Date Published: August 21, 2014
Publisher: Public Library of Science
Author(s): Erin R. Greiner, Jeffery W. Kelly, Fernando L. Palhano, Salvador Ventura.
Amyloid fibrils are associated with many maladies, including Alzheimer’s disease (AD). The isolation of amyloids from natural materials is very challenging because the extreme structural stability of amyloid fibrils makes it difficult to apply conventional protein science protocols to their purification. A protocol to isolate and detect amyloids is desired for the diagnosis of amyloid diseases and for the identification of new functional amyloids. Our aim was to develop a protocol to purify amyloid from organisms, based on the particular characteristics of the amyloid fold, such as its resistance to proteolysis and its capacity to be recognized by specific conformational antibodies. We used a two-step strategy with proteolytic digestion as the first step followed by immunoprecipitation using the amyloid conformational antibody LOC. We tested the efficacy of this method using as models amyloid fibrils produced in vitro, tissue extracts from C. elegans that overexpress Aβ peptide, and cerebrospinal fluid (CSF) from patients diagnosed with AD. We were able to immunoprecipitate Aβ1–40 amyloid fibrils, produced in vitro and then added to complex biological extracts, but not α-synuclein and gelsolin fibrils. This method was useful for isolating amyloid fibrils from tissue homogenates from a C. elegans AD model, especially from aged worms. Although we were able to capture picogram quantities of Aβ1–40 amyloid fibrils produced in vitro when added to complex biological solutions, we could not detect any Aβ amyloid aggregates in CSF from AD patients. Our results show that although immunoprecipitation using the LOC antibody is useful for isolating Aβ1–40 amyloid fibrils, it fails to capture fibrils of other amyloidogenic proteins, such as α-synuclein and gelsolin. Additional research might be needed to improve the affinity of these amyloid conformational antibodies for an array of amyloid fibrils without compromising their selectivity before application of this protocol to the isolation of amyloids.
Maintenance of protein homeostasis, or proteostasis, is accomplished by the proteostasis network comprising biological pathways that control the rate of protein synthesis and the efficiency of protein folding, trafficking and degradation . The aggregation of peptides or proteins, exacerbated by aging, is genetically and pathologically linked to degenerative disorders, including Alzheimer’s disease (AD), Parkinson’s disease, and the systemic amyloid diseases . A wide range of proteins, including those normally existing in a soluble folded state or as an intrinsically disordered monomer, can form cross-β-sheet amyloid fibrils owing to a mutation or because of environmental alterations . Amyloid deposits can be detected using Congo red birefringence or thioflavin T fluorescence, and are often associated with glycosaminoglycans, the amyloid P component, or other proteins . Amyloid fibrils are made up of multiple interacting filaments, which are each comprised of thousands of monomers arranged at least as two-layer cross-β-sheets . Amyloid is generally relatively resistant to denaturation and proteolysis . Because amyloid is stabilized by backbone H-bonding and side chain-side chain hydrophobic interactions, it has been proposed that any protein, regardless of its amino acid sequence, can form amyloid fibrils if subjected to appropriate solution conditions , .
We produced amyloid fibrils using three different proteins, namely, Aβ1–40, α-syn and the 8 kDa fragment of gelsolin. Aβ1–40 is the peptide associated with Alzheimer’s disease , α-syn is associated with Parkinson’s disease , and gelsolin with Familial amyloidosis of Finnish type ,  (Figure 1A). The fibrils formed from the three different proteins presented with the typical amyloid structure (panels B–D), as seen by transmission electron microscopy (TEM). The aggregates formed by α-syn were long and twisted (Figure 1B), whereas gelsolin and Aβ1–40 aggregates were shorter, with some clusters (Figure 1C, 1D). These results were confirmed by Congo Red (CR; Figure 1E, red bars) and ThT (Figure 1E, yellow bars) binding. These two compounds are amyloid-specific dyes that change their spectroscopic behavior when bound to the cross-β fold present in amyloid fibrils , . As a negative control, we assessed ThT and CR binding using buffer alone (Figure 1E) or soluble peptides of Aβ1–40, α-syn or gelsolin (not shown due to the similarity with the buffer control). On average, we observed a 10–20 fold higher ThT and CR signal with aggregated peptides than with buffer alone (Figure 1E).
Several groups have described the use of bioinformatics – and in vivo screening ,  to find new amyloid. Biochemical analytical methods are useful for this purpose but few or no targets were subsequently validated by other assays, showing that the isolation of amyloid fibrils is challenging . Our two-step strategy lays the groundwork for developing a sensitive assay for the purification and detection of amyloid fibrils. One limitation of our strategy was the IP step, probably imposed by the affinity of the LOC antibody for α-syn and gelsolin amyloid fibrils. It is important to emphasize that the LOC antibody was able to recognize all amyloid fibrils tested as presented before by Glabe’s group . However, probably due the complexity of the reaction medium used in this work, the ability of LOC antibody to immunoprecipitate different kinds of fibrils was compromised. Nevertheless, we could immunoprecipitate picograms of Aβ fibrils by the use of the protocol described here. We envision a scenario where new amyloid conformational antibodies can be created, making the use of this methodology generic and not restricted to purification of Aβ fibrils. The LOC antibody was efficient in imunoprecipitating Aβ fibrils produced in vivo and methodology described here can be useful to purify Aβ fibrils from biological samples, rendering the fibrils available for more accurate structural and biochemical characterization. We hope that the goals and limitations presented in this work give new insight to the research community to enable the development of a method that can be used to isolate amyloid fibrils from complex solutions.