The polyglutamine (polyQ)-containing protein ataxin-3 (AT3) triggers the neurodegenerative disease spinocerebellar ataxia type 3 (SCA3) when its polyQ tract is expanded beyond a critical length. This results in protein aggregation and generation of toxic oligomers and fibrils. Currently, no effective treatment is available for such and other polyQ diseases. Therefore, plenty of investigations are being carried on to assess the mechanism of action and the therapeutic potential of anti-amyloid agents. The polyphenol compound epigallocatechin-3-gallate (EGCG) and tetracycline have been shown to exert some effect in preventing fibrillogenesis of amyloidogenic proteins. Here, we have incubated an expanded AT3 variant with either compound to assess their effects on the aggregation pattern. The process was monitored by atomic force microscopy and Fourier transform infrared spectroscopy. Whereas in the absence of any treatment, AT3 gives rise to amyloid β-rich fibrils, whose hallmark is the typical glutamine side-chain hydrogen bonding, when incubated in the presence of EGCG it generated soluble, SDS-resistant aggregates, much poorer in β-sheets and devoid of any ordered side-chain hydrogen bonding. These are off-pathway species that persist until the latest incubation time and are virtually absent in the control sample. In contrast, tetracycline did not produce major alterations in the structural features of the aggregated species compared with the control, but substantially increased their solubility. Both compounds significantly reduced toxicity, as shown by the MTT assay in COS-7 cell line and in a transgenic Caenorhabditis elegans strain expressing in the nervous system an AT3 expanded variant in fusion with GFP.
Amyloidoses are clinical disorders caused by deposition of proteins that abnormally self-assemble into insoluble fibrils and impair tissue–organ function. More than 20 unrelated precursor proteins can undergo misfolding, which is followed by protein aggregation into β-sheet rich, amyloid fibrils (1,2). Increasing evidence suggests that the most toxic species in cells are not the mature amyloid fibrils, but the pre-fibrillar oligomeric structures (3,4). In line with this idea, some evidence also suggests that formation of mature fibrillar aggregates may even be a defense mechanism for the cell (5). The discovery of molecules that inhibit protein deposition or reverse fibril formation could certainly disclose new avenues for developing therapeutic strategies aimed at preventing or controlling the corresponding amyloid-related diseases. Different classes of structurally unrelated compounds have been investigated for their ability to interfere with protein self-aggregation and stability of amyloid fibers (6). Among these, the flavonoids represent a large group of naturally occurring polyphenolic substances, well tolerated and abundant in some foods (7). Epigallocatechin-3-gallate (EGCG) is the most represented tea catechin (50–80% of total catechins) and it is known as a potent antioxidant via direct scavenging of reactive oxygen species (ROS) and reactive nitrogen species, induction of defense enzymes and binding of divalent metals, such as copper and iron (8,9). Furthermore, EGCG was reported to cross the blood–brain barrier in mammals (10) and to be safe for humans when tested in clinical studies (11). Current data also show that EGCG interacts with a large variety of amyloid-forming proteins such as amyloid beta (Aβ) (12), α-synuclein (13), transthyretin (14) and huntingtin (15), producing unstructured, off-pathway oligomers and reducing toxicity.
Another well-studied group of inhibitors are tetracyclines, a class of drugs capable of crossing the blood–brain barrier and already used in clinical practice, offering the advantage of a safe toxicological profile and well-characterized pharmacological properties (16–18). It has been shown that such compounds, to different extents, prevent fibrillogenesis of prion protein (PrP) (17), transthyretin (19), α-synuclein (20), β2-microglobulin (21) and Aβ (22,23). In most of these cases, tetracyclines are also able to redissolve mature fibrils. Although the exact mechanism of anti-amyloidogenic activity of tetracyclines is largely unknown, it is likely related to their ability to interfere with the formation of fibrillar aggregates (17,24). They may also contribute to improving other pathological events associated with amyloid deposit formation, including inflammation, ROS generation causing oxidative stress, apoptosis and uncoupling of metal homeostasis (25).
In this study, we sought to evaluate the effect of EGCG and tetracycline on the aggregation process and toxicity of the expanded ataxin-3 (AT3), the polyglutamine (polyQ)-containing protein responsible for spinocerebellar ataxia type 3 (SCA3). To date, no effective treatment has been developed for such disease and no compounds have been tested for their effect on AT3 aggregation process. We therefore studied the in vitro effects of EGCG and tetracycline on expanded AT3 aggregation by taking advantage of different analytical methods, in particular Fourier transform infrared (FTIR) spectroscopy and atomic force microscopy (AFM). We demonstrated that the two compounds differently modulate the protein aggregation. EGCG interferes within the early steps of aggregation, accelerating the misfolding of the Josephin domain (JD) and preventing the formation of mature fibrils; in contrast, tetracycline modulates the process by increasing the solubility of aggregated species, with no major alterations in their structural features. In both cases, co-incubation of AT3 with these compounds reduces the toxicity of protein aggregates in COS-7 cells. We also confirmed the effects of the two inhibitors in vivo, using transgenic Caenorhabditis elegans as a simplified SCA3 model, which provides a clear evidence of the beneficial effect of the two compounds on the disease, although neither EGCG nor tetracycline were able to disassemble preformed AT3-amyloid fibrils in vitro.
EGCG and tetracycline differently affect AT3 aggregation kinetics and solubility
In recent years, plenty of evidence has highlighted a critical role for soluble oligomeric amyloid species in triggering cellular toxicity (26,27). Here, we have examined whether and how EGCG or tetracycline affect the amyloid aggregation pattern of an expanded variant of AT3, thus preventing its toxic effects. We produced His-tagged, monomeric AT3Q55 to a degree of purity of at least 93%, as assessed by densitometric analysis of SDS-gels (Supplementary Material, Figs S1 and S2A). Also, no cross-reactive degradation products were detected by western blotting in preparations of freshly purified AT3Q55 (Supplementary Material, Fig. S2B). The protein (25 μm) was then incubated at 37°C in the presence or the absence of either compound at 1 : 1 or 1 : 5 protein–drug molar ratios. Aliquots were taken at different incubation times and the soluble fraction isolated as the supernatant from a centrifugation at 14 000g. The protein content was determined using Bradford assay. Compared with the control, both treatments resulted in a substantially slower decline in soluble protein content starting from 24 h of incubation (Fig. 1A and B). SDS-soluble protein fraction was also quantified by SDS–PAGE of the supernatants and subsequent densitometric analyses. Starting from the earliest incubation time (3 h), EGCG treatment resulted in a significant reduction in SDS-soluble amount of the protein (Fig. 1C, E and F; Supplementary Material, Fig. S3). In contrast, tetracycline somewhat retarded its disappearance (Fig. 1D, E and G; Supplementary Material, Fig. S3). The effects were best detected at the highest protein–drug molar ratios.
Surprisingly, the reduced solubility of EGCG-treated protein was paralleled by the appearance of large SDS-resistant aggregates in the soluble fraction (Fig. 1F). These soluble aggregates are large in size and do not enter the separating gel (>250 kDa). In contrast, in time course experiments without EGCG, large SDS-resistant complexes were hardly detected in the soluble fraction (Fig. 1E). Tetracycline treatment yielded a pattern qualitatively similar to that of untreated protein, as regards SDS-resistant species accumulation (Fig. 1G). Further characterization of the aggregation products was performed by SEC (Supplementary Material, Fig. S4). In the control sample (Supplementary Material, Fig. S4A), higher molecular-weight products appeared in the void-volume starting from 3 h of incubation, which correspond to a molecular mass of 300 kDa or higher. Scanty, if any, accumulation of intermediate forms between void-volume and monomeric protein was observed. Tetracycline treatment (Supplementary Material, Fig. S4C) did not appreciably altered this pattern, the only significant difference being a somewhat faster formation of higher molecular-weight forms compared with the control sample. In contrast, EGCG treatment (Supplementary Material, Fig. S4B) resulted in a much faster disappearance of the monomeric form and accumulation of aggregates, with also appreciable accumulation of intermediate forms. This fits well with the aggregation pattern determined in SDS–PAGE (Fig. 1E–G). The much higher void-volume-peak detected in the case of EGCG treatment quite likely results from tight protein–drug interaction, as substantiated by our observations (data not shown).
EGCG, but not tetracycline, drastically affects the structural features of the aggregation intermediates
By FTIR spectroscopy, we investigated the structural features of AT3Q55 aggregation products arising in the presence of either compound. This technique provides insights into protein secondary structures through the analysis of the amide I band (1700–1600 cm−1 spectral region), due to the C=O stretching vibration of the peptide bond. Noteworthy, it also makes it possible to detect intermolecular β-sheets in protein aggregates (28,29). The second derivative spectrum of freshly purified AT3Q55 (25 µm) in the amide I region is reported in Figure 2A. This mathematical procedure allows disclosing the different absorption components, which in the derivative spectrum appear as negative peaks (30). The spectrum of the native protein (Fig. 2A) is dominated by the component at ≈1657 cm−1, assigned to α-helices and random coils, while the band at ≈1635 cm−1, along with the shoulder at ≈1690 cm−1, is due to intramolecular β-sheets, as previously discussed (31). We then collected spectra of AT3 incubated at 37°C in the presence of either drug. The complete incubation mixture was analyzed, i.e. without prior centrifugation. During the incubation of untreated protein, the component at ≈1635 cm−1 decreased in intensity, almost disappearing after 24 h. Instead, the component at ≈1657 cm−1 first slightly decreased and subsequently increased again. Moreover, after 6 h two new components at ≈1624 and ≈1694 cm−1 were seen in the spectra. Finally, at longer incubation times, a new component appeared at ≈1604 cm−1 (Fig. 2B). These spectral changes are in agreement with our previous FTIR study (31), whereby we demonstrated that the disappearance of the 1635 cm−1 component is representative of native intramolecular β-sheets disruption, while the appearance of the ≈1624 and ≈1694 cm−1 peaks is due to the formation of intermolecular β-sheets. We also could unambiguously assign the components at 1604 and at 1657 cm−1 in the final aggregates to ordered side-chain hydrogen bonding of expanded polyQ tracts, a hallmark of AT3 mature and SDS-insoluble amyloid fibrils (31).
Spectral changes observed upon incubation of AT3Q55 in the presence of EGCG at a 1 : 5 molar ratio were substantially different from those of the control sample (Fig. 2C). In particular, the ≈1635 cm−1 component, assigned to native β-sheets, strongly decreased immediately after EGCG addition, indicating that this compound induces unfolding/misfolding of the protein. Two shoulders at ≈1628 and ≈1691 cm−1 appeared in the spectra and increased in intensity only at longer times of incubation (Fig. 2C). The comparison of the spectra in Figure 2C and B demonstrates that EGCG induces the formation of intermolecular β-sheet structures to a much lesser extent compared with those observed in its absence. Noteworthy, the 1604 cm−1 peak, assigned to the glutamine side-chain hydrogen bond network, was not detected in the spectra of the AT3Q55-EGCG solution, although most protein in solution was found to be SDS-resistant (Fig. 1F). Under these conditions, only a very small protein pellet was obtained by centrifugation at 14 000g after 2 weeks of incubation, whose FTIR spectrum was very similar to that of SDS-insoluble fibrils (data not shown). It displayed, in particular, the peaks assigned to glutamine side-chain hydrogen bonding (31).
The spectral changes observed upon incubation of AT3Q55 in the presence of tetracycline at a 1 : 5 molar ratio (Fig. 2D) were, instead, very similar to those found for the untreated protein (Fig. 2B).
Incubation of AT3 in the presence of EGCG or tetracycline gives rise to aggregates displaying different morphologies
AT3Q55 solutions incubated in the presence or the absence of either compound were analyzed by tapping mode AFM to get insight into the morphology of the resulting aggregates. Representative images are reported in Figure 3. Bundles of fibrils were observed for AT3Q55 alone after 24 and 48 h. The height of these bundles was between 20 and 60 nm, in keeping with previous observations (31). Instead, in the presence of EGCG, no such bundles were detected. Actually, after 24 h the sample mainly consisted of globular particles of height between 2.0 and 3.5 nm, isolated or associated in small clusters. After 48 h, large clusters of non-fibrillar material were found. The cluster height was between 20 and 80 nm and the typical cluster size in the scan plane was 0.5–1 μm.
In contrast, tetracycline did not significantly affect the aggregation pattern. Indeed, fibrils morphologically indistinguishable from those observed in the control sample were found at both incubation times, although many irregular, compact and relatively flat aggregates with height of 26 ± 2 nm and irregular edges also appeared along with mature fibrils.
EGCG and tetracycline do not disrupt AT3Q55 preformed fibrils
To assess the effect of EGCG or tetracycline on preformed amyloid aggregates, we first produced AT3Q55 fibrils by incubating the protein at 37°C for 2 weeks in PBS buffer. The FTIR spectrum of the pellet of the resulting sample (Supplementary Material, Fig. S5A) displayed four main components: 1695 and 1624 cm−1 due to intermolecular β-sheets, 1657 and 1604 cm−1 assigned, respectively, to the C=O stretching and NH2 deformation modes of glutamine side chains involved in strong hydrogen bonding (31). The fibrils were resuspended in PBS alone (Supplementary Material, Fig. S5B), in PBS in the presence of EGCG (Supplementary Material, Fig. S5C) or tetracycline (Supplementary Material, Fig. S5D) at a molar ratio of 1 : 5 (protein compound). No significant spectral changes were observed in the three samples during a 1-week incubation at 37°C, indicating that EGCG and tetracycline are not able to redissolve mature AT3Q55 fibrils.
Both EGCG and tetracycline treatments reduce AT3Q55 cytotoxicity
We also examined the toxicity of AT3Q55 species formed in the presence or the absence of EGCG (Fig. 4A) or tetracycline (Fig. 4B) on the COS-7 cell line. AT3Q55 aliquots were added to the cellular medium and toxicity was assessed using the MTT assay (Fig. 4). When 3- or 6 h-preincubated AT3Q55 preparations were added to the cell cultures, MTT reduction dropped to ∼60% of the untreated cells. However, a statistically significant lower toxicity was determined when administering AT3Q55 aggregates generated in the presence of either EGCG (Fig. 4A) or tetracycline (Fig. 4B). The lack of any statistically significant differences between treated and not treated cells after a 24-h preincubation can be quite plausibly justified by the fact that oligomeric, toxic forms of amyloidogenic proteins, including AT3, evolve into non-toxic fibrils at later incubation times (31,32).
Transgenic worms expressing AT3 variants display a SCA3 phenotype
To evaluate the effect of the two compounds on AT3 toxicity in vivo, we used a SCA3 C. elegans model. A wild type (Q17) and an expanded form (Q130) of AT3 were expressed in the nervous system in fusion with GFP under the control of the pan neuronal promoter unc-119. The expression of the two proteins was monitored by confocal analysis using GFP fluorescence (Supplementary Material, Fig. S6). We observed that the fluorescence of both proteins was diffuse in the young animals (1 day; Supplementary Material, Fig. S6A) but in the older animals (4 days) a focal fluorescence distribution was detected in both nematode strains, although to a lesser extent in the AT3Q17-GFP-expressing worms (Supplementary Material, Fig. S6B). The amyloid nature of the aggregates formed only by the expanded protein was confirmed by X-34 staining (Supplementary Material, Fig. S6C).
We then evaluated the effect of AT3 expression on worm survival, which highlighted a significant reduction in lifespan compared with the wild-type strain (Fig. 5A; median survival: 6 days for N2, 4 days for AT3Q17-GFP and 3 days for AT3Q130-GFP strains). Since the lack of movement coordination is a hallmark of SCA3, we monitored the changes in mobility of the transgenic worms and determined the locomotion activity using the body bends assay (Fig. 5B). AT3Q130-GFP-expressing worms displayed an age-dependent uncoordinated movement, whereas older transgenic AT3Q17-GFP expressing worms showed a behavior similar to that of the N2 worms (data not shown). The quantification of locomotion revealed a significant reduction in the number of body bends per min in AT3Q130-GFP worms in all days tested. A lesser reduction was also observed in AT3Q17-GFP transgenic worms (Fig. 5B). This characterization confirms the robustness of the model and supports the goodness of the body bends assay to test the pharmacological effect of drugs.
EGCG and tetracycline protect against AT3 toxicity in C. elegans model
We performed the body bends assay on worms treated with 0.1 mm EGCG (Fig. 6A) or 0.1 mm tetracycline (Fig. 6B). After 24 and 48 h of EGCG treatment, AT3Q130-GFP-expressing worms displayed a statistically significant increase in the number of body bends, i.e. 30 and 25%, respectively, compared with untreated worms (Fig. 6A). In the presence of tetracycline, we also observed a statistically significant increase, by 48 and 23%, respectively (Fig. 6B). Both compounds did not elicit any effect in AT3Q17-GFP-expressing worms, except for the treatment with EGCG for 48 h (15%; Fig. 6A). N2 motility was not affected at any times. For both compounds, no significant differences in motility were detected at earlier and later times of treatment (data not shown).
The green tea polyphenol EGCG and the antibiotic tetracycline are attractive candidates for the treatment of neurodegenerative diseases for their well-established anti-amyloidogenic effect (33–35), proven safety record in humans and blood–brain barrier permeability (16,36). As no cure or suitable therapeutic compound is available to date for SCA3 treatment, this study was undertaken to verify possible anti-amyloidogenic activity of EGCG and tetracycline on AT3 aggregation.
As regards the effects that the polyphenol EGCG exerts on AT3Q55 aggregation kinetics, we observed that, starting from 24 h, it substantially retarded the disappearance of soluble matter from supernatants of protein solutions incubated under aggregative conditions. Although this observation points to the compound capability to contrast amyloidogenesis, we observed that it also induced a significantly faster decline in the SDS-soluble fraction of the supernatants, especially at the molar ratio 1 : 5 protein-EGCG (Fig. 1A and C). This, in turn, was paralleled by a substantial accumulation of soluble, SDS-resistant species, which were unable to enter the separating gel (Fig. 1E and F). This pattern can be represented quantitatively (Supplementary Material, Fig. S7A and B) as time courses of SDS-soluble, SDS-resistant soluble and insoluble (pelleted) species. This makes it apparent that the massive appearance of soluble, SDS-resistant species is a hallmark of EGCG treatment, as in its absence they are scanty, if any, at any time.
Our findings raise the question as to whether soluble, SDS-resistant species are on- or off-pathway intermediates. Aggregation kinetics support the latter hypothesis, as they do not evolve into insoluble material, either after 72 h or even after a 2-week incubation, as outlined below. This idea was even more reinforced by FTIR analyses that highlighted dramatic structural differences between soluble, SDS-resistant species appearing in the presence of EGCG, and pelleted final aggregation products that accumulate in its absence. In particular, the former differ from final insoluble aggregates in that they do not display the 1604 cm−1 peak, assigned to glutamine side-chain hydrogen bond network, and also are much poorer in intermolecular β-sheets (peaks at ≈1635 and 1690 cm−1). Thus, EGCG quite likely interferes at a very early step of the amyloid pathway, accelerating misfolding of the JD and redirecting the protein toward off-pathway aggregates, whose precise structural features have still to be defined in detail. Albeit not conclusively, our data support the idea that the drug is capable of binding monomeric rather than oligomeric AT3, as substantiated by the appearance of peculiar structural changes from the very beginning of the incubation (1 h), when the untreated protein has not yet undergone any appreciable structural modification. This plausibly suggests that EGCG directs AT3 towards the off-pathway by primarily acting on the monomeric protein. In keeping with this assumption, a previous report shows interaction of the drug with monomeric human serum albumin (37). Interestingly, another report shows that EGCG binds transthyretin tetramer, thus preventing its dissociation into monomers, which is the rate-limiting step for amyloid fibril formation (38). This finding also supports the view that the drug can interact with proteins before they undergo amyloid aggregation.
Consistent with the proposed aggregation pattern, AFM analysis confirmed substantial structural differences between the aggregates arisen in the presence and those in the absence of EGCG. In its presence, no mature fibrils were generated, but only larger spherical amorphous species, possibly resulting from oligomer clustering. Our results fit with the common hypothesis that EGCG prevents on-pathways, which lead to toxic amyloid oligomers and protofibrils (15). Actually, highly stable, off-pathway aggregates were assembled. This substantial change in the aggregation pattern quite likely underlies the protective effect of the drug we have detected on COS-7 cells by the MTT assay. The beneficial effect of the compound was also confirmed in the SCA3 C. elegans model, in which the EGCG treatment resulted in a significant increase in motility and improvement in locomotion in the diseased worms only.
Unlike EGCG, tetracycline did not apparently impact on the structural features of the aggregation intermediates, but drastically reduced the formation of insoluble (pelleted) aggregates, with scanty accumulation of soluble, SDS-resistant species (Supplementary Material, Fig. S7C). This was substantiated by both FTIR spectroscopy (Fig. 2D) and AFM analyses (Fig. 3). FTIR did not demonstrate substantial modifications in aggregation kinetics and secondary structure compared with the untreated protein; in addition, AFM confirmed the formation of mature amyloid fibrils along with irregular and compact aggregates. Thus, although tetracycline treatment also results in some accumulation, at the latest incubation times, of soluble, SDS-resistant species (Fig. 1G; Supplementary Material, Fig. S7C), these are quite likely on-pathway intermediates committed to evolve into final insoluble fibrils, with which they share internal structure, as assessed by FTIR.
Unlike the present study, previous reports show that tetracycline dramatically inhibits fibrillogenesis of a set of other amyloidogenic proteins, notably PrP and α-synuclein (21). In spite of the lack of such an effect, our treatment also leads to a significant reduction in toxicity, as shown in the COS-7 cell line and the SCA3 C. elegans model. We suggest that the drug might bind to the surface of the growing oligomers and fibrils, with resulting substantial increase in their solubility (Fig. 1B, D and G; Supplementary Material, Fig. S5C), but with no major changes in their structural features (Figs 2D and 3). This, in turn, might also explain how the drug can substantially prevent the toxic effects of the oligomeric species, i.e. by masking the exposed hydrophobic patches. This does not rule out, of course, that subtle structural changes in the AT3 aggregates, undetectable by our analytical methods, may also occur.
Although previous reports show that EGCG and tetracycline are capable of remodeling and redissolving mature amyloid fibrils of different proteins (13,25), our FTIR results clearly demonstrate that these drugs do not affect the secondary structures of the AT3 mature fibrils (Supplementary Material, Fig. S5). This might be plausibly accounted for by the presence of glutamine side-chain hydrogen bonding that cooperatively contributes to the stability and irreversible aggregation of the SDS-insoluble polyQ mature fibers (31). Based on these observations, we conclude that the remodeling action does not represent an essential prerequisite for the protective effect to be exerted.
Our findings might also help better clarify previous results showing the drug protective effect towards a set of amyloid-generating proteins (39), considering in particular another polyQ-containing peptide, i.e. huntingtin exon-1 (15).
We are currently carrying on further structural analyses by NMR spectroscopy to precisely elucidate the mode of interaction between AT3 and either compound, including the functional groups involved, thus clarifying their different mechanisms of action. Finally, our work also confirms that our SCA3 C. elegans model proves to be a suitable tool for assessing a drug's capability of counteracting AT3 toxicity in living organisms.
MATERIALS AND METHODS
AT3Q55 gene was previously cloned in pQE30 vector and the protein was expressed in SG13009 (Escherichia coli K12 Nals, StrS, RifS, Thi−, Lac−, Ara+, Gal+, Mtl−, F−, RecA+, Uvr+, Lon+; Qiagen Hamburg GmbH, Hamburg, Germany) as His-tagged protein (31). Cells were grown at 37°C in 2TY–ampicillin–kanamycin medium, induced with 1 mm IPTG at OD600 0.8 for 45 min at 30°C. To obtain crude extract, pelleted cells were resuspended in lysis buffer (5 ml/g wet weight; 25 mm potassium phosphate, pH 7.2, 150 mm NaCl, 0.5 mm phenylmethanesulfonyl fluoride, 10 mm imidazole, 10% glycerol, 1 mm 2-mercaptoethanol, 1 mg/ml lysozyme plus protease inhibitors cocktail) and incubated for 30 min at 4°C. The cell suspension was then sonicated in three pulses of 30 s each. DNase I (0.2 mg/g of cells, wet weight) was added, and the sample further incubated for 30 min at room temperature. Finally, it was centrifuged for 45 min at 20 000g. The supernatant was filtered through a 0.45-μm pore size SFCA membrane (Corning), loaded onto HisPur™ Cobalt Resin (Thermo Fisher Scientific, Rockford, IL, USA) and washed with 20 bed volumes of wash buffer (25 mm potassium phosphate, pH 7.2, 150 mm NaCl, 2 mm phenylmethanesulfonyl fluoride, 10 mm imidazole, 10% glycerol, 1 mm 2-mercaptoethanol). The bound protein was then eluted with elution buffer (25 mm potassium phosphate, pH 7.2, 150 mm NaCl, 2 mm phenylmethanesulfonyl fluoride, 150 mm imidazole, 10% glycerol, 1 mm 2-mercaptoethanol). Protein was stored at −80°C. Before each experiment, protein fractions were thawed and loaded onto a Superose 12 10/300 GL gel filtration column (GE Healthcare, Life Sciences, Little Chalfont, UK), pre-equilibrated with PBS buffer (25 mm potassium phosphate, pH 7.2, 150 mm NaCl). Elution was performed at a flow rate of 0.5 ml/min in the same buffer. Fractions were collected and protein content determined using Coomassie brilliant blue G-250 (Thermo Fisher Scientific) and bovine serum albumin as a standard protein.
SDS–PAGE and densitometry analysis of soluble protein fraction
Purified AT3Q55 (25 μm) was incubated at 37°C in PBS buffer in the presence or the absence of EGCG or tetracycline (Sigma-Aldrich, St Louis, MO, USA) at a protein–compound molar ratio of 1 : 1 or 1 : 5. AT3 aliquots at different times of incubation (0, 3, 6, 24, 48 and 72 h) were centrifuged at 14 000g for 15 min and 10 μl of the supernatants were subjected to SDS–PAGE. The gels were stained with Imperial Protein Stain (Thermo Fisher Scientific), scanned at 700 nm with Odyssey® Fc System (LiCor, Lincoln, NE, USA) and analyzed with Image Studio software (LiCor).
For FTIR analyses, purified AT3Q55 (25 μm, corresponding to ≈1 mg/ml) was incubated at 37°C in PBS buffer (25 mm potassium phosphate, pH 7.2, 150 mm NaCl) in the presence or the absence of EGCG or tetracycline at a protein–compound molar ratio 1 : 5. FTIR measurements of the protein semi-dry films were performed in attenuated total reflection (ATR) (40) as previously described (31,41,42). Briefly, a 3-μl aliquot of the samples at different times of incubation (0, 1, 3, 6, 24, 30, 48, 144 h and 2 weeks) was deposited on the single reflection diamond ATR plate (Golden Gate, CA, USA) and dried at room temperature so as to obtain a semi-dry protein film. ATR/FTIR spectra were measured using a Varian 670-IR spectrometer (Varian Australia Pty Ltd, Mulgrave, VIC, Australia) equipped with a nitrogen-cooled mercury cadmium telluride detector under the following conditions: 2 cm−1 spectral resolution, 25 kHz scan speed and 1000 scans co-addition and triangular apodization. The ATR–FTIR spectra of PBS and of the two compounds at 125 µm in PBS were also collected at each incubation time at 37°C. The AT3Q55 spectra were obtained by subtraction of the proper reference spectra (Supplementary Material, Fig. S8). Second derivatives of the spectra were obtained by the Savitzky-Golay algorithm (5 points), after an 11-point binomial smoothing of the measured spectra, using the software Grams/AI (Thermogalactic, MA, USA). In a control experiment, the FTIR spectrum of freshly purified AT3Q55 in form of a semi-dry film was also measured in the transmission mode by an infrared microscope. In particular, a 3-µl aliquot of AT3Q55 in PBS was deposited on a BaF2 window and dried at room temperature. The microFTIR absorption spectrum was acquired in the transmission mode using the Varian 610-IR infrared microscope coupled to a Varian 670-IR spectrometer (43,44). The second derivative of the absorption spectrum collected in the transmission mode (data not shown) displayed the same spectral features of that collected in the ATR mode, indicating that the ATR spectra are not affected by the interaction of the protein with the diamond surface of the ATR device.
Atomic force microscopy
AT3Q55 was purified by gel filtration, frozen at −80°C and thawed before the AFM experiments. AT3Q55 (25 μM) was incubated at 37°C in PBS buffer in the presence or absence of EGCG and tetracycline at a molar ratio protein–compound of 1 : 5. At fixed aggregation times, a 10-μl aliquot was withdrawn, incubated on a freshly cleaved mica substrate for 5 min, then rinsed with Milli-Q water and dried under mild vacuum. AFM images were acquired in tapping mode in air using a Dimension 3100 Scanning Probe Microscope equipped with a ‘G’ scanning head (maximum scan size 100 μm) and driven by a Nanoscope IIIa controller, and a Multimode Scanning Probe Microscope equipped with ‘E’ scanning head (maximum scan size 10 μm) and driven by a Nanoscope IVcontroller (Digital Instruments—Bruker). Single beam uncoated silicon cantilevers (type OMCL-AC160TS, Olympus) were used. The drive frequency was between 270 and 330 kHz, the scan rate between 0.5 and 0.8 Hz.
COS-7 cells were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum, 100 U/ml penicillin, 100 μg/ml streptomycin and 4 mml-glutamine, maintained at 37°C in a humidified 5% CO2 incubator. For MTT assays, cells were trypsinized and plated at a density of 10 000 cells per well on 96-well plates in 100 μl fresh medium without phenol red. After 24 h, 25 μm AT3Q55 alone or co-incubated with the two compounds (1 : 1 and 1 : 5 molar ratio) at different times (3, 6 and 24 h) was added to the cell medium at a final concentration of 2.5 μm and cells were further incubated for 1 h at 37°C. Then MTT was added to the cells at a final concentration of 0.5 mg/ml. Absorbance values of formazan were determined at 570 nm with an automatic plate reader after 2 h.
Caenorhabditis elegans strains
AT3Q17 (wild type) and AT3Q130 (pathological form) cDNAs, cloned in pPD95.77 plasmid in frame with GFP, under control of pan neural promoter unc-119, were kindly provided by Dr Nobuyuki Nukina (45). The recombinant DNAs were injected at concentration of 5–20 ng/ml into the lin-15(n765ts) strain together with the LIN-15 rescuing plasmid. A wild-type Bristol N2 strain was used as control. All strains were grown at 25°C on solid nematode growth medium (NGM) seeded with E. coli (OP50) for food according to standard procedures (46). At least three independent lines for each construct were tested for the phenotype.
Worms age synchronization
To prepare age-synchronized animals, a small plate (3 ml) of nematodes was transferred onto a new large plate (25 ml) with fresh NGM agar seeded with E. coli (OP50) to obtain many eggs. After 2 or 3 days of incubation at 25°C, the population was collected in 2.5 ml of M9 buffer (42 mm Na2HPO4, 22 mm KH2PO4, 86 mm NaCl and 1 mm MgSO4) and an equal volume of 4% glutaraldehyde was added. The suspension was incubated for 4 h at 4°C and, after a brief centrifugation, washed twice in M9 buffer. Half sector of a 25 ml plate was seeded with E. coli. Collected worms were put down to uninoculated sector of plate and incubated at 25°C (46).
One day-synchronized adult worms were isolated and placed daily onto a fresh plate seeded with E. coli at 25°C. Surviving and dead animals were counted daily until all worms had died. The test was performed on 40 animals for each strain.
Body bends frequency test
One day-synchronized adult worms were placed onto a new plate and body bends per minute were counted under a microscope (Leica MZ FLIII, Leica Microsystem). The test was performed on 20 animals.
One day-synchronized adult worms were isolated and placed onto a fresh plate in the presence or the absence of 0.1 mm EGCG or tetracycline. The body bends test was performed after 24 and 48 h. For each treatment, 15 animals were used.
All experiments were done at least in triplicate. Data are presented as means ± SE. P-values were calculated using the Student's t-test. (*P < 0.05, **P < 0.01). For the life-span assay, animal survival was plotted using Kaplan–Meier survival curves. Significant differences at the P < 0.05 level were calculated by one-way ANOVA.
This work was supported by grants from Regione Lombardia, Italy (Network-Enabled Drug Design), Fondazione Cariplo, Italy (progetto Nobel: Network Operativo per la Biomedicina di Eccellenza in Lombardia), University of Milano-Bicocca (Fondo di Ateneo per la Ricerca) and University of Genoa (Fondi di Ateneo per la Ricerca).
Conflict of Interest statement. None declared.
We thank Dr Nobuyuki Nukina for the gift of pPD95.77 plasmids and the Caenorhabditis elegans Genetics Centre for providing the N2 Bristol strain. We also thank Dr Anna Maria Villa for technical support in confocal microscopy analysis and Dr Diletta Ami for preliminary characterization of transgenic nematodes.