Inhibition of Amyloid Peptide Fragment Aβ25–35 Fibrillogenesis and Toxicity by N-Terminal β-Amino Acid-Containing Esapeptides: Is Taurine Moiety Essential for In Vivo Effects?

Authors


Corresponding author: Cesare Giordano, cesare.giordano@uniroma1.it

Abstract

We report the synthesis and fibrillogenesis inhibiting activity of the new peptide derivatives 1–4, related to the pentapeptide Ac-LPFFD-NH2 (iAβ5p), proposed by Soto and co-workers and widely recognized as one of the most active β-sheet breaker agents. The Aβ25–35 fragment of the parent full-length Aβ1–42 was used as fibrillogenesis model. The activity of peptide derivatives 1–4 was tested in vitro by thioflavin T binding assay, far UV CD spectroscopy, and scanning electron microscopy. Their ability to hinder the toxic effect of Aβ25–35in vivo was studied by monitoring the viability of human SH-SY5Y neuroblastoma cells and the prevention of superoxide anion radical release from BV2 microglial cells. The results point to a favourable role in the fibrillogenesis inhibitory activity of the sulphonamide junction for compounds 1 and 2, containing an N,N-dimethyltaurine and a taurine amino-terminal moiety, respectively. Furthermore, compounds 1 and 2 show a significant protective effect on cell viability, rescuing the cells from the toxicity exerted by Aβ25–35 treatment.

Alzheimer’s disease, a progressive degenerative form of dementia in adults, belongs to a class of diseases known as amyloidoses whose common feature is the extracellular cerebral accumulation of amyloid β-peptides (Aβ) (1–3). The main components of Aβ, normally constituted by 39–43 amino acids, are produced by cleavage of a larger amyloid precursor protein, and their self-association is consequent to a conformational switch from random coil or α-helical to highly toxic β-sheet structure, responsible for the amyloidogenesis and Aβ aggregates formation (1,4,5). In this context, the inhibition of amyloid fibril formation by using compounds able to specifically stabilize the α-helical (4) and/or destabilize the β-sheet conformations by a direct interaction with the Aβ has been largely suggested and developed as pharmacological strategy (6,7). The observation that short synthetic peptides, named β-sheet breaker peptides (BSBp), are able to establish an interaction, through a mechanism not yet fully understood, with soluble oligomers or amyloid aggregates, destabilizing the amyloidogenic conformer and precluding amyloid polymerization (8), prompted us to undertake a study on novel BSBp derivatives.

We have recently published a report on six new peptide analogues (9) related to the synthetic five-residue BSBp derivative Ac-Leu-Pro-Phe-Phe-Asp-NH2 (Ac-LPFFD-NH2; iAβ5p), previously adopted by Soto and coworkers (10) as lead compound, on which several strategic chemical modifications have been applied to improve its activity and metabolic stability (3). In the previous work (9), we found that the esapeptide N,N-dimethyl-Tau-Leu-Pro-Phe-Phe-Asp-NH2 (1; Figure 1), in which an N,N-dimethyl-taurine (DM-Tau) substitutes the acetyl group of iAβ5p, demonstrated a higher β-sheet breaker activity on Aβ1–40 than iAβ5p and had an increased resistance to proteolysis. Prompted by this finding and in continuation of our efforts in obtaining new potential BSBp, we synthesized and tested the activity of the three new analogues 2–4 (Figure 1) with the main purpose to elucidate the influence of the terminal N,N-dimethyl group and/or of the sulphonamide junction both present in peptide 1.

Figure 1.

 Schematic representation of β-sheet breaker peptides 1–4.

Essentially, derivatives 2–4 maintain the original sequence of 1, with the exception of substitutions at DM-Tau residue: a taurine (Tau), an N,N-dimethyl-β-alanine (DM-βAla) and a β-alanine (βAla) were incorporated in compounds 2, 3 and 4, respectively.

In this study, the Aβ25–35-fragment (Aβ25–35) of Aβ1–42 was chosen as model for fibrillogenesis. Aβ25–35, sequence GSNKGAIIGLM, is reported as the biologically active region of Aβ1–42 being the shortest fragment among the Aβ derivatives that contains a high hydrophobic region and forms consistent β-sheet aggregates retaining the toxicity of the parent full-length peptide (1,11,12). Interestingly, it has been reported that Aβ25–35 is formed in vivo and that it is localized to plaque cores and in vascular amyloid deposits (13). It has been hypothesized that Aβ25–35 is the true toxic species (14). In addition, the short length of Aβ25–35 allows a more ready and affordable synthesis (15).

Peptide 1 and the new analogues 2–4 were tested in vitro on Aβ25–35 as inhibitors of fibrillogenesis using thioflavin T (ThT) binding assay, far UV CD spectroscopy and scanning electron microscopy (SEM), while their efficacy in vivo was studied monitoring their ability to counteract the toxic effect on neuronal cells in culture and in preventing the release of ROS from cultured microglial cells.

Experimental Section

Peptide synthesis

The synthesis of peptide 1 was performed in solution as previously described (9). Aβ25–35 and peptide analogues 2–4 were prepared manually by conventional solid phase chemistry (16) on Fmoc-Met-Wang resin (150 mmol scale, 250 mg) and Rink-amide resin (65 mmol scale, 100 mg), respectively. Fmoc-AA-OH, DIC and HOBt (3 eq. respectively) in N-methyl-2-pyrrolidone were used for couplings. The Fmoc-taurylsulfonyl chloride (Fmoc-Tau-Cl), required for the preparation of 2, was obtained according to Brouwer et al. (17) and coupled to the resin-bound protected pentapeptide in dichlromethane in the presence of 4-methylmorpholine (7 eq.). For the synthesis of 3, DM-βAla hydrochloride was prepared (18) and coupled following usual protocol. The reaction mixtures were shaken in a mechanical shaker for 3 h at room temperature. The completion of the coupling reaction was checked by ninhydrin-Kaiser test, and the removal of Fmoc was effected with a piperidine (20%) solution in DMF. For the deprotection of final peptides and cleavage from resin, a solution of TFA/EDT/TA/H2O/TIS (75:10:10:4:1) was used and the peptides were precipitated from the cleavage solution with peroxide-free dry diethyl ether at 0 °C. The precipitate was repeatedly washed with ether, redissolved with water and lyophilized. The crude peptides were purified by RP-HPLC on a Waters 600E liquid chromatography system by using a Waters Bondapac C-18 column (1.9 × 30 cm, 5 μm, 300 Å) for semipreparative scale with elution at 8 mL/min by a linear gradient of 10–60% acetonitrile in 0.1% aqueous trifluoroacetic acid in 30 min. Peptide purity was >97% by analytical HPLC (Waters μ-Bondapack C-18 column, 0.39 × 30 cm, 5 μm, 300 Å).

The peptides were characterized on a Q-TOFMICRO spectrometer (Micromass, now Waters, Manchester, UK) equipped with an ESI source, in the positive ion mode and data were analysed using the MassLynx software (Waters). Mass spectrometry: found [M + H]+ (calcd [M + H]+, ppm deviation); Aβ25–35, 1061.2615 (1061.2665, 4.7); 1, 772.3649 (772.3625, −3.1); 2, 744.3351 (744.3312, −5.2); 3, 736.3937 (736.3956, 2.6); 4, 708.3670 (708.3643, −3.8).

Thioflavin T binding assay

The assembly of Αβ25–35 was monitored using a ThT binding assay (19). Thioflavin T fluorescence intensity was monitored continuously in a Perkin Elmer LS 50 B spectrofluorimeter, thermostated at 20 °C, with excitation and emission wavelengths set at 450 and 484 nm, respectively. Fluorescence changes were monitored in a 1 mL cuvette, under continuous stirring, at 484 nm every 6 seconds, with 5 seconds integration time over at least 13 200 seconds. One millimolar ThT stock solution in water was prepared, aliquoted and stored at −20 °C. Before each experiment, an aliquot of ThT stock solution was thawed at 4 °C shielded from light. A solution of 200 μm lyophilized Αβ25–35 in 20 mm Tris–HCl pH 7.4 (1 mL final volume), alone or in the presence of equimolar 1–4 or iAβ5p, was transferred to a 1 mL cuvette, and 10 μL of 1 mm ThT was added. Dead time for measurements of fluorescence changes did not exceed 12 seconds.

Far UV CD spectroscopy

Far UV (190–250 nm) CD measurements were performed with a Jasco J-720 spectropolarimeter (Jasco, Easton, MD, USA), in a 0.05 cm path length quartz cuvette with detachable windows, thermostated at 20 °C. The results are expressed as the mean residue ellipticity [Θ] assuming a mean residue weight of 110 per amino acid residue. Aliquots of lyophilized Αβ25–35 were dissolved in 20 mm Tris–HCl, pH 7.4, at 200 μm final concentration in the absence or in the presence of equimolar BSBp and the far UV CD spectral changes, occurring upon fibrils formation, were monitored at 20 °C, at time 0, 1, 3, 6 and 24 h.

Data analysis

Far UV CD spectra of Aβ25−35 in the presence or absence of the synthetic fibrillogenesis inhibitors were analysed by singular value decomposition algorithm (SVD) (20–22) using the MATLAB (MathWorks, South Natick, MA, USA) software. Singular value decomposition algorithm is useful to find the number of independent components in a set of spectra and to remove the high-frequency noise and the low-frequency random error. CD spectra at increasing incubation time in the 190–250 nm region (0.5 nm sampling interval) were placed in a rectangular matrix A of n columns, one column for each spectrum collected at the selected incubation time, ranging from 0 to 24 h. The A matrix is decomposed by SVD into the product of three matrices: A = USVT, where U and V are orthogonal matrices and S is a diagonal matrix. The columns of U matrix contain the basis spectra and the columns of the V matrix contain the time dependence of each basis spectrum. Both U and V columns are arranged in terms of decreasing order of the relative weight of information, as indicated by the magnitude of the singular values in S. The diagonal S matrix contains the singular values that quantify the relative importance of each vector in U and V.

Preparation of Aβ25–35 stock solutions

25–35 (10 mg) was dissolved in 35 mL of ultrapure water (UHQ Sartorius Arium® 611 UV; Sartorius, Göttingen, Germany) and the peptide concentration determined by UV-VIS spectroscopy in a Lambda 16 Perkin Elmer spectrophotometer using an εM at 214 nm of 923/m/cm (23). After dilution to 200 μm with ultrapure water, the Aβ25–35 solution was divided into aliquots, frozen at −20 °C, lyophilized and stored at −20 °C until use.

Scanning electron microscopy studies

Morphologies of Aβ25–35 were investigated using a scanning electron microscope LEO 1450VP with tungsten filament as electrons emitter with 20 keV acceleration potential and a resolution of 4 nm. For the SEM study, aliquots of 600 μm25–35 incubated at 20 °C in 30 mm Tris–HCl, pH 7.4 in the absence or in the presence of 1.2 mm1–4 were withdrawn after 24 h, frozen, lyophilized and gold coated.

Cell culture

Human dopaminergic neuroblastoma cell line SH-SY5Y was obtained by the Cell Bank ICLC (Genova, Italy). Tissue culture reagents were from Gibco (Invitrogen, Monza, Italy). The SH-SY5Y neuroblastoma cells were maintained in a humidified incubator under 5% CO2 at 37 °C. Cells were grown in Dulbecco’s modified Eagle’s/F12 medium supplemented with 10% heat-inactivated foetal bovine serum and 2 mm glutamine. The cells were routinely harvested twice a week by trypsinization (0.05% trypsin-EDTA) and plated in 25 cm2 culture flasks (split 1:4–1:6).

The BV2 microglial cells were grown and maintained in Dulbecco’s modified Eagle medium (D-MEM) supplemented with 10% heat-inactivated foetal bovine serum, 4 mm glutamine and 50 μg/mL gentamicin at 37 °C under a humidified atmosphere of 5% CO2/95% air. For superoxide assay experiments, 2 days before the treatment, BV2 cells were seeded onto 6-well plates (350 000 cells/well). Before being harvested by scraping in 1 mL/well of D-MEM without phenol red, cells were washed once with 2 mL/well of same medium. A cell suspension containing 500 000 cells/mL was used for spectrophotometric measurements.

Cells viability assays

25–35 was dissolved in sterile PBS at a final concentration of 1 mm and incubated overnight at 37 °C, in the presence or absence of the BSBp (molar ratio 1:2) and then 100-fold diluted with culture medium immediately before performing viability experiments and superoxide assay. Control experiments were performed on the solvent alone and on each BSBp 1–4 at the same final concentration used for aggregation inhibition tests.

Cell viability was evaluated as follows: the cells were seeded in 96-well plates in 100 μL medium at a density of 15 000 cells/well and allowed to grow for 24 h, then the medium was replaced with that containing 10 μm25–35 alone or Aβ25–35 preincubated with 20 μm1–4. After incubation for up to 72 h, 20 μL of MTT reagent (5 mg/mL in PBS) was added to each well and incubated for 2 h at 37 °C. The medium was then discarded, and the resulting insoluble formazan was extracted with isopropanol. The absorbance was measured with a spectrophotometric microplate reader at 570 nm, with a reference at 660 nm. Absorbance of wells without cells was also measured and subtracted as background from each sample.

For superoxide assay, the amount of superoxide produced in microglia was continuously determined in a double beam spectrophotometer by measuring at 37 °C, for 30 min, the superoxide dismutase (SOD)-inhibitable reduction of cytochrome c at 550 nm, according to (24). Briefly, 10 μm25–35 alone or in the presence of 1–4 was added under continuous stirring to the cuvettes containing 500 000 cells/mL, 4 × 10−5 m cytochrome c and 0.5 mm CaCl2. SOD (100 U/mL) was added only to the reference. It was assumed that an 1.0 AU absorbance change at 550 nm corresponds to 47.4 nmols of reduced cytochrome c. Experiments were repeated at least three times and were congruent.

All the results are expressed as the mean values ± SD. Student’s t-test was used for groups’ statistical comparison and p-values <0.05 were regarded as significant.

Results and Discussion

Thioflavin T binding assay

As shown in Figure 2, the relative fluorescence intensity at 484 nm increased as a function of time in the presence of 200 μm25–35. When Aβ25–35 was dissolved with equimolar 1–4 or iAβ5p, the increase in ThT fluorescence was lower than that detected in the presence of Aβ25–35 alone (Figure 2). In particular, the amplitude of the fluorescence changes measured for 13 200 seconds progressively decreased from 3, 4, iAβ5p, 1 and 2. The amplitudes observed, compared with that of iAβ5p, were 1.4- and 1.8-fold lower for 1 and 2, respectively, and 1.2-fold higher for 4, whereas for 3 was closely similar to that observed for Aβ25–35 alone.

Figure 2.

 Thioflavin T binding assay for Aβ25–35. The effect of equimolar 1–4 on 200 μm25–35 assembly was monitored at 20 °C in 20 mm Tris–HCl, pH 7.4 in the presence of 10 μm thioflavine T.

In the absence of Aβ25–35, the fluorescence intensity at 484 nm of ThT alone was identical to that measured in the presence of any of the BSBp tested and did not change over 13 200 seconds.

Far UV CD spectroscopy

The in vitro inhibitory activity of 14 was tested by monitoring their efficacy in preventing the secondary structure conformational transition of Aβ25–35 preliminary to the formation of fibrils and aggregation. Conformational studies on the synthetic lyophilized Aβ25–35 have been performed by monitoring the far UV CD spectroscopic changes as a function of time in the absence or in the presence of equimolar 1–4. Typically, 200 μm lyophilized Aβ25–35 undergoes to the far UV CD spectral changes reported in Figure 3.

Figure 3.

 Far UV CD spectra of Aβ25–35. Aliquots of lyophilized Aβ25–35 were dissolved in 20 mm Tris–HCl, pH 7.4, at 200 μm final concentration, and the far UV CD spectral changes were monitored at 20 °C, at time 0, 1, 3, 6 and 24 h. The inset shows the far UV CD spectrum of the freshly prepared Aβ25–35, dissolved in water prior to lyophilization.

During 24 h the spectral changes observed by far UV CD at pH 7.4 in 20 mm Tris–HCl can be summarized in two main events: an overall decrease in molar ellipticity mainly at around 206–208 nm, a typical turns-containing structure region (25), followed by a blue-shift of the zero intercept, and then a further decrease at around 218 nm that parallels a visible solubility loss. Far UV CD spectrum of the freshly prepared Aβ25–35, dissolved in water prior to lyophilization, shows a strong negative band near 200 nm (Figure 3, inset) and a weak band between 220 and 230 nm, typical of a random coil-rich structure (26) and, differently than the lyophilized Aβ25–35, does not change significantly over the 24 h incubation time: in this conformation, the Aβ25–35 peptide is less prone to form fibrils (15).

The far UV CD spectra of the lyophilized Aβ25–35 (Figure 4) in the presence of equimolar 14 are all substantially different from the spectrum of the Aβ25–35 alone, as they show a negative band near 200 nm and do not significantly change from 0 to 24 h incubation time, suggesting a possible interaction between 1–4 and Aβ25–35. In particular, at time zero, the far UV CD spectra of Aβ25–35 in the presence of equimolar 14 are all characterized by the lack of a zero intercept and by a strong negative contribution at around 200 nm, suggesting the absence of β-sheet structure. In the presence of 2, the modest negative contribution at around 216 nm observed at time 0 decreases at increasing incubation time.

Figure 4.

 Far UV CD spectra of Aβ25–35 alone or in the presence of β-sheet breaker peptides (BSBp). Aliquots of lyophilized Aβ25–35 were dissolved in 20 mm Tris–HCl, pH 7.4, at 200 μm final concentration in the absence or in the presence of equimolar 14. Far UV CD spectra were monitored at 20 °C at time 0, 1, 3, 6 and 24 h.

The far UV CD spectra of 1–4 in the absence of Aβ25–35 were also monitored and did not change from 0 to 24 h (data not shown). The interaction between Aβ25–35 and 1–4 is suggested by the significant difference between the spectra of Aβ25–35 in the presence of 1–4 (Figure 4) and the calculated sum of individual spectra of free Aβ25–35 with free 1–4. Figure 5 shows the calculated sum of the far UV CD spectrum of free Aβ25–35 with free 1 at increasing time.

Figure 5.

 Calculated sum far UV CD spectra of Aβ25–35 and 1. Far UV CD spectra of 200 μm25–35 and 200 μm1 in 20 mm Tris–HCl, pH 7.4, were recorded separately at time 0, 1, 3, 6 and 24 h at 20 °C and then summed. The inset shows the far UV CD spectra of lyophilized Aβ25–35 dissolved in 20 mm Tris–HCl, pH 7.4, at 200 μm final concentration containing equimolar 1. The data in the inset are the same as those reported in Figure 3.

The considerable difference between the calculated sum spectra of free Aβ25–35 with free 1 (Figure 5) and the measured spectra of the 1–1 complex reported in the inset of Figure 5, strongly suggests an interaction between Aβ25–35 and 1. Notably, the far UV CD spectrum of Aβ25–35 in complex with 1 (Figure 5, inset), differently from that of free Aβ25–35 (Figure 3), does not show the negative minimum at around 216 nm or the positive ellipticity around 200 nm, typical of a high β-sheet content. Similar results were obtained from the comparison of the far UV CD spectra of Aβ25–35 in complex with 2–4 (Figure 4) with the calculated sum spectra of free Aβ25–35 with free 2–4 (data not shown).

Scanning electron microscopy

The effect of inhibitors of Aβ25–35 fibrillogenesis was studied also by SEM. Figure 6 shows SEM images of Aβ25–35 after 24 h incubation at 20 °C in the absence (A, B) or in the presence (C, D) of 2.

Figure 6.

 Effect on Aβ25–35 aggregation. Scanning electron microscopy (SEM) images at two different magnifications of Aβ25–35 (600 μm in 30 mm Tris–HCl, pH 7.4) incubated at 20 °C alone (A,B) or with 2-molar excess of 2 (C,D). Five milliliter aliquots were withdrawn after 24 h, frozen, lyophilized and gold coated, by sputtering, prior to SEM analyses.

In the absence of the inhibitor (Figure 6A,B), Aβ25–35 sample shows the formation of protofibrils <1 μm in length (27,28). In the presence of 2 (Figure 6C,D), the protofibrils as well as any kind of fibrillar structure are absent; the few globular structures visible do not show any morphological resemblance of amyloid aggregates and are presumably formed by 2, as suggested by SEM images of 2 alone (data not shown).

Cell viability studies

Human SH-SY5Y neuroblastoma cells were treated with Aβ25–35 previously incubated overnight in the absence or in the presence of 1–4. As shown in Figure 7, after 24 h cell viability was reduced by Aβ25–35 to the extent of about 70% (p < 0.01), similar to what previously reported (29,30).

Figure 7.

 Comparison of cytoprotective effects of β-sheet breaker peptides (BSBp) 1–4. SH-SY5Y cells were seeded in 96-well microplates and grown for 24 h, then exposed to 10 μm25–35 previously incubated overnight alone or in the presence of 20 μm1–4. Cell viability was assessed by the MTT method after 24 h of incubation. Data are means ± SD. **p < 0.01 versus Control, *p < 0.05 versus Control and §p < 0.05 versus Aβ25–35.

Incubation of Aβ25–35 in the presence of 1 and 2 rescued the cells from the toxic effects of Aβ25–35, suggesting a cytoprotective effect for DM-Tau and Tau containing peptides. A statistically significant protective effect on cell viability was observed for both peptides, with p-values <0.05 versus Aβ25–35 treated cells (Figure 7). Peptides 3 and 4 showed no cytoprotective activity and remained significantly different than control (p < 0.05), although for 4 cell viability was slightly improved compared to Aβ25–35 treated cells. None of the BSBp in the range 10–50 μm showed toxicity toward SH-SY5Y cells, and cell viability was unaffected up to 48 h treatment (data not shown).

Superoxide assay on microglia

Microglia is known to mediate Aβ25–35-induced neurotoxicity releasing different proinflammatory factors and free radicals, such as superoxide anion (31), which can be determined by measuring the (SOD)-inhibitable reduction in cytochrome c at 550 nm (data not shown). To evaluate the protective effect of BSBp against the Aβ25–35-induced superoxide anion release, microglia was treated with Aβ25–35 alone or previously incubated with 1–4. The decrease in Aβ25–35-induced superoxide anion release was 85%, 50%, 40% and 25% in the presence of 1, 2, 4 and 3, respectively.

These findings point to the role that the sulphonamide junction can play in terms of both metabolic stability and alteration on usual H-bond patterns (32). In addition, the cytoprotective effect of 1 and 2 could be attributed to the capability of the taurine moiety to interfere with the Aβ aggregation (33,34), with the oxidative stress regulation (35,36) as well as with the membrane stabilization through a direct interaction with phospholipids (37).

Conclusion

In summary, the reported results demonstrate the in vivo protective effect of the taurine-containing peptides and confirm the capability of its analogues to hinder fibrillogenesis in vitro. The efficacy of BSBp 14 underlines the usefulness of modification performed at the N-terminal position of the peptide derivatives related to iAβ5p. With regard to the better efficacy of 1 and 2, it may be related to their enhanced binding capability because of the increased acidity of the sulphonamide junction that makes it a better hydrogen bond donor compared to the peptide amide moiety present in 3 and 4. The sulphonamide junction, which is completely stable to proteases and possesses two negatively charged oxygen atoms directly bound to the tetrahedral sulphur, can significantly alter polarity and H-bonding patterns of the native counterparts. For the same reasons, the sulphonamide junction in 1 and 2 may play a key role in the cytoprotective action against the amyloid toxicity in vivo.

Acknowledgment

The authors wish to thank Dr. Cinzia Fabrizi (Department of Anatomical Sciences, Histology, Forensic and the Locomotor Apparatus, Sapienza University, Rome, Italy) who kindly provided the BV2 murine microglial cell line, Dr. Daniela Ferro (CNR) for SEM analysis and Dr. Pasqualina Punzi (Department of Chemistry, Sapienza University, Rome, Italy) for Mass Spectra.

Ancillary