phe4 peptide bond processing and are known to cleave A 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
[EC]
; description available on the World Wide Web at us.expasy.org/enzyme). Using synthetic A40 and A42 as substrates, this enzyme was able to efficiently cleave both peptides at Glu3-Phe4 and Leu34-Met35 bonds in in vitro experiments.3 Although collagenase I has been commonly used in amyloid purification protocols to increase the yield of the extracted fibrils (48-52), we purposely avoided its use in our FDD isolation protocol in order to prevent unwanted artifacts.
In sporadic and familial AD as well as in other A-related disorders, amyloid deposits in cerebral vessels are primarily constituted of fibrillar forms of A40, although the presence of A42 in detectable amounts is considered a common finding in these lesions (reviewed in Ref. 53). Similarly, in the various transgenic models of A deposition, A40 is the main component of vascular deposits (54-56), including the recently published APP-Dutch mouse (57). The fact that the main A components of the vascular FDD lesions were A-(X-42) species was somehow an unexpected finding of still unknown connotations for the disease pathogenesis. The importance of A42 for the formation of vascular deposits is also highlighted by the recent report on the chimeric BRI-A42 mice, which selectively overproduce A42 but not A40, and the demonstration of extensive A42 cerebral amyloid angiopathy, whereas the BRI-A40 chimeric mice overproducing A40 did not show overt amyloid pathology, calling into question the importance of A40 in the initiation of vascular amyloid deposits (58). It is also not clear whether the co-localization of ADan with A 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 A molecules were retrieved in co-immunoprecipitation experiments from tissue extracts and A-(1-42) and A-(4-42) bound to ADan dimers and higher oligomers (not monomers) in ligand blot analysis, suggesting conformational specificity. It is important to note that A was detected in neither parenchymal nor vascular ABri deposits in FBD (15, 16) nor identified as an A ligand in the ligand blot analysis shown in Fig. 7 despite being ABri and ADan molecules identical in their 22 N-terminal residues.
Mounting evidence indicates that plaque burden correlates poorly with degree of dementia and that soluble oligomers rather than the final amyloid fibrils are the harmful A 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 AD23N (63). Despite the structural differences among ABri, ADan, and A, the absence or limited number of compact plaques and extensive neurofibrillar degeneration with tangles identical to those found in Alzheimer brains are main features of these disorders. Therefore, not only are different amyloid peptides associated with similar neuropathological changes, but these disorders call into 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 preamyloid lesions as well as vascular and perivascular deposits, 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 A 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 preamyloid and vascular amyloid in the process of neurodegeneration.
|
FOOTNOTES |
|---|
1 To whom correspondence should be addressed: Depts. of Pathology and Psychiatry, New York University School of Medicine, 550 First Ave. (TH-432), New York, NY 10016. Tel.: 212-263-7997; Fax: 212-263-6751; E-mail: ghisoj01{at}popmail.med.nyu.edu.
2 The abbreviations used are: FDD, familial Danish dementia; ADan, Danish amyloid; ABri, British amyloid; AD, Alzheimer's disease; FBD, familial British dementia; A, -amyloid; MALDI-TOF, matrix-assisted laser desorption ionization-time-of-flight; PBS, phosphate-buffered saline; FA, formic acid; Ab, antibody; Tricine, N-[2-hydroxy-1, 1-bis(hydroxymethyl)ethyl]glycine.
3 Y. Tomidokoro, B. Frangione, and J. Ghiso, unpublished observations.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESDISCUSSIONREFERENCES |
|---|
, peaks also obtained in negative controls (using brain extracts but uncoated Dynabeads) and considered nonspecific.
