*2.5. Docking Studies*

Selection of the biological activity for docking study was done with the help of computer program PASS [20]. Software estimates predicted activity spectrum of a compound according to the structural formulas of synthesized fluorinated pyrazole aldehydes as probable activity (Pa) and probable inactivity (Pi). Activity with Pa > Pi is considered as possible for a particular compound. Phosphodiesterase inhibition has been selected for docking study since the PASS program calculated Pa > Pi of all compounds for that activity.

The molecular docking of compounds (**3a**–**j**) was performed using iGEMDOCK (BioXGEM, Taiwan). Crystal coordinates of the catalytic domain of phosphodiesterase type 5 (PDE5) (PDB ID: 4OEW) in the complex with monocyclic pyrimidinones (PDB ID: 5IO) were downloaded from Protein Data Bank (PDB, https://www.rcsb.org/). The PDE5 structure was prepared, including the removal of water molecules and optimized protein structure using BIOVIA Discovery Studio 4.5 (Dassault Systèmes, San Diego, CA, USA). Avogadro 1.2.0 (University of Pittsburgh, Pittsburgh, PA, USA) was applied for optimizing the 3D structures of 28 molecules using the molecular mechanic's force field (MM+) [21]. In addition, the semiempirical PM3 method was used for geometry optimization of all structures [22].

The protein binding site was outlined according to the bounded ligand (PDB ID: 5IO) [23]. Genetic parameters were set (population size 200, generations 70, the number of a solution or poses: 2) after the preparation of the protein target and set of optimized structures of 10 fluorinated pyrazoles as ligands. Docking into the binding site and generation of protein-compound interaction profiles of electrostatic (*Elec*), hydrogen-bonding (*Hbond*), and van der Waals (*vdW*) interactions was performed for each compound in the library. Finally, by combining pharmacological interactions and energy-based scoring function, the compounds were ranked. Energy-based scoring function or total energy (*E*) is:

$$E = \upsilon d\mathcal{W} + Hbond + Elec.\tag{1}$$

### **3. Results and Discussion**

Desired 4,5-dihydro-1*<sup>H</sup>*-pyrazole derivatives were synthesized from corresponding chalcones. Chalcones were obtained in the typical aldol condensation reaction of aldehydes and acetophenones, while pyrazoles were synthesized from corresponding chalcones in the presence of hydrazine hydrate and formic acid (Figure 2). Their structures were confirmed by 1H NMR, 13C NMR, and mass spectra. All compounds show characteristic peaks for –CHO proton around 8.67 ppm, pyrazole C-5 proton peak around 5.58 ppm and pyrazole C-4 proton peaks around 3.24 ppm and 3.90 ppm. Other peaks correspond to aromatic protons (6.90–7.90 ppm) and specific substituents on the phenyl ring. Mass spectra for each compound corresponds to its molar mass. All compounds were further characterized by their melting points and *R*f values as indicated in the Materials and Methods Section.

**Figure 2.** Synthetic pathway for fluorinated pyrazole aldehydes.

After synthesis, purification, and full characterization, all compounds were investigated for their antibacterial activity against two Gram-positive and two Gram-negative bacteria (Table 1) and tyrosinase inhibiting activity (Table 2).

**Table 1.** Antibacterial activity of synthesized compounds in terms of minimum inhibitory concentration (MIC) against *Escherichia coli, Pseudomonas aeruginosa, Bacillus subtilis* and *Staphylococcus aureus*(μg mL−1).


**Table 2.** Tyrosinase inhibiting activity of synthesized compounds \*.


\* concentration of tested compound in the reaction mixture was 100 μM. Results present mean value ± standard deviation of triplicate measurements.

The antibacterial assay revealed that Gram-negative bacteria were more susceptible to the tested compounds than Gram-positive ones. The cell wall of Gram-positive bacteria ranges from 20 to 80 nm, while for Gram-negative bacteria it ranges from 1.5 to 10 nm [24,25]. Of the four tested bacteria, the Gram-positive had higher resistance against tested compounds, with up to threefold higher MIC values (62.5–250 μg mL−<sup>1</sup> for *B. subtilis* and *S. aureus*) than for Gram-negative bacteria (62.5 μg mL−<sup>1</sup> for *E. coli* and *P. aeruginosa* for all tested compounds). Electrostatic interactions can regulate the interaction of the bacterial surface with acidic and basic functional groups and various agents, which can lead to changed cell surface permeability and thus lead to the death of the cell. The target protein often tolerates the replacement of a hydrogen atom with fluorine, since it has a small atom volume. In order to increase the half-life of the drug and human exposure, many drugs on the market contain introduced fluorine atoms [26]. The introduction of a fluorine atom into a molecule can change the distribution of electrons and thus a ffect pKa, dipole moment, and even chemical reactivity and stability of adjacent functional groups since it is the most electronegative element. The bioavailability of the compounds can also be improved by higher membrane permeability for the compound as a result of reduced compound basicity due to the introduced fluorine [27]. Threefold higher MIC was found with the compounds **3h**–**j** that had various halogen (bromo/chloro) or dimethylamino group containing substituents implying that electrostatic distribution a ffects membrane permeation of the compound.

Determination of tyrosinase inhibitory potential revealed that none of the compounds significantly inhibited mushroom tyrosinase at 100 μM concentration, in comparison with the kojic acid as a standard inhibitor, which exhibited IC50 of 16.96 ± 1.05 μM for monophenolase, and 13.10 ± 1.02 μM for diphenolase activity. Nevertheless, in most cases, greater tyrosinase inhibition could be observed for monophenolase than diphenolase activity. Among synthesized compounds, compound **3i** was found as the strongest inhibitor of monophenolase activity of mushroom tyrosinase (32.07 ± 3.39%), but its inhibiting activity of diphenolase activity was found twofold lower. The most probable reason for the lack of significant tyrosinase inhibitory activity of synthesized compounds is pyrazole ring *N*-substitution with aldehyde group. Zhou et al. (2013) have described that *N*-acetylation at pyrazole ring causes diminished inhibitory activity when compared to non-substituted compounds [6]. In addition, based on the report of Zhou et al. [6] it seems obvious that presence of hydroxyl groups on phenyl rings might be the prerequisite for tyrosinase inhibiting activity, which was not the case in the present study where phenyl ring A was fluorinated, and ring B had a various non-hydroxyl group containing substituents.

Experimentally proven inactivity of synthesized compounds towards mushroom tyrosinase was the motive to find another potential biological activity for these compounds. PASS online program (http://www.pharmaexpert.ru/passonline/) provides the prediction of several hundred biological activities based on structural formulas. For almost all synthesized compounds PASS program has predicted the highest *Pa* for the phosphodiesterase inhibition. All the compounds showed greater *Pa* than *Pi* (Table 3).


**Table 3.** Results of PASS program for the phosphodiesterase inhibition.

\* probable activity (*Pa*) and probable inactivity (*Pi*).

Phosphodiesterase type 5 (PDE5) is a cyclic guanosine monophosphate (cGMP-specific) enzyme and mostly expressed in smooth muscle tissue of corpus cavernosum. PDE5 inhibitors have vasodilative effects, therefore, are used for treating erectile dysfunction, pulmonary hypertension and cardiovascular diseases [28]. In order to provide virtual screening of synthesized compounds as potential inhibitors of PDE5 molecular docking study was performed. Docking score and energy of interactions between protein residue and ligand are tabulated in Table 4.


**Table 4.** Docking scores for fluorinated pyrazoles in interaction with PDE5.

According to the docking scores, compound **3g** showed the lowest total energy, which indicates it best fits into the active site of PDE5. The energy of the interactions between protein residue and ligand **3g** are tabulated in Table 5.

**Table 5.** The energy of the main interactions between protein PDE5 residue and ligand **3g.**


(M = main chain; S = side chain).

Potential surface representation of PDE5 binding site with docked compound **3g** is presented in Figure 3, while Figure 4 illustrates the interactions of ligand **3g** with receptor PDE5 in the binding site.

**Figure 3.** Potential surface representation catalytic domain of PDE5 with docked compound **3g**. (Range of potential: from min. 1.77 mV (blue) to max. 0.541 mV (red)).

The binding site of PDE5 was defined according to the inhibitor--halogen derivate of monocyclic pyrimidinones (PDB ID: 5IO). Molecular docking confirmed the previous findings of characteristic binding interactions of inhibitors with the PDE5 catalytic domain [29]. Key interactions of compound **3g** are bidentate hydrogen bond (H-bond) with the side-chain of Gln817. One H-bond is formed with an oxygen atom of the carbaldehyde, and the second one with the nitrogen atom of pyrazole ring (Figure 4). Based on van der Waals interactions, pyrazole ring interacts with the side chain of Phe820. Side chain of Val782 forms π-π interactions with pyrazole ring and dimethoxyphenyl ring. The same ring is bonded to the Tyr612 by the π-donor hydrogen bond. Interactions of fluorophenyl ring are mediated through the π-σ interactions with Phe786 and π- sulfur interactions with Met816.

**Figure 4.** The main interactions of compound **3g** with residues in catalytic domain of PDE5: (**A**) 3D representation and hydrophobic surface of the binding site, (**B**) 2D representation (green = conventional hydrogen bond, light green = van der Waals, very light green = π-donor hydrogen bond, purple = π-σ interactions, light purple= π- π interactions, pink=alkyl and π-alkyl interactions, brown= π-sulphur bond).

Docking results of this study are in accordance with the solved crystal structure of PDE5 catalytic domain in complex with di fferent inhibitors. The catalytic domain of PDE5 includes three subdomains: N-terminal cyclin-fold region, a linker region and a C-terminal helical bundle in which the center is an active site of PDE5 core pocket (Q pocket) of the binding site contains Gln817, Phe820, Val782, Tyr612 [30]. Typical interactions include mono or bidentate hydrogen bonds of inhibitors with Gln817 and mainly π-π interactions of aromatic rings with hydrophobic clamp, which contains residuals, Val782, and Phe820. In the crystal structure of complex PDE5/sildenafil (PDB ID: 1UDT and 2H42) was confirmed the bidentate H-bonds are formed between the amide moiety of the pyrazolopyrimidinone of sildenafil and the side-chain of Gln817. Sildenafil (Viagra ®) is a PDE5 inhibitor, which is approved as the first oral medicine for the treatments of male erectile dysfunction and for treatment of pulmonary arterial hypertension [29]. Previously mentioned ligand, 5IO [23], also formed classical bidentate H-bonds with residue Gln817, π-π interactions of phenyl ring with Phe820 and hydrophobic interactions with residues Leu765, Val782, Ala783, and Phe786. Moreover, halogen bonding interactions, between Tyr612 and I atom have been recognized.
