In vitro MAO-B assay

The MAO-B effects were tested on a recombinant human MAO-B employing the fluorimetric method Amplex UltraRed reagent with several modifications (Mateev et al. 2022a). Tyramine hydrochloride was utilized as a substrate, and Selegiline as a positive control.

The crystallographic structures of MAO-B (PDB: 2V5Z) and AChE (PDB: 4EY6) were retrieved from the Protein Data Bank (PDB). The initial protein structures of the target enzymes were refined with the Protein Preparation module in Maestro. Applying the former, hydrogen bonds were added, het states were generated and the crystallographic structures were minimized with the OPLS4 force field. The docking simulations were carried out with two robust docking softwares - GOLD 5.3 and Glide (Schrödinger Maestro Suite). The grid boxes were arranged around each co-crystallized ligand (Safinamide for MAO-B, and Galanthamine for AChE). Induced-fit docking (IFD) (Schrödinger) was used to enhance the robustness of the active conformations. The IFD enables full side chain flexibility and scores each new active solution with the Glide’s XP score.

All of the theoretical computations were carried out with Jaguar (Schrödinger Release 2022-1: Jaguar, Schrödinger, LLC, New York, NY, 2021.) (Bochevarov et al. 2013). Initially, the geometries of the most potent multitarget drugs were refined with conformational simulations with 250 iterations and OPLS4 force field. Subsequently, DFT calculations with Becke’s three-parameter hybrid exchange–correlation functional (B3LYP) and 6-311⁺⁺G (d,p) basis set were performed. The frontier molecular orbitals (FMO) and the global reactivity descriptors for the most prominent drug were calculated at the same level of theory. The computational simulations were performed on an AMD Ryzen 9 5950X 16 core CPU and 64 GB of installed RAM. Windows 10 Pro was used as an operating system.

The synthetic route of the title pyrrole-based derivatives was recently reported (Mateev et al. 2022a). One novel carboxylic acid 3, its corresponding hydrazide 5 and four new hydrazide-hydrazones 5a-d were obtained through Paal-Knorr condensation, hydrazide formation and condensations with various aldehydes, respectively (Fig. 1).

Figure 1. 

Synthetic route of the applied pyrrole-based hydrazide-hydrazones. All compounds were fully elucidated by instrumental (¹H NMR, FT-IR, and HRMS), and chromatographic analyses.

In vitro MAO-B assay

Compounds 3, 5 and 5a-5d were screened for their MAO-B inhibitory activity by an in vitro fluorometric method. Selegiline was used as a standard drug. The results are given in Fig. 2.

Figure 2. 

Activities of the title compounds against MAO-B applied as 1 μM concentrations. Data are presented as means from three independent experiments ± SD. * P < 0.1, ** P < 0.01, *** P < 0.001 vs control (pure hMAO-B).

The standard Selegiline demonstrated 55% hMAOB activity, compared to the control (pure hMAOB). The unsubstituted hydrazide 5 revealed a good inhibitory capacity of 30%, while the pyrrole acid (3) showed moderate activity of 25%. All hydrazide-hydrazones exerted no MAO-B inhibitory potential. Recent study held by Altintop et al. discussed the synthesis and evaluation of MAO-B effects of novel pyrrole-based compounds (Altintop et al. 2018). The most promising ligand displayed IC₅₀ of 1.642 ± 0.082 μM. The molecular weights of the compounds varied in the range of 290–350 g/mol, which is significantly lower compared to the employed in this study dataset. Thus, further optimizations of the title pyrroles could be pointed towards lowering the molecular weights of the hydrazide-hydrazones.

In vitro AchE assay

The inhibitory capacity of the title compounds against eeAChE (electric eel acetylcholinesterase) was measured according to a modified Ellman‘s method (Chigurupati et al. 2016). Donepezil was used as a reference compound. The blocking capacities of the test compounds, at 10 μM concentrations, are provided in Fig. 3.

Figure 3. 

Activities of the title compounds against AChE applied as 10 μM concentrations. Data are presented as means from three independent experiments ± SD. * P < 0.1, ** P < 0.01, *** P < 0.001 vs control (pure eeAChE).

The most active compound was the hydrazide-hydrazone – 5d, with 75% inhibitory activity against the AChE enzyme. The unsubstituted hydrazide – 5, showed a good blocking ability of 51%. Interestingly, the pyrrole-based compounds substituted with 4-methoxy (5a), 2,4-dimethoxy (5b), and 2,3-dimethoxy (5c) benzaldehydes revealed no significant effects (13%, 19%, and 22%, respectively). The initial pyrrole-carboxylic acid (3) was inactive against AChE. The results determined the importance of -OH in the benzene ring for significantly higher AChE activity.

Overall, it was noted that the most prominent dual acting MAO-B/AChE inhibitor is the hydrazide 5. The former compound exerted good MAO-B inhibitory activity (30% at 1 μM concentration) and excellent AChE blocking capacity (51% at 10 μM concentration).

In an attempt to gain further understanding of the in vitro results, and to obtain additional insights into the active binding conformations of the inhibitors, molecular docking simulations of the title compounds within the active site of MAO-B (PDB: 2V5Z) and AChE (PDB: 4EY6) were carried out.

Initially, the docking simulations in MAO-B with GOLD 5.3 (ChemPLP scoring function) were performed. All ligands were successfully situated in the active site, however, only compounds 3 and 5 showed good theoretical results. Interestingly, Glide did not return any poses of the pyrrole-based ligands. Therefore, we included IFD calculations, which achieved good scores for compounds 3 and 5 (-11.20 kcal/mol, and -14.80 kcal/mol, respectively) (Table 2).

The GOLD’s ChemPLP scoring algorithm obtained acceptable results for the N-pyrrolylcarboxilic acid 3 and the hydrazide 5. The substituted hydrazide-hydrazones demonstrated many steric clashes, which led to drastically lowered fitness scores. The docking calculations concluded that the IFD protocol is the most suitable considering the positive correlations with the in vitro data. The former simulations did not acquire any results for 5a, 5b, 5c and 5d due to the inability of IFD to situate the ligands in the active site of MAO-B. Thus, the stabilizing intermolecular interactions were provided only for 3, 5 and the standards Safinamide and Selegiline (Table 3).

The active cavity of MAO-B is constructed of a substrate cavity and an entrance cavity. Furthermore, the substrate cavity includes an aromatic cage, which is of great importance for the MAO-B activity (Dasgupta et al. 2021). The active amino residues Ile199, Tyr326, Tyr398 and Tyr435 have been reported as essential for the selectivity against the B isoform (Tzvetkov et al. 2014). The N-pyrrolylcarboxylic acid (3), formed four π–π interactions with Tyr326, Tyr398 and Tyr435. The fragments, which participate in the former stable bonds, were the pyrrole and tryptophan aromatics. No hydrogen bond between 3 and MAO-B was formed. Importantly, the amino acids from the “aromatic cage” of the active site of MAO-B were included in the stabilizations of both 3 and 5. The visualization of the complex 5-MAO/B is provided in Fig. 4.

Figure 4. 

Visualizations of the intermolecular interactions of 5 with the active site of MAO-B (PDB: 2V5Z). A. 2D panel; B. 3D panel.

The tryptophan moiety of the hydrazide 5 was situated in the “aromatic cage” of the active site where it was stabilized by π–π interactions with Tyr398 and Tyr435. The former stabilizing interactions are important for the selective MAO-B activity. The pyrrole moiety was located in the substrate cavity of the enzyme, where Cys172, Ile198, Ile199, and Try326 were involved in π–π interactions with the aromatic heterocycle. The ethyl ester of 5 was placed in the entrance cavity. Importantly, the hydrazide was further stabilized by H-bond with the active amino acid Gln206. A stable H-bond was not present in the case of the N-pyrrolylcarboxilic acid 3 which could explain the enhanced experimental MAO-B inhibitory activity of 5.

Overall, the docking simulations in MAO-B demonstrated that shorter pyrrole-based compounds exert better MAO-B blocking capacities.

Both GOLD 5.3 and Glide successfully returned active poses of all ligands when docked in AChE (PDB: 4EY6). The docking scores are presented in Table 4.

The most prominent AChE inhibitor from the in vitro studies – 5d (75% activity at 10 μM) revealed the best XP score of -14.27 kcal/mol and moderate ChemPLP score of 96.40. The most unfavorable scores were retrieved after docking simulations with the N-pyrrolylcarboxilic acid 3. The former calculations were in good agreement with the experimental results. Interestingly, the highest GOLD 5.3 fitness scores was achieved by the hydrazide-hydrazone condensed with anisaldehyde - 5a (103.81). However, the in vitro data showed insignificant AChE inhibitory activity – 13% (10 μM). The former data illustrated the major disadvantage of the docking programs – increased risk for false-active ligands (Makeneni et al. 2018). The active amino acids involved in hydrogen bonds, π–π and hydrophobic interactions are provided in Table 5.

Importantly, most of the stabilizing interactions were with hydrophobic nature. The former bonds have been described in the stabilization of several active AChE inhibitors (Chigupati et al. 2016; Tang et al. 2019). The active amino acid Trp86 was involved in π–π bonds with most of the title pyrrole-based ligands. The significance of Trp86 for the formation of potent inhibitors is discussed by several research groups (Ordentlich et al. 1995; Ranjan et al. 2018).

Subsequent visualization of the intermolecular interactions between the most prominent dual inhibitor - 5, and the active site of AChE is given in Fig. 5.

Figure 5. 

Major intermolecular interactions between the active site of 4EY6 and 5. A. 2D visualization; B. 3D visualization.

The pyrrole-based hydrazide 5 formed compact active conformation in the active site of AChE. The tryptophan moiety was facing the entrance cavity of the enzyme, while the pyrrole fragment was situated in the substrate pocket. Additionally, a hydrogen bond between the active amino acid His447 and the secondary amino group stabilized the complex. Interestingly, the hydrazide 5 did not participate in any stabilizations with amino acids from the peripheral active site (PAS) which negatively affects the docking scores.

A search for the most energetically favorable geometry of a ligand is fundamental prior to subsequent computational simulations. Therefore, the most prominent dual MAO-B/AChE inhibitor (5) was applied for full optimization calculations. Initial conformational search was performed by 250 iterations with OPLS4 force field. The energetically favorable conformations were further optimized by full DFT geometry optimization at B3LYP/6-311⁺⁺ (d,p) level of theory. The most favorable conformation is given in Fig. 6.

Figure 6. 

Optimized geometry of the most effective dual MAO-B/AChE inhibitor 5 at B3LYP/6-311⁺⁺(d, p).

A conceptual DFT analysis of the stability and reactivity of the most active dual inhibitor 5 was conducted by calculations of the highest occupied molecular orbital (HOMO), the lowest unoccupied molecular orbital (LUMO). Subsequently, the values of the global reactivity descriptors, such as ionization potential (IP), electron affinity (EA), molecular hardness and softness, electronegativity and electrophilicity were obtained. The HOMO and LUMO electronic densities of 5 with DFT/B3LYP/6-311⁺⁺(d,p) based calculations are given in Fig. 7. The positive phase was colored in blue, while the negative one in red.

Figure 7. 

HOMO (A) и LUMO (B) of the best dual acting MAO-B/AChE inhibitor – 5.

It was found that the HOMO of 5 is mainly localized on the tryptophan moiety while the LUMO is centered around the p-bromophenyl fragment. Subsequently, the energies (in atomic units) of the FMOs and the global reactivity descriptors (hardness (η), softness (S), electronegativity (χ), chemical potential (μ), and electrophilicity index (ω)) were calculated by employing the Koopman’s theorem (Tsuneda et al. 2010). The DFT calculations of FMOs and global reactivity descriptors (in eV) are provided in Table 6.

The energies of HOMO and LUMO of one compound provide data about the distribution of the internal energy in a system. The ionization potential of compound 5 was significantly higher (5.5 eV) compared to the electron affinity value (0.86 eV). Therefore, the electron donor capacity of the pyrrole-based hydrazide is drastically enhanced than the electron donor ability. The big gap between the energies of the LUMO and HOMO represents the hydrazide 5 as a molecule with low polarity, low reactivity and high stability (Tawari et al. 2010). The global hardness η can be regarded as a direct measure of the electron density deformation and of the chemical reactivity. The η value of 5 (2.30 eV) revealed low reactivity (Pradeep et al. 2021). The negative value of the chemical potential (-3.16 eV) of compound 5 establishes good stabilities, and resistance to sudden decomposition. The calculated electrophilicity index value (2.17 eV) is related to the chemical potential and hardness, and it is an indicator for the nucleophilicity power (Choudhary et al. 2019). The discussed DFT results showed that the best dual acting MAO-B/AChE inhibitor - 5, possess high ionization potential and low electron affinity. The noted big HOMO-LUMO gap represents the hydrazide as a stable molecule with low reactivity.