Interactions of Aβ_1-42 Peptide and Its Three Fragments (Aβ_8-12, Aβ_8-13, and Aβ_5-16) with Selected Nonsteroidal Drugs and Compounds of Natural Origin

Abstract

In the following paper, we present the results of our studies on the interactions of the Aβ_1-42 peptide and its three short fragments, namely Aβ_5-16 (RHDSGYEVHHQK; HZ1), Aβ_8-13 (SGYEVH; HZ2), and Aβ_8-12 (SGYEV; HZ3) with selected painkillers (ibuprofen and aspirin) and compounds of natural origin (anabasine and epinephrine). Steady-state fluorescence spectroscopy was used to study the binding properties of the selected systems. Additionally, based on molecular dynamics (MD) calculations supported by NMR-derived restrains, we have proposed the most likely area of the interactions of Aβ_1-42 and Aβ_5-16 peptides with the investigated compounds. The influence of symmetrically oriented side chains of amino acid residues present in the first part of the Aβ_1-42 sequence on the stability of the resulting complexes has been discussed. Finally, the changes in the peptide structures on account of complex formation were analyzed.

1. Introduction

Pharmacotherapeutic procedure in Alzheimer’s disease (AD) involves the use of pain relievers (e.g., drugs that enlarge the metabolic activity of the brain tissue), symptomatic relievers (e.g., antidepressant drugs, neuroleptics), substitution relievers (e.g., drugs that increase the concentration of acetylcholine in the brain), combined agents (drugs with different mechanisms of action), pleiotropic agents (estrogens, aluminum chelating agents), as well as agents reducing tau protein synthesis and deposition of amyloid β deposits with the use of agents inhibiting neurofibrillary degeneration [1,2]. On the other hand, it allows only to alleviate symptoms, but is not effective in stopping a development of the disease and reversing the resulting neurodegenerative changes [3,4] induced among others by chronic ethanol administration [5], chronic unpredictable mild stress [6], imbalance in various metal ions intake [7], or excessive amounts of reactive oxygen [8] and nitrogen species [9]. Thus, new drugs (agents currently used in the therapies of other diseases, as well as compounds found in raw materials of natural origin) and alternative treatment methods are sought and—along with an appropriate diet and reduction of deleterious environmental factors—could be very helpful in preventing Alzheimer’s disease [10].

Studies on potential drugs for Alzheimer’s disease include, among others, nonsteroidal anti-inflammatory drugs (NSAIDs) which exhibit anti-inflammatory, analgesic and antipyretic properties [11]. NSAIDs inhibit the activity of cyclooxygenase (COX) responsible for the altering of arachidonic acid to prostaglandins [12]—the mediators of inflammatory processes, resulting in the expansion of blood vessels, increased body temperature, increased capillary permeability, and inflammatory cells [13]. NSAIDs also probably affect amyloid precursor protein (APP) metabolism, independent of COX, thereby decreasing the amount of amyloid Aβ_1-42 in the brain. Finally, in vitro studies have shown that NSAIDs modulate neuronal responses by, for example, inhibiting glial cell responses to amyloid β [14,15,16]. Despite some speculations about differences in the reliability of action between aspirin and various NSAIDs and a fact that additional studies in that field are required [17], aspirin has been found to have a protective effect in AD [18]. Ibuprofen can be used in chronic inflammatory Alzheimer’s disease as well—it probably contributes to the inhibition of γ secretase activity, which is responsible for the formation of amyloid β [19]. However, similarly to aspirin, the data on the beneficial effects of that drug on AD are not entirely clear and require further research [20]. Apart from NSAIDs, there are also reports in the literature on substances of natural origin as potentially active substances in Alzheimer’s disease [21], for example pyridine and piperidine alkaloids [22,23,24]. Anabasine is an example of nicotinic alkaloids, a component of a tobacco smoke [25,26]. Next to nicotine, it is an example of a ligand for neuronal nicotinic acetylcholine receptors (nAChRs) [27]. Since the most available drugs for the treatment of Alzheimer’s disease are indeed acetylcholinesterase inhibitors and muscarinic M1 receptors agonists [28], the studies on anabasine in AD development seem to be justified. In the same context, many neuroanatomic experiments revealed the influence of epinephrine (adrenaline) on the modulation of behavioral and cardiovascular function [29,30,31].

In this manuscript, we present the results of our studies on the interactions of human amyloid β protein fragment 1-42 (Aβ_1-42; the structure is shown in Figure 1a) and its three chosen fragments, namely Aβ_5-16 (RHDSGYEVHHQK; HZ1), Aβ_8-13 (SGYEVH; HZ2), and Aβ_8-12 (SGYEV; HZ3) (the sequences are presented in Figure 1b) with selected painkillers: ibuprofen and aspirin as well as two compounds of natural origin: anabasine and epinephrine (the structures are presented in Figure 2). The principal aim of this work was to study the binding properties of Aβ_1-42 polypeptide to the chosen biologically active compounds. It should be noted that the conformation of the dominant family of HZ1 aligns appropriately on the corresponding section of the native Aβ_1-42 [32], particularly when compared orientation of residues from the central part of the sequence with the most expanded side chains (i.e., tyrosine or histidine). Therefore, in particular, the attention has been focused on the influence of symmetrically oriented side chains of the amino acid residues in peptides under study on the stability of the resulting complexes. Furthermore, the studies were focused on the fragments from the first part of the sequence because previously it has been proven that probably three residues in the original sequence Aβ_1-42 (namely, His6, His14 and His16) serve very often as binding sites for other ligands [33,34].

2. Experimental

2.1. Materials and Methods

Human amyloid β protein fragment 1-42 (95%, by HPLC) was obtained from AnaSpec Inc. (CA, USA) and used as received. Its shorter fragments chosen for the experiments (Aβ_8-12, Aβ_8-13, and Aβ_5-16) were synthesized [36] and chromatographically purified [37] using procedures described previously. The peptides were then characterized using mass spectrometry. The concentration of the probes was confirmed spectrophotometrically using Perkin Elmer Lambda 650 UV/Vis spectrophotometer on the basis of the measurement of absorbance of tyrosine at 280 nm (ε₂₈₀ = 1280 M⁻¹cm⁻¹) [37]. Aspirin (acetylsalicylic acid, >95%), ibuprofen (α-methyl-4-(isobutyl)phenylacetic acid, 99%), anabasine (neonicotine, >97%), and epinephrine (adrenaline, >95%) were obtained from Sigma Aldrich (Poland) and used as purchased.

2.2. Fluorescence Spectroscopy

Fluorometric experiments were carried out at 20 °C on a Cary Eclipse Varian spectrofluorometer equipped with a temperature controller and a multicell holder. The fluorescence emission spectra (λ_ex = 275 nm) of all studied peptides were recorded from 280 to 400 nm. During titration experiments, 2 mL of each peptide solutions in 5 mM MES buffer pH 6.0 (c_HZ1 = 38.8 µM, c_HZ2 = 39.2 µM, c_HZ3 = 34.5 µM) were titrated with six 5 μL aliquots of aspirin, ibuprofen, anabasine, and epinephrine solutions in DMSO (c_aspirin = 0.1 M, c_ibuprofen = 0.1 M, c_anabasine = 0.1 M, and c_epinephrine = 0.01 M). The fluorescence intensity values determined at 305 nm (the maximum of emission of tyrosine and thus all peptides under study) were measured in the presence of increasing concentrations of selected drugs (pure DMSO was used as a control probe) and were further used to estimate the strength of interactions and determine association constants (where possible). The absorption of light by ibuprofen, aspirin, anabasine, and epinephrine at the excitation and emission wavelength of tyrosine (275 and 305 nm, respectively) has been considered and consequently fluorescence intensity values have been corrected for inner filter effect based on the Equation (1):

$${\mathrm{F}}_{\mathrm{corr}}={\mathrm{F}}_{\mathrm{obs}}\times {10}^{\frac{{\mathrm{A}}_{275}+{\mathrm{A}}_{305}}{2}},$$

(1)

where F_corr and F_obs correspond to the corrected and observed fluorescence intensity values, respectively; while A₂₇₅ and A₃₀₅ correspond to the absorbance values measured at the excitation and emission wavelength, respectively [38,39].

2.3. Molecular Dynamics (MD)

Theoretical studies were performed using the AMBER 16 program [40] at constant temperature and volume (NVT scheme) and with the AMBER force field, the version ff14SB. All calculations were carried out in the periodic box, composed of water molecules, type TIP3P. During calculations, the Ewald procedure was implemented with a mesh of particles for electrostatic long-range interactions at a temperature of 10 °C. Molecular dynamics simulations were carried out with aspirin, ibuprofen, anabasine, and epinephrine added to Aβ_1-42 and HZ1 (for shorter peptides, namely HZ2 and HZ3, the MD calculations were not performed due to the excessive mobility of the systems). In the calculations involving Aβ_1-42, the peptide structure was defined on the basis of data from the PDB bank. In theoretical considerations for HZ1 peptide, limitations resulting from the NMR experiment, in the distance between the selected atoms and angles were taken into account (175 distance restraints and 32 dihedral angles restraints). In addition, during simulations for protons for which no NOE signals were observed, so-called “anti-NOE” limitations also were performed. This approach usually minimizes any deviation from the AMBER ff16SB force field that aims to favor the a-helical conformations. For each trajectory, the time of simulation was t = 10 ns, and the integration time step was 2 fs. In each simulation, counter ions have been added to neutralize systems.

3. Results

The fluorescence intensity of all selected peptides increases systematically with the increase of the amount of the added drug (painkiller and/or compound of natural origin), which may be a result of a variety of processes, among others excited state reactions, ground-state complex formations or collisional interactions (Figure 3). The dissociation constants (K_D) of the resulting complexes can be determined by Lineweaver-Burk Equation (2):

$$\frac{1}{{\mathrm{F}}_0-\mathrm{F}}=\frac{1}{{\mathrm{F}}_0}+\frac{K_{\mathrm{D}}}{{\mathrm{F}}_0\left[\mathrm{Q}\right]},$$

(2)

In the case of systems where a significant and linear relationship between $\frac{\mathrm{F}-{\mathrm{F}}_0}{{\mathrm{F}}_0}$ and ${\mathrm{c}}_{\mathrm{drug}}$ was observed, the Lineweaver-Burk plots were constructed as the relationship of $\frac{{\mathrm{F}}_0}{{\mathrm{F}}_0-\mathrm{F}}$ vs. ${\mathrm{c}}_{\mathrm{drug}}^{-1}$ (Figure 4). From the regression equations of these curves, association constants ($K_{\mathrm{A}}=K_{\mathrm{D}}^{-1}$) of the resulting complexes have been determined as the averages of two independent experiments. The values of these association constants are presented as insets in Figure 4.

The molecular dynamics simulations (MD) were used to investigate the interactions of ibuprofen, aspirin, epinephrine, and anabasine with Aβ_1-42 and HZ1 polypeptides. It should be noted that for HZ1 peptide the NMR-derived restrains were used for calculations. It has been previously confirmed, that the main conformation of HZ1 forms a well-defined bent structure in its central part (Ser4–His10) with quite flexible ends [32]. Figure 5, Figure 6, Figure 7 and Figure 8 present the results of theoretical calculations obtained for both peptides with ibuprofen, aspirin, epinephrine, and anabasine, respectively.

4. Discussion

From the inspection of Figure 3 and Figure 4, it can be observed that all studied peptides exhibit the strongest affinity to epinephrine. It is demonstrated by significantly higher values of association constants in case of peptide-epinephrine complexes (when compared to complexes with ibuprofen, aspirin, and anabasine), and is in a great agreement with MD simulations. It can be supposed that the presence of hydroxyl groups capable to form hydrogen bonds (under experimental conditions; pH 6.0) with the side chains of the peptides is probably the most important factor responsible for the strong epinephrine–peptide interaction. In case of interactions with Aβ_1-42, epinephrine maintains a distinct distance in relation to the N-terminus and fluctuates steadily around the C-terminus (Phe4, His6 residues). The dominant spatial arrangement is probably stabilized by п-п interactions of two aromatic rings, Phe4 and epinephrine. Such a stabilization may be responsible for the obtained K_A value which is approximately 30 times higher in case of complex of epinephrine with Aβ_1-42 rather than with HZ1, HZ2 or HZ3. According to the HZ1 peptide it can be seen that epinephrine molecules—similarly to Aβ_1-42—still locate around the beginning of the peptide chain causing its deeper bend. Epinephrine is located in the space defined by the electrostatic fields of Asp3 and His2. It is worth emphasizing that Tyr10 is also located within the space occupied by epinephrine, which probably stabilizes the position of the molecule. The observed possible interaction with Tyr10 explains similar (within the experimental error) K_A values of complexes of epinephrine with HZ1, HZ2 and HZ3.

Both fluorescence spectroscopy and theoretical simulations revealed a significant affinity of the studied peptides towards ibuprofen, as well. However, in the case of Aβ_1-42, the ibuprofen molecule shifted significantly towards the C-terminus during the simulation. The ligand found itself in a space defined by the electrostatic field of Met35, Val40 and Ile41 and clearly shifted from the N-terminus. In the case of HZ1, we clearly observed the sliding of the ibuprofen, especially its ring moiety, into the space defined by the side chains of tyrosine and histidine. Here, we observed clearly that the system is stabilized by п-п interaction of the side chain of tyrosine with ibuprofen. After the ibuprofen molecule was arranged parallel to the tyrosine side chain, the system showed stability during further calculations and did not change its position. Similarly to the results observed in case of epinephrine, the participation of Tyr10 in a stabilization of the system may be responsible for the comparable values of association constants of complexes of ibuprofen with HZ1, HZ2 and HZ3.

The results obtained from calculations for anabasine revealed the lack of clear affinity of that compound for both Aβ_1-42 and HZ1. In case of the latter, the observed minimum distance to anabasine was above 15 Å. In case of Aβ_1-42, during the simulations, it was possible to observe the beginning of the unfolding of the helix from the C-terminus part of the sequence, even though anabasine was oscillating around the middle of the sequence. Unfortunately, the dominant conformations of the two-component system clearly indicate the stability of the system when these two molecules are spaced, and are far away from each other. These observations are in an agreement with the results of spectrofluorometric titrations, since Aβ_1-42 was the sole peptide in case of which the use of Lineweaver–Burk equation enabled estimation of association constant—its very low value confirms extremely slight interactions with anabasine.

The results obtained for aspirin are similar to those for anabasine. Molecular dynamics simulations revealed that aspirin destabilizes the Aβ_1-42 structure, causing unfolding of the helix in the C-terminal part of the sequence, apart from the fact that it does not clearly interact with the system during the simulation. According to the HZ1 peptide, the dominant conformations of the two-component system clearly indicate the stability of the system when these two molecules are spaced, and are far away from each other. The results of spectrofluorometric titrations clearly confirm that, although the use of Lineweaver-Burk formula made it possible to determine appropriate K_A values, they are very low and demonstrate very low interactions.

Furthermore, it has been proven that the sequence of the amino acid residues and the orientation of the side chains has the impact on the strength of the interaction on account of a steric effects which hinder the interactions between the tested compounds.

5. Conclusions

The fluorescence spectroscopy supported by molecular dynamics simulations (MD) with NMR-derived restrains was used to study the interactions of Aβ_1-42 polypeptide and its derivatives with selected low-molecular weight organic compounds, namely ibuprofen and aspirin (the painkillers) as well as anabasine and epinephrine (naturally occurred compounds). In all cases, the shortening of the Aβ_1-42 polypeptide chain caused changes in peptides conformation that were significant enough to affect their ability to bind the investigated compounds. Experimental results and MD calculations showed that epinephrine (adrenaline) reveals the highest affinity to the investigated peptides. It can be supposed that the presence of hydroxyl groups capable to form hydrogen bonds (under experimental conditions; pH 6.0) with the side chains of the peptides is probably the most important factor responsible for the strong epinephrine–peptide interaction. Furthermore, it has been proven that the sequence of the amino acid residues and the orientation of the side chains have the impact on the strength of the interaction on account of a steric effects which hinder the interactions between the tested compounds. Finally, it is worth emphasizing that only in the case of aspirin a strong effect on the disturbance of the structure of Aβ_1-42 was observed. This phenomenon manifested itself in the unfolding of the C-terminal strand of the protein and can be explained by the specific arrangement of the peptide chain which folded so uniquely.