Using bacterial inclusion bodies to screen for amyloid aggregation inhibitors
- First Online:
- Cite this article as:
- Villar-Piqué, A., Espargaró, A., Sabaté, R. et al. Microb Cell Fact (2012) 11: 55. doi:10.1186/1475-2859-11-55
The amyloid-β peptide (Aβ42) is the main component of the inter-neuronal amyloid plaques characteristic of Alzheimer's disease (AD). The mechanism by which Aβ42 and other amyloid peptides assemble into insoluble neurotoxic deposits is still not completely understood and multiple factors have been reported to trigger their formation. In particular, the presence of endogenous metal ions has been linked to the pathogenesis of AD and other neurodegenerative disorders.
Here we describe a rapid and high-throughput screening method to identify molecules able to modulate amyloid aggregation. The approach exploits the inclusion bodies (IBs) formed by Aβ42 when expressed in bacteria. We have shown previously that these aggregates retain amyloid structural and functional properties. In the present work, we demonstrate that their in vitro refolding is selectively sensitive to the presence of aggregation-promoting metal ions, allowing the detection of inhibitors of metal-promoted amyloid aggregation with potential therapeutic interest.
Because IBs can be produced at high levels and easily purified, the method overcomes one of the main limitations in screens to detect amyloid modulators: the use of expensive and usually highly insoluble synthetic peptides.
KeywordsAmyloidsInclusion bodiesProtein foldingProtein aggregationMetalsAlzheimer
In the last few years, protein aggregation has emerged from a neglected area of protein chemistry as a transcendental issue in biological and medical sciences, mainly because the deposition of proteins into insoluble amyloid fibrils is being found behind an increasing number of human diseases such as Alzheimer’s disease (AD) or type II diabetes[1–4]. Therefore, there is an increasing interest in developing methods to identify molecules that trigger the aggregation of these proteins inside the organism as well as to discover chemical compounds that can interfere with these pathways, having thus therapeutic potential[5–7].
The pathological hallmark of AD is brain deposition of amyloid plaques composed predominantly by the Aβ42 peptide isoform[8–10]. The origin of these insoluble extracellular neurotoxic deposits is still not completely clear, and multiple factors such as pH, peptide concentration, oxidative stress and metal ions have been reported to trigger their formation[11, 12]. Here we present a fast, cost-effective high-throughput approach to study conditions and molecules that affect Aβ42 aggregation. The assay is based on the use of the inclusion bodies (IBs) formed by an Aβ42-GFP fusion protein in bacteria. IBs formation has long been regarded as an unspecific process relaying on the establishment of hydrophobic contacts[13, 14]. However, there are now strong evidences demonstrating that bacterial IBs formation shares a number of common features with the formation of the highly ordered and pathogenic amyloid fibrils linked to human diseases[15–18]. Therefore, IBs have become an attractive model to study protein aggregation and their consequences in simple but biologically relevant environments[19–21]. IBs are formed inside the cell when the folding of proteins into native conformations is competed by a faster establishment of anomalous intermolecular interactions that leads to the formation of insoluble aggregates. In the present work, we exploit this kinetic competition in vitro to screen for compounds that can modulate protein aggregation. As a proof of principle, we demonstrate the ability of the approach to detect the effect of metal ions on Aβ42 aggregation as well as to identify compounds that block this metal-induced reaction.
Results and Discussion
Refolding Aβ42-GFP IBs is sequence specific
We wondered if the kinetic competition between GFP folding and Aβ42 aggregation can be reproduced in vitro. To this aim we used the IBs formed by the wild-type peptide fusion (Aβ42wt-GFP) and the F19D mutant (Aβ42F19D-GFP), which display the highest and lowest aggregation propensities in our library, respectively. Purified IBs were denatured to remove the polypeptide contacts supporting the aggregate structure. This provides unfolded and isolated species for the subsequent in vitro refolding step and guarantees that all inter- or intra-molecular contacts are established de novo as it happens after protein synthesis in the cell. IBs were chemically denatured using two chaotropic agents, 10 M urea and 8 M Gu·HCl. Each unfolded Aβ42-GFP fusion was diluted in refolding buffer and the amount of recovered active GFP monitored using fluorescence spectroscopy (see Methods). The same conditions were used to unfold and refold equimolar concentrations of native untagged GFP. As it can be seen in Figure1A, independently of the IBs peptide variant, the level of recovered GFP activity was higher when Gu·HCl was used as denaturant. This is in contrast with the results obtained with untagged GFP, for which denaturation with urea resulted in higher fluorescence recovery (Figure1B), suggesting that the used denaturant might affect the aggregation/refolding pathway. The proportion of fluorescent GFP recovered after refolding was always higher than that in the original IB (Figure1A). Aggregation usually corresponds to a second or higher order reaction and therefore, aggregation rates are extremely dependent on protein concentrations. Since the protein concentrations used during in vitro refolding are much lower than those existent in vivo, the folding of the GFP domain can compete more efficiently with the aggregation process, providing a larger dynamic response than in bacteria. However, the refolding efficiency of Aβ42-GFP IBs is about ~10-fold and ~4-fold lower than this of untagged GFP after denaturation in urea and Gu·HCl, respectively, suggesting that, as it happens in vivo, the aggregation of the Aβ42 moiety competes the folding of GFP. Importantly, the activity recovery from the mutant IBs is higher than that from IBs formed by the wild-type sequence, supporting a kinetic competition between GFP folding and Aβ42 aggregation in vitro. The predicted lower aggregation rate of the mutant would account for the higher fluorescence recovery. By analogy, any agent that would increase the intrinsic aggregation rate of Aβ42 will decrease the final amount of functional GFP and vice versa, allowing to screen for promoters or inhibitors of the protein aggregation process.
Detection of the Aβ42 aggregation-promoting effect of ionic metals
Identification of inhibitors of metal-triggered Aβ42 aggregation
Although, to our knowledge, no in vivo effects of o-Vanillin on Aβ42 promoted neuronal toxicity have been reported so far (work in progress). A closely related compound differing only in a CH2 group, 2-Hydroxy-3-ethoxybenzaldehyde, completely blocked the neurotoxicity of the peptide to rat hippocampal neurons in culture, indicating that despite the simplicity of our assay, it may identify physiologically relevant hit compounds.
To obtain further insights on the effects of copper, zinc and o-Vanillin on Aβ42 aggregation, we monitored the kinetics of GFP refolding after IBs denaturation in the presence and absence of these molecules by following the changes in fluorescence emission (Figure5E). In PBS, GFP fluorescence was recovered following a double exponential curve with a rate constant of 0.90 ± 0.02 s-1 and a half-life of 46.21 min for the fast reaction phase. The presence of both copper and zinc abrogated completely the fluorescence recovery already at the beginning of the refolding reaction, likely indicating that they promote a very fast aggregation of the fusion protein that totally competes the GFP domain folding reaction. The presence of o-Vanillin has a negligible effect on copper containing solutions. In contrast, this molecule allows recovery of 70 % of the fluorescence at the end of the reaction in the presence of zinc. The rate constants and half-life for the fast phase were very close to those exhibited in the absence of metals, with values of 0.87 ± 0.03 s-1 and 48.29 min, respectively. This indicates that this compound acts interfering with zinc promoted Aβ42 aggregation without affecting GFP folding. Interestingly, the GFP fluorescence recovery reaction is completed after 3.5 h, being thus a faster assay than those relying on the aggregation of synthetic peptides, which usually require at least overnight incubation. We used the metallochromic Zincon reagent to quantify the free levels of Zn2+ and Cu2+ in the absence and presence of o-Vanillin using spectrophotometry. No differences in free ion metal levels were observed (data not shown) suggesting that the compound does not act as a chelator but rather affects the refolding/aggregation kinetics of misfolded GFP fusions.
Production and purification of inclusion bodies
Escherichia coli BL21DE3 competent cells were transformed with pET28 vectors (Novagen, Inc., Madison, WI, USA) encoding the sequences for Aβ42wt-GFP fusion and the mutant Aβ42F19D-GFP, as previously described.
10 mL of bacterial cultures were grown at 37°C and 250 rpm in LB medium containing 50 μg/mL of kanamycin. At an OD600 of 0.5, 1 mM of isopropyl-β-D-1-thiogalactopyranoside (IPTG) was added to induce recombinant protein expression.
After 4 hours, cells were harvested by centrifugation and pellets were re-suspended in lysis buffer (100 mM NaCl, 1 mM EDTA and 50 mM Tris pH 8) to purify intracellular inclusion bodies (IBs), as previously described. Briefly, protease inhibitor PMSF and lysozyme were added at the final concentrations of 15 mM and 300 μg/mL, respectively. After incubating at 37°C for 30 min, detergent NP-40 was added at 1 % and cells were incubated at 4°C for 50 min under mild agitation. To remove nucleic acids, cells were treated with DNase and RNase at 15 μg/mL at 37°C for 30 min. IBs were collected by centrifugation at 12,000xg for 10 min and washed with lysis buffer containing 0.5 % Triton X-100. Finally, they were washed three times with PBS to remove remaining detergent.
In vitro refolding assay
15 μL of purified IBs at OD360 = 10 were centrifuged for 10 min at 12000xg. To denature the aggregates, the pellets were re-suspended in 10 μL of 8 M Gu·HCl or 10 M urea and incubated at room temperature for 4 h. For the refolding process, denatured aggregates were dissolved in 990 μL of refolding buffer. These buffers were based on PBS, previously treated with Chelex 100 chelating resin from Sigma-Aldrich (St. Louis, MO, USA), and the following salts and compounds according to the different refolding assays: CaCl2, FeCl3, MgCl2, NaCl, NiCl2, ZnCl2, CuCl2, apigenin, azure C, basic blue 41, congo red, curcumin, hemin chloride, meclocycline sulfosalicylate, myricetin, nordihydroguaiaretic acid, o-Vanillin (2-hydroxy-3-methoxybenzaldehyde), thioflavin -T and quercetin, all obtained from Sigma-Aldrich (St. Louis, MO, USA). Equimolar concentrations of purified untagged GFP were used in control experiments. GFP fluorescence of the solutions containing refolded IBs or untagged GFP were measured in a 96 well plate in a Victor 3 Plate Reader (Perkin-Elmer, Inc., Waltham, MA, USA) using excitation and emission wavelength filters of 405 nm and 510 nm, respectively or in a Jasco FP-8200 spectrofluorometer using excitation and emission wavelengths of 480 nm and 510 nm, respectively. Measurements were performed in triplicate. For kinetic experiments, the refolding step was followed using the same parameters and reading the fluorescence emission every 2 min for 16 h. In order to homogenize the samples, these were briefly shacked (for 5 s) before each determination.
Transmission electronic microscopy
IBs or aggregates containing solutions were placed on carbon-coated copper grids, and left to stand for five minutes. The grids were washed with distilled water and stained with 2 % (w/v) uranyl acetate for another two minutes before analysis using a HitachiH-7000 transmission electron microscope (Hitachi, Tokyo, Japan) operating at accelerating voltage of 75 kV.
Secondary structure determination
Aggregates present in refolding solutions were precipitated by centrifugation at 12.000 xg (g en cursiva i sense espais) for 30 min, resuspended in Milli-Q water and analyzed, together with purified untagged GFP, by FT-IR spectroscopy using a Bruker Tensor 27 FT-IR Spectrometer (Bruker Optics Inc) with a Golden Gate MKII ATR accessory. Each spectrum consists of 16 independent scans, measured at a spectral resolution of 2 cm-1 within the 1700–1500 cm-1 range. All spectral data were acquired and normalized using the OPUS MIR Tensor 27 software.
This work was supported by grants BFU2010-14901 from Ministerio de Ciencia e Innovación (Spain) and 2009-SGR 760 from AGAUR (Generalitat de Catalunya). RS is recipient of a contract from the Ramón y Cajal Programme from Ministerio de Ciencia e Innovación. NSG is recipient of a FEBS long-term fellowship. SV has been granted an ICREA ACADEMIA award (ICREA).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.