|Home | About | Journals | Submit | Contact Us | Français|
This chapter outlines protocols that produce homogenous preparations of oligomeric and fibrillar amyloid -β peptide (Aβ). While there are several isoforms of this peptide, the 42 amino acid form is the focus because of its genetic and pathological link to Alzheimer’s disease (AD). Past decades of AD research highlight the dependence of Aβ42 function on its structural assembly state. Biochemical, cellular and in vivo studies of Aβ42 usually begin with purified peptide obtained by chemical synthesis or recombinant expression. The initial steps to solubilize and prepare these purified dry peptide stocks are critical to controlling the structural assembly of Aβ. To develop homogenous Aβ42 assemblies, we initially monomerize the peptide, erasing any “structural history” that could seed aggregation, by using a strong solvent. It is this starting material that has allowed us to define and optimize conditions that consistently produce homogenous solutions of soluble oligomeric and fibrillar Aβ42 assemblies. These preparations have been developed and characterized by using atomic force microscopy (AFM) to identify the structurally discrete species formed by Aβ42 under specific solution conditions. These preparations have been used extensively to demonstrate a variety of functional differences between oligomeric and fibrillar Aβ42. We also present a protocol for fluorescently labeling oligomeric Aβ42 that does not affect structure, as measured by AFM, or function, as measured by a cellular uptake assay. These reagents are critical experimental tools that allow for defining specific structure/function connections.
Currently, research is focused on soluble oligomeric assemblies of Aβ42 as the proximate cause of the neuropathology that defines AD. Controlling Aβ assembly is critically important as Aβ structure determines its function (Figs. 1 and and2).2). Numerous experiments have addressed methods to characterize Aβ structure (for review, refs. (1, 2)). These studies demonstrate that peptide conformation and aggregation behavior are highly dependent on initial solvent conditions (Fig. 3) and subsequent solution conditions (Fig. 1a). Oligomer preparations are defined using a variety of different methods, including neurotoxic activities, isolation techniques (primarily size exclusion chromatography (SEC)), size estimation such as by SDS or native PAGE, imaging techniques, and reactivity with various Aβ conformation-specific antibodies. These multiple operative definitions of oligomeric Aβ have resulted in a literature that is often difficult to interpret and almost impossible to compare. A rigorous approach is particularly important with Aβ42, which aggregates faster and to a significantly greater extent than Aβ40 and other shorter forms of the peptide (Fig. 1a, b).
AFM is particularly well suited to the analysis of amyloidogenic peptides and proteins that can assemble into a variety of structurally discrete species, specifically those like Aβ. Polydispersity of morphologies and sizes often complicates or precludes the use of other biophysical techniques (such as NMR or light scattering methods), or is masked by solvent incompatibilities of the bulk solution (as for secondary structure detected by far-UV circular dichroism). Techniques based on separation by size (SDS-PAGE, Native PAGE and SEC) may lead to apparent multimers/sizes arising from technical artifacts due to matrix effects. AFM is one of the few techniques that provide direct, high-resolution, 3-dimensional morphological images of the broad range of structures present in a single scan without the need for chemical manipulation of the sample. Numerous studies have demonstrated several advantages of tapping mode AFM for Aβ42 morphological characterization (3–8). We have used AFM for developing conditions that consistently produce homogenous preparations of oligomeric or fibrillar assemblies of Aβ42 (9, 10). We have used these preparations extensively to demonstrate significant functional differences between Aβ42 oligomers and fibrils using a variety of experimental models (for example, Figs. 1c and and2b)2b) (9, 11–14).
To directly assess the conformation-dependent differences among Aβ assemblies, we have developed protocols for the preparation of homogeneous unaggregated, oligomeric, and fibrillar Aβ42 (9, 10) (Figs. 1 and and4).4). Because Aβ42 is the isoform of the peptide most associated with AD, we chose to utilize it almost exclusively, with Aβ40 used occasionally as a negative control (Fig. 1b, d). Using AFM to image Aβ42, we remove preexisting aggregates and β-sheet secondary structure from Aβ42 with a strong fluorinated alcohol, hexafluoroisopropanol (HFIP) (Fig. 3), followed by solubilization of the now monomerized peptide in dimethylsulfoxide (DMSO). Starting with this monomeric peptide preparation, we further developed two aggregation protocols that consistently produce extensively oligomeric or fibrillar populations of Aβ42 (Fig. 1). For the “unaggregated” peptide preparation, the DMSO-solubilized peptide is diluted in the experimental solution (for example, culture media) and used immediately (Fig. 1a, 0 h). To grow a “plaque in a dish,” follow the fibril forming procedure, with the addition of salt at physiological concentrations (Fig. 5a2). Note that AFM is an optimal method for determining the aggregation state of Aβ42 as it is difficult to consistently identify Aβ42 assemblies by Western analysis of SDS-PAGE (Fig. 5) (15).
These distinct assemblies are derived from chemically identical and structurally homogeneous starting materials and are thus particularly well suited for comparative structure–function studies. We have demonstrated that in vitro, oligomeric Aβ42 is ~10-fold more neurotoxic than the fibrillar (plaque-forming) assembly, and ~40-fold more toxic than the unaggregated peptide, with oligomeric Aβ42-induced toxicity significant at 10 nM (Fig. 1c). Under Aβ42 oligomer- and fibril-forming conditions, Aβ40 remains predominantly as unassembled monomer (Fig. 1b) and had significantly less effect on neuronal viability than preparations of Aβ42 (Fig. 1d). We applied the aggregation protocols developed for wild type (WT) Aβ42 to Aβ42 with the Dutch (E22Q) or Arctic (E22G) mutations (Fig. 2). Oligomeric preparations of the mutant peptides exhibited extensive protofibril and fibril formation, respectively, but were not consistently different from WT Aβ42 in terms of inhibition of neuronal viability. However, fibrillar preparations of the mutants appeared larger in diameter and induced significantly more inhibition of neuronal viability than WT Aβ42 fibril preparations. These data demonstrate that protocols developed to produce oligomeric and fibrillar Aβ42 are useful in distinguishing the structural and functional differences between Aβ42, Aβ40, and Aβ containing known genetic mutations.
As researchers become increasingly conscientious of utilizing structurally uniform, well-characterized Aβ preparations, the same criteria need to be applied to fluorophore-labeled-Aβ, prior to their widespread use as experimental tools. Numerous recent studies utilizing fluorophore-labeled Aβ42 peptides demonstrate this need for defined methods of consistently preparing well-characterized fluorescent Aβ assemblies (16–31). The fluorescent Aβ42 reagents used to date are prepared from different sources of Aβ assemblies, in many cases using Aβ42 preparations that have not yet been structurally/morphologically characterized. Thus, structural comparisons between the unlabeled and labeled Aβ assemblies are not possible. Establishing the specific structural form of the assemblies, by AFM and other methods, is necessary to be able to interpret and compare results from the various fluorescent Aβ42 species. We present a method for preparing Alexa Fluor® 488-labeled Aβ oligomers, extending our structural and functional characterization to fluorophore-labeling of Aβ42 oligomers. Structural characterization by AFM establishes a method for labeling uniform oligomeric assemblies that is comparable to unlabeled oligomeric Aβ42 (Fig. 6a). To compare function, we demonstrate that the uptake of labeled and unlabeled oligomeric Aβ42 by neurons in vitro is also similar (Fig. 6b) (see Note 1). These well-characterized fluorophore-Aβ42 oligomers are an exciting new reagent for use in a variety of studies designed to elucidate critical cellular and molecular mechanisms underlying the functions of this Aβ42 assembly form in AD.
Ham’s F-12, phenol red-free cell culture media, supplemented with 146 mg/L L-Glutamine (see Note 4).
10 mM hydrochloric acid solution (prepared in ultrapure H2O from a 1 M HCl stock).
10 mM HCl containing 150 mM NaCl.
Steps 1–7 need to be done in a fume hood.
We gratefully acknowledge financial support for these studies from NIH 1F32AG030256-01 (LMJ), Alzheimer’s Association NIRG-06-26957 (CY), NIH R01 AG19121 (MJL), NIH PO1AG030128-A2 (MJL), Alzheimer’s Association Zenith Award ZEN-08-89900 (MJL), and NIH (NIA) PO1AG021184 (MJL). We also gratefully acknowledge Kevin Laxton and Amy Pham for technical and intellectual contributions.
1Cellular uptake by neurons is reported only for oligomeric Aβ42, as treatment with fibrillar Aβ42 does not result in any detectable uptake in the model described (data not shown).
2In-house synthetic peptide or peptide from other vendors will also work, but it must be of very high purity and quality. The TFA salt (as opposed to the acetate or ammonium salt) is preferred. In-house material should be accurately weighed in clean glass vials with a HFIP-resistant closure.
3Dry DMSO stocks can be made by transferring DMSO from a freshly opened ampule to a 1–2 mL glass vial with a DMSO-resistant closure (such as Teflon – VWR, Cat. No. 66009-556). Store vials containing the dry DMSO in a dessicated glass jar in the dark at RT and discard after 2 weeks.
4The glutamine supplementation is to match the composition of the Biosource phenol red-free F-12 media, which was described in the original oligomer protocol (10) but is no longer available.
5Originally, we used the 3-[4,5-dimethylthiazol-2-yl]-2,5-di-phenyl tetrazolium bromide (MTT) assay (Roche Molecular Biochemicals) as a measure of neurotoxicity (9, 11) (Figs. 1c, d and and2b).2b). This method is based on the reduction of internalized MTT tetrazolium to a colored formazan compound by cellular redox potential. The formazan production is proportional to viable cells in culture. However, MTT reduction does not necessarily reflect cellular metabolic activity, as some Aβ assemblies may also enhance exocytosis of MTT formazan (32). This assay also requires relatively long staining and extraction times. Therefore, we now use the Promega CellTiter-Glo® Luminescent Cell Viability Assay as a measure of in vitro neurotoxicity. This has resulted, as one would predict, in lower toxicity for comparable doses of Aβ, and so higher doses of Aβ42 are now required to achieve the same toxicity (12).
6HFIP is corrosive and very volatile. Avoid contact and work in the fume hood; take care not to contact septum or other surfaces during solubilization.
7Peptide comes stored under vacuum, and the peptide in the bottom of the vial needs to be in solution before the vacuum is broken. After the peptide is in solution, pierce the septum with a syringe needle to release the vacuum. For other peptides, add enough HFIP such that the final peptide concentration is 1 mM. Use proper sterile technique to avoid any bacterial contamination when the peptide stocks are resuspended in culture media or buffer.
8Solution should be clear and colorless. Any trace of yellow color or cloudy suspension indicates poor peptide quality and should not be used. Some peptides may require brief (~5 min) bath sonication.
9Do not use siliconized tubes for the preparation of HFIP stocks. Be careful when dispensing HFIP solution and watch for bubbles. Leave tubes open when evaporating HFIP overnight.
10The peptide should not be white or chunky. An even clear film is a strong indicator of good peptide quality.
11These stocks should be stable for several months to years.
12DMSO stock should be clear and colorless. Remember to use proper sterile technique.
13Do not store peptide as a DMSO stock for more than 1 h to avoid protofibril formation.
14While the “unaggregated” prep is an ideal control for conformation, it is most useful in assays that require either a very low concentration of peptide (9) or a short incubation period (14). Prolonged incubation at higher concentrations result in the uncontrolled aggregation of the peptide and unpredictable functional activity.
15Do not keep 5 mM Aβ stock on ice because the DMSO will solidify.
16Western analysis by SDS-PAGE is not a method for assessing the conformation/assembly of Aβ42 (Fig. 5) (15). However, it is useful for visualizing the relative amount of peptide for comparison between samples.
17Depending on the age of the electrode/power supply equipment, voltage, and current settings may affect the pattern and abundance of bands typically observed for Aβ (monomer, dimer, trimer, and tetramer). We have found that power supply limits set at 90–100 V, 80 mA for 80–90 min for electrophoresis yield consistent results.
18For Aβ preparations in F-12, the HCl pretreatment of the mica improves consistent and uniform peptide adsorption to the mica. For Aβ preparations in HCl or PBS, including the Alexa Fluor® 488-labeled oligomers, no mica pretreatment is performed.
19Dried sample disks can be stored in a helium-purged desiccator for several months.