Design, synthesis, and greener pasture biological assessment of a novel nucleoside: 1-(α-D-ribofuranosyl)-6,7-difluoro-2-methyl-4-quinazolinone as an inhibitor of COVID-19 and Alzheimer’s disease

Abstract

Synthesis of a new fluorinated nucleoside of 6,7-difluoro-2-methyl-4-quinazolinone was described. 2-Amino-4,5-difluorobenzoic acid 1 reacts with (CH₃CO)₂O followed by ammonia to form (1H)-6,7-difluoro-2-methyl-4-quinazolinone 3a. Ribosylation of a silylated 4 with l-O-acety1-2,3-5-tri-O-benzoyl-α-D-ribofuranose 5 forms a protected nucleoside 6 then unprotected from 6 to give a free nucleoside 7. Greener pasture biological docking of the cystine protease of COVID-19 [M^pro, code 7BQY, PDB] by novel nucleoside and fluoroquinazoline compounds is presented. LIGPLOT (2D) representations calculated for the same ligands are shown. A superposition of remdesivir approved medicine, N3 inhibitor, and our ligands docked together into the binding protein of 7BQY is also given for a fair comparison. The binding affinities of remdesivir, N3 inhibitor, the nucleoside 7, and fluoroquinazoline 3a, 3b compounds with 7BQY calculated under the same conditions are −7.7, −7.4, −7.6, −6.1, and −6.1 kcal mol⁻¹, respectively. The high values were due to the existence of many hydrophobic interactions and hydrogen bonds between the ligands and the active amino acid residues of the receptor, indicating a promising candidate as a COVID-19 inhibitor. Pro Tox -II server showed that compound 7 has a similar feature to the approved antiviral drug remdesivir for COVID-19. Additionally, a fascinating molecular modeling investigation showed that our nucleoside demonstrated good binding inhibition of AChE enzyme towards advancing an efficient medication against Alzheimer’s disease. Finally, DFT has been conducted to illustrate the MD results in terms of the molecular descriptor-based structural activity relationship calculated from FMOs.

1. Introduction

4-quinazolinones are solids and stable to alkaline treatment and mild acid. The most common approach to synthesizing 4(3H)-quinazolinone products is the direct reaction of 2-aminobenzoic acid (anthranilic acid) derivatives with acetic anhydride to give the benzoxazinone, followed by condensation with nitrogen nucleophiles [1–4]. The 2-methyl group in 3H or 1H-4-quinazolinone is more responsive than 4-quinazoline. In general, quinazolinones have several biological applications such as preventing cancer [2,5,6], anti-HIV [7], cytotoxicity in vitro [8], antiinflammatory [6,9], antibacterial [10,11], antifungal [12], and antiviral properties [13,14]. Furthermore, the kinase inhibitory potency of many N-aryl thiazolo[5,4-f]quinazolin-4-amines has been revealed, intending to improve the healing of Down syndrome (DS) and early AD [15–18]. The fluorine atom is a significant substituent in medicinal chemistry due to its electronic properties and improved molecular lipophilicity [19]. New fluorinated hydroquinazoline derivatives were used as antifungal agents [20] and anticancer agents [21]. A difluorinated inhibitor showed 4.23 times larger potency against the epidermal growth factor receptor (EGFR) than a nonfluorinated inhibitor [22]mutations of which have been linked specifically to nonsmall-cell lung cancer. For the L858R/T790M/C797S triplet mutant (EGFRTM. In addition, quinazolinone nucleosides [23–26] and fluorinated nucleosides have been used extensively in biological activity such as the prevention of cancer cells from forming, anti-HCV activity in vitro [27], and anti-HBV agents [28].

This article aims to report a new fluorinated nucleoside quinazolinone containing two fluorine atoms at positions 6 and 7 of the quinazolinone moiety [29]. When this compound was reacted with l-O-acety1-2,3-5-tri-O-benzoyl-β-D-ribofuranose 5 by silated method [30], the fluorinated nucleoside; 6,7-difluoro 2-methylquinazolin-4-one 7 was obtained. Also, given the current several ongoing concerns in medicine finding to monitor the horrific effect of the virus on our daily lives [3,31–33], we have been encouraged to screen, in silico as a greener pasture preliminary step, the interaction between our novel nucleoside: 1-(2,3,5-tri-O-benzoyl-β-D-ribofuranosyl)-6,7-difluoro-2-methyl-4-quinazolinone, 7 and fluoroquinazoline, 3 ligands with the main protease (downloaded from RCSB, PDB) deposition code 7BQY) active site as possible candidate drugs for COVID-19. It is worth mentioning here that nucleoside was tested before, in vitro, for antiviral activity against two representative cowpox viruses, ortho poxviruses, and vaccinia virus [34]. We also investigated, in silco, the molecular modeling of our nucleoside as a binding inhibition of AChE enzyme towards the advancement of an efficient drug against Alzheimer’s disease.

2. Materials and methods

TLC was done using silica gel 60 (aluminum sheet, Fluka company) and revealed by UV-vis light. Melting points (mp) of all synthesized compounds were determined using an electrothermal device and are uncorrected. The ¹H and ¹³C NMR spectra were measured on an NMR spectrometer in CDCl₃, CD₃OD at 213 and 850 MHz. Mass spectra were measured on GC MS-QP 2000 EX mass spectrometer at 70 e.V (King Abdel Aziz University).

2.1. Synthesis

Synthesis of 6,7-difluoro-2-methylbenzo[2,3-d]oxazin-4-one (2)2-amino-4,5-difluorobenzoic acid 1 (1.6416 g, 0,009 mol) with a suitable quantity of acetic anhydride was refluxed for 1 h to reveal compound 2 as a brown powder. Yield: 1.6971 g (89.5%); mp: 158 °C; molecular formula: C₉H₅NO₂F₂; molecular weight (mol. wt): 197.13.

Synthesis of 1H-6,7-difluoro-2-methyl-4-quinazolinone (3a)1.69 g (0.009 mol) of compound 2 with a suitable quantitative amount of ammonia solution was refluxed for 6 h, cooled, and then treated with acetic acid to give crystals collected by filtration as a white powder of compound 3a. Yield: 0.3690 g (21.95%); mp: 263–270 °C; molecular formula: C₉H₆F₂N₂O, mol. wt: 196.15, MASS m/z (%): M⁺= 196.06 (100%), 196.04 (100%) [29,35].

Ribosylation of 1H-6,7-difluoro-2-methyl-4-quinazolinone

1-(2,3,5-tri-O-benzoyl-β-D-ribofuranosyl)-6,7-difluoro-2-methyl-4-quinazolinone (6)A mixture of 1H-6,7-difluoro-2-methyl-4-quinazolinone 3a (0.369 g, 0.0018 mol), dry HMDS (20 mL) and a catalytic quantity of (NH₄)₂SO₄ were heated under reflux for 24 h (TLC). The product was evaporated to dryness in anhydrous media to afford the silylated derivative 4 as an intermediate compound, which added (10 mL) of dry 1,2-dichloroethane, (0.4694 g, 0.9 mmol) 1-O-acetyl-2,3,5-tri-O-benzoyl-β-D-ribofuranose (5) and (2 mL, 10 mmol) trimethylsilyl triflate; (TMSOTf as a catalyst). The obtained solution was stirred for 14 days at room temperature and then washed with an aqueous NaHCO₃ followed by water and dried over sodium sulfate. The silica gel column chromatography with ethyl acetate and chloroform (2:98) was used in order to separate the pure product. After evaporation, the main fraction 6 was obtained as a sticky yellow material. Yield: 0.378 g (49.93%); mp: 180 °C; molecular formula: C₃₅H₂₆F₂N₂O₈; mol. wt: 640.59; ¹HNMR (CDCl₃, 850 MHz, TMS), δ H: 8.06–7.13 (m, 17H Aromatic protons), 6.0–6.1 (d, 1H, J = 4.25 Hz, H-1′), 5.62 (t, J = Hz,1H, H-2′), 4.97 (t, J = Hz, 1H, H-3′), 4.49–4.72 (m, 1H, H-5′), 4.01–4.06 (m, 1H, H-4′), 1.56 (s, 3H) CH₃); ¹³CNMR (CD₃OD, 213 MHz, TMS), δ: 190.00, 167.63, 167.32, 164.32, 157.06, 155.73, 141.90, 138.05, 132.93, 132.87, 132.83, 130.57, 130.54, 129.71, 130.59, 129.92, 129.77, 129.74, 129.72, 129.50, 128.66, 128.43, 128.42, 128.40, 128.32, 128.29, 119.10; 117.75, 101.43, 79.16, 75.76, 70.25, 60.32, 65.14 CH₂, 17.65 CH₃.

Deprotection of 1-(2,3,5-tri-O-benzoyl-α-D-ribofuranosyl)-6,7-difluoro-2-methyl-4-quinazolinone-1-(α-D-ribofuranosyl)-6,7-difluoro-2-methyl-4-quinazolinone (7)

A mixture of compound 6 (0.5 mmol), and sodium metal (0.001g) in methanol (dry, 10 mL), 0.04 mmol) was stirred at room temperature for 1day. It was then treated with drops of acetic acid to neutralize the solution. The residue was recrystallized from water to offer 7 as light-yellow crystals. Yield: 0.152 g (81.28%); mp: 255 °C; molecular formula: C₁₄H₁₄F₂N₂O₅, mol. wt: 328.27; ¹HNMR (CD₃OD, 850 MHz, TMS), δ: 7.88 (s, 1H, H-5′), 6.94 (s, 1H, H-8), 6.9 (d, 1H, J = 6.80 Hz, H-1′), 6.60–6.63 (m, 1H, H-2′), 5.65 (d, 1H, J = 4.25 Hz, OH2′), 5.33 (d, 1H, J = 1.7 Hz, OH3′), 5.18 (d, 1H, J = 3.4 Hz, OH5′), 4.70–4.95 (m, 3H, H-3′, OH-3′, OH-2′), 3.63–3.68 (m, 2H, H-4′, OH-4′), 3.52(t, J₁ = 6.80, J₂ = 6.80 Hz, 1H, H-5′), 1.88 (s, 3H, CH₃); ¹³CNMR (CD₃OD, 213 MHz, TMS), δ: 160.0, 129.3, 128.8, 128.2, 127.2, 127.1, 126.0, 124.0, 78.0, 61.6, 51.2, 51.0, 50.9, 29.3; MASS m/z (%): M⁺ = 327.21, 326.00, 322.24, 319.18 (100), 306.24, 301.10, 298.21, 293.98, 287.08, 279.07, 263.12, 260.05, 328.14, 252.20, 243.94, 232.96, 220.02, 203.96, 193.11, 190.90, 184.95, 171.95, 163.98, 155.16, 142.95, 123.99, 107.03, 102.90, 91.97. (calc. 328.27).

2.2. Docking in silico studies

The molecular docking studies of our nucleoside compound and donepezil were done using the PyRx 0.8 (https://sourceforge.net/projects/pyrx/). It is a recommended and powerful visualization engine [36]offering preprocessing and postprocessing adapted (to date. The settings in the PyRx 0.8 include: Grid box (19.61, 29.01, 36.27 A˚), (3.54, 64.29, 64.01 A˚) centered at (19.61, 29.10, 26.27), (23.06, 25.0, 25.0) for 7BQY and 1ELE. Energy range = 4 and exhaustiveness = 8. Water molecules and the N3 ligand were deleted from the proteases (PDB code 7BQY, 1EVE). The key residues of 7BQY used in this study were identified before [31, 37].

2.3. Molecular descriptor-based structural activity relationship calculated from FMOs

The optimal architectural structures of the synthesized compounds 3a, 3b, and 7 were computed in the gas phase using Gaussian 9 on the B3LYP 6-311G basis set and have been used in the calculation of the molecular chemical descriptors. The energy levels of the frontier molecular orbitals highest occupied molecular orbitals (HOMOs) and lowest occupied molecular orbitals (LUMOs) could be used to compute several chemical descriptors (least unoccupied molecular orbitals). Furthermore, HOMOs and LUMOs in the examined compounds could be used as a qualitative predictor of their ability to donate or receive electrons from the neighboring receptor [38–40]. FMOs, in general, are a powerful component for obtaining realistic qualitative data on excitation qualities in a variety of chemical and pharmacological processes [39,41–44]. Furthermore, FMO-derived chemical descriptors have been employed to estimate biological activities [45–54]4-dihydro-[1,2,4]triazole-3-thione was synthesized and structurally characterized by elemental analysis, FT-IR, Raman, ¹H and ¹³C-NMR and UV–Vis studies. A density functional theory (DFT. Likewise, the FMOs energy levels and the energy gaps may influence the kind and amount of binding during their interactions with receptors. As a result, nonbonding intermolecular interactions such as hydrophilic interactions and H-bonding occur with the receptor.

Table 1 presents many estimated thermodynamic molecular descriptors, including dipole moment (μ), electronegativity (χ), charge transfer prevention extent, global hardness (η), and electrophilicity (ω) determined from electronegativity and chemical hardness values.

Table 1.

Dipole moment and chemical reactivity descriptors (Debye, μ) of the prepared compounds 3 and 7.

Parameter	3a	3b	7
LUMO	−2.17	−1.97	−2.23
HOMO	−6.96	−6.96	−6.96
ΔELUMO-HOMO	4.80	5.00	4.74
$\chi =-\frac{1}{2}(E_{HOMO}+E_{LUMO})$	4.57	4.47	4.60
$\eta =-\frac{1}{2}(E_{HOMO}-E_{LUMO})$	2.40	2.50	2.37
$\delta =\frac{1}{\eta }$	0.42	0.40	0.42
$\omega =\frac{{\chi }^2}{2\eta }$	4.34	3.99	4.46
I=-EHOMO	6.96	6.96	6.96
A=-ELUMO	2.17	1.97	2.23
μ	4.87	5.87	6.87

3. Results and discussion

Compounds 2–7 were prepared as displayed in Schemes 1 and 3. The structures of the newly synthesized organic compounds were confirmed using ¹H, ¹³C NMR, and mass spectra.

Scheme 1.

Synthesis of nucleoside (3H)-6,7-difluoro-2-methyl-4-quinazolinone

Scheme 3.

Synthesis of nucleoside 6,7-difluoro-2-methyl-4-quinazolinone.

Benzoxazinone compounds can be prepared by treatment of anthranilic acid derivatives with acid chloride [1] or acetic anhydride [2,55]5-dimethyl-2-thiazolyl. The 6,7-difluoro-4-benzoxazinone 2 was prepared from the reaction of 2-amino-4,5-difluorobenoic acid 1 with acetic anhydride for 1 h. Treatment of compound 2 with ammonia solution for 6 h afforded the (3H)-6,7-difluoro-2-methyl-4-quinazolinone 3, Scheme 1.

Weddige identified the tautomeric characteristics of (3H) 4-quinazolinones, which could exist in three tautomeric forms. The existence of 4-hydroxy quinazoline was displayed by its stability in aqueous alkali at pH 12 to give the anion form. The 4-quinazolinones usually do not dissolve in alkali, mainly when a substitute is present on N1 or N3 [15]. The compound, 6,7-difluoro-2-methyl-4-quinazolinone is expected to have three tautomeric forms of (1H) 6,7-difluoro-2-methyl-4-quinazolinone 3a, 3b, and 3c, Scheme 2.

Scheme 2.

Tautomeric properties of (1H)-6,7-difluoro-2-methyl-4-quinazolinone 3a,3b, and 3c.

A mixture of 3a, dry HMDS, and a catalytic quantity of (NH₄)₂SO₄ was heated under reflux for 24 h (TLC). The product was evaporated to dryness in anhydrous media to afford the silylated derivative 4 as an intermediate compound. Compound 5 was treated with trimethylsilyl triflate; (TMSOTf as a catalyst), an aqueous NaHCO_3, and silica gel column chromatography with ethyl acetate and chloroform (2:98) to obtain 6 as a sticky yellow material. The ¹H NMR of the protected nucleoside 6 shows a doublet signal at δ = 6.1 ppm allocated to the anomeric protons of the ribose moiety with a J coupling constant equal to 4.25 Hz matches the 1′-proton. The spectra appeared as multiple signals of the configuration at δ = 8.06–7.13 ppm due to benzoyl groups and quinazolinone protons in an aromatic region, see Scheme 3 and Figure 1.

Figure 1.

Calculated ¹H and ¹³C NMR of an optimized molecular geometry of compound 7 in deuterated methanol solvent. MEP graph is also presented.

Deprotection of the benzoyl group of protected nucleoside 6 was achieved in sodium metal in dry methanol at room temperature for 24 h to give the corresponding free nucleoside 7. The ¹H NMR of 7 showed the prospective base moiety protons and the sugar moiety protons, though no signal for benzoyl protons appeared. Also, the ¹H NMR spectrum of 7 shows a doublet at δ = 6.9 ppm allocated to the anomeric proton of the ribose moiety with J coupling constant equal to 4.25 Hz that matches the 1′-proton the β-configuration. With the appearance of aromatic group complex signals, a singlet peak appeared at δ = 7.88 ppm allocated to the H-5, and another singlet signal at δ = 6.94 ppm assigned to the H-8. The ¹H NMR of 7 showed the predicted base moiety protons in addition to the sugar moiety protons. Calculated ¹H, and ¹³C NMR of an optimized molecular geometry of compound 7 in deuterated methanol solvent are provided for comparison (Figure 1; experimental section). Compound 7 was confirmed using mass spectra which showed a molecular ion peak ion (M⁺) at m/z = 327.21 (2.77%), (calc. 328.27) for molecular formula C₁₄H₁₄F₂N₂O₅, the base peak = 319.18 (100).

3.1. DFT theoretical calculations

3.1.1. Molecular geometry

The optimal molecular structures were predicted using DFT calculations at the B3LYP 6311G (d, p) basis set to assess the stability of the expected positional isomers. To estimate the most stable positional isomer of the produced compound 3, calculations were performed for the proposed two isomers, 3a and 3b. These computations entailed doing a geometry structure optimization on each isomer to obtain the least energy structure, then calculating the frequency at the optimized geometry. In addition, several thermochemical parameters were calculated (Figure 2; Table 2). Free energy (G) and enthalpy (H) were calculated to determine the relative stabilities, corrected energy, and thermodynamic parameters of both positional isomers of compound 3.

Figure 2.

Calculated optimized molecular geometry of both isomers of compound 3.

Table 2.

Thermal parameters (hartree/particle) of both positional isomers of compound 3.

Parameter	3a	3b
Ecorr	0.139237	0.139796
ZPVE	−730.840481	−730.858147
Etot	−730.829206	−730.847069
H	−730.828262	−730.846124
G	−730.878053	−730.895234
ΔE in kcal/mol	11.209	0.00

The DFT calculations revealed that the 3b isomer had the lowest energy structure and the most stability in comparing both geometrical isomers. The 3b isomer, on the other hand, is the least stable. The energy difference between both isomers is 11.2 kcal/mol. However, the amid derivative 3b higher stability could be illustrated in its tautomeric ability with the conjugated C=O group (Figure 3).

Figure 3.

Tautomeric isomerism of compound 3.

On the other hand, the intermolecular H-bonding could affect the stability of the predicted stable isomer 3b. Isomer 3b can form intermolecular H-bonding with two strong H-bonds than the other isomer 3a. The formation of two H-bonds will enhance the stability of the amid isomer 3b more than the other isomer (Figure 4).

Figure 4.

Intermolecular H-bonding of amid isomer 3b.

3.2. Docking analysis

This docking study investigates how a nucleoside ligand might interact in the active site of the main protease (M^pro; PDB code 7BQY, 1EVE) for Alzheimer’s disease and COVID-19 [15]. Hydrogen bonding, hydrophobic interactions, and other factors, e.g., entropy and solvation, can control the structural reorganization of both the ligand and the receptor upon binding. It remains challenging to predict the conformational changes of 7BQY, 1EVE, and the ligand, as both are expected to display different degrees of adjustment after binding [33]. The docked molecule in 7BQY is shown in Figure 5. The confirmation of all molecules has been demonstrated with respect to the known medicine Remdesivir Figure 5. It has been noticed that the Met165(A), Arg188(A), Gln189(A), and Gly143(A) amino acids are the common residues, forming hydrophobic interactions among all the compounds used in this study. At the same time, Remdesivir forms four hydrogen bonds (Leu141(A), Asn143(A), Ser144(A) and Glu166(A)). Similarly, compound 7 also forms 4 hydrogen bonds ((Leu141(A), Ser144(A), His163(A) and Glu166(A)). N3 displays three hydrogen bonds (Leu41(A), Asn142(A) and His163(A)). Compounds 3a and 3b show one and two hydrogen bonds, respectively, Figures 6 and 7. A superposition of compounds 3a, 3b, 7, remdesivir drug, and N3 inhibitor docked into the binding pocket of 7BQY using the identical parameters for a fair comparison. The outcomes are presented in Figure 5. The number and type of interactions between the remdesivir, N3, 7, 3a, 3b and the main protease (7BQY) are summarized in Figure 6 and partly in Figure 7. The laydown of all docked ligands is presented together (with and without 7BQY), showing their fitting in the same active site but different positioning (Figure 5). Also, the display of the superposition of each ligand compared to remdesivir is displayed for easy comparison. This is represented by PyMOL by Schrödinger [56].

Figure 5.

The superposition of compounds 3a, 3b, 7, remdesivir drug, and N3 inhibitor docked into the binding pocket of 7BQY using the identical parameters for a fair comparison (circled, middle bottom window). For clarity, all docked ligands are shown in the top central window without the 7BQY, M^pro. Finally, the display of each ligand after docking is compared to remdesivir (orange in color). This is represented by PyMOL by Schrödinger (downloaded from https://pymol.org/2/) [56].

Figure 6.

Number and type of interactions between the remdesivir, N3, 7, 3a, 3b, and the main protease (7BQY).

Figure 7.

(a) A schematic 2D LIGPLOT illustration of remdesivir and the main protease 7BQY. The purple stick at the center represents remdesivir, whereas the remaining are the amino acids residues. The green dashes are the hydrogen bond, and the thin red dashes and the spoked arcs pointing towards the ligand represent the hydrophobic residue bonds with remdesivir. The active site residues: Leu141(A), Asn143(A), Ser144(A), and Glu166(A) are involved in making a hydrogen bond with the ligand. Atoms marked by spokes in the ligand or protein are involved in the interactions without the red lines. (b) A schematic 2D LIGPLOT illustration of compounds 7 and 7BQY represented by PyMOL [56].

3.3. Binding prediction of compound 7 compared to donepezil using molecular docking

Docking analysis of compound 7 and donepezil was carried out on the Torpedo California acetylcholinesterase to study their binding affinity (TcAChE) (PDB, 1EVE), [57–62]. Compound 7 demonstrated a close-fitting binding against the TcAChE enzyme with close binding energy of −9.7 kcal/mol compared to donepezil (−10.9 kcal/mol) (Figures 8 and 9). The molecular structure of 7 and donepezil after binding with the protein are shown in Figures 8 and 9. Compound 7 shows conventional hydrogen bonds, fluorine-hydrogen bonds, carbon-hydrogen bonds, and van der Waals interaction. In contrast, the donepezil drug does not exhibit any of these interactions, indicating probably considerable flexibility of compound 7 that facilitates the tight interaction with the binding site for the quaternary nitrogen of AChE enzyme [33]. Donepezil drug demonstrates mainly p-s and p-p stacked interactions (Figure 9). Superimposition of 7 (blue line) and donepezil (orange line) is shown in Figure 8, which demonstrates the benzylpiperidine group oriented towards PHE, TYR, and TRP residues. Similarly, the quinazoline moiety in compound 7 is directed towards PHE, TYR, and TRP residues. The inden-1-one group of donepezil oriented towards TRP, ARG, and LEU residues (Figure 9), whereas the hydrofuran moiety in 7 is directed towards SER, TYR, and Val residues. TRP84 interacts via π-sigma interaction with a distance of 3.37 Å. In contrast, in the case of donepezil, it exhibits a π - π stacking with a distance of 4 Å (Figure 9). Three conventional hydrogen bonds were found between the hydroxyl groups and SER81, TYR70, and TYR121 residues with distances of 3.3, 3.72, and 5.95 Å, respectively (Figure 9). Interestingly, as can be seen, at the bottom of the gorge, a C=O···HIS440 hydrophobic interaction was formed in the active site of AChE [63]. This interaction was not detected in the donepezil drug case. This might enhance the affinity to the enzyme and therefore improve the inhibition effect of compound 7.

Figure 8.

(a) The superposition of compound 7 and donepezil docked to 1EVE using the identical parameter for a reasonable comparison; (b) the display of donepezil drug after docking; (c) the display of both donepezil drug and compound 7; (d) the display of compound 7. Results are presented using PyMOL [56].

Figure 9.

2D representations of interactions in (a) TcAChE- 7 complex; (b) in TcAChE-donepezil complex. Distances are in A^o.

3.4. Toxicity prediction (in silico) for our ligands in comparison to remdesivir and N3

ProTox-II virtual lab was used to predict the toxicity of our small molecules in comparison to the authorized drug remdesivir [64–70] and N3 inhibitor [31]. The oral toxicity presented as lethal dose (LD) at 50% (LD50) milligrams per kilograms weight of the test population. ProTox-II predicted the toxicity classes as class 3 for 3a and 3b and LD50 of 200 mg/kg with the same average similarity and prediction accuracy of 60.65% and 68.07%, respectively (Table 3). Interestingly, the toxic activity of 7 is predicted as class 4, which is similar to the approved drug, remdesivir, with LD50 of 1000 mg/kg. The average similarity and prediction accuracy of 7 and remdesivir were 66.03%, 68.07%, and 40.93%, 54.26%, respectively (Table 3). Compound N3 was predicted as the lowest toxic compound in this study as class 5 with LD50 of 4000 mg/kg and average similarity and prediction accuracy of 45.06% and 54.26, respectively (Table 3).

Table 3.

Acute oral toxicity predicted by ProTox-II web server for ligands 3a, 3b, 7, remdesivir, and N3.

Ligands drug and N3	Oral toxicity prediction results
	Predicted LD50 (mg/kg)	Predicted toxicity class	Average similarity (%)	Prediction accuracy (%)
3a	200	3	60.65	68.07
3b	200	3	60.65	68.07
7	1000	4	66.03	68.07
Remdesivir	1000	4	40.93	54.26
N3	4000	5	45.06	54.26

The ProTox-II web server can also predict organ toxicity. For example, the hepatoxicity estimation of the three ligands 3a, 3b, the approved medicine for COVID-19 remdesivir and N3 inhibitor were all not active. In contrast, 7 was predicted as a functional ligand on organ toxicity (Lever) with a probability of 0.52. Predicted activities for all studied ligands and the controls (remdesivir and N3) were inactive (noncarcinogenic, non immunotoxin, nonmutagenic, noncytotoxic) (Table 4).

Table 4.

Organ toxicity and toxicological endpoints predicted activity calculated using the ProTox-II web server for ligands 3a, 3b, 7, remdesivir, and N3.

Ligands drug and N3	Hepatotoxicity		Carcinogenicity		Immunotoxicity		Mutagenicity		Cytotoxicity
	Activity	Probability	Activity	Probability	Activity	Probability	Activity	Probability	Activity	Probability
3a	x	0.54	x	0.51	x	0.99	x	0.64	x	0.90
3b	x	0.54	x	0.51	x	0.97	x	0.64	x	0.90
7	√	0.52	x	0.66	x	0.96	x	0.77	x	0.77
Remdesivir	x	0.56	x	0.55	x	0.90	x	0.62	x	0.55
N3	x	0.58	x	0.52	x	0.96	x	0.61	x	0.67

Figure 10 depicts the prediction of the FMOs energy levels and their energy gap of 3a and 3b and its nucleoside 7. The energy levels of the FMOs in compound 3a are lower than those in compound 3b, which could be explained in terms of conjugation. The presence of an H-atom adjacent to the C=O group enhances the tautomerism with the OH. It may decrease the conjugation of the C=O group with the ring, inhibiting the conjugation with the benzene ring. This effect on MO levels has an impact on the chemical descriptors that are used to illustrate the biological activity of these compounds. However, the attachment of the hydrophilic carbohydrate moiety of the isomer 3a affects the energy difference between the orbitals with a small value ΔE = 4.74 e.V. Also, it is evident that the isomerism highly affects the level of the LUMO than the HOMO. Isomer 3b shows the lowest-lying LUMO. The low-lying LUMO orbital of 3b could predict its ability to accept the electron than its tautomer and its nucleoside derivative 7. However, the higher topological, polar surface area, lower hydrophobicity, in silico absorption and high H-bonding acceptor percent of compound 7 could illustrate its high predicted biological activity against the inhibition of COVID 19 and Alzheimer infections.

Figure 10.

Molecular orbital distribution and localization for FMOs of 3a, 3b, and 7.

4. Conclusions

Synthesis of some of the ribosylation of silated compound 4 with 1-O-acetyl-2,3,5-tri-O-benzoyl-β-D-ribofuranose 5 gave β-anomeric of the benzoylated nucleoside derivatives 6. Deprotection of the latter using dry absolute methanol and sodium metal gave the new free N-nucleosides 7, in moderate yields. The new compounds obtained have been characterized by their spectral analysis. The prediction analysis using the Pro Tox -II server showed that compound 7 has a similar feature to the approved antiviral drug remdesivir for COVID-19. Compound 7 behaved similarly in all tests except on hepatotoxicity. This suggests that compound 7 may be worth additional study in the context of a possible drug for COVID-19. A molecular modeling investigation confirmed, in silco, that our nucleosides 7 is an excellent binding inhibition of AChE enzyme. Compound 7 could be a possible effective drug against Alzheimer’s disease. Finally, DFT was used to demonstrate molecular geometry and the thermodynamic parameters that could be used to illustrate the MD results in terms of the structural activity correlation computed from FMOs using molecular descriptors. Compound 7 demonstrated its predicted activity towards binding inhibition of AChE enzyme and Alzheimer’s disease due to higher hydrophilicity, a larger topological polar surface area, and a strong H-bonding acceptor.

Funding Statement

This research was funded by Researchers Supporting project Taif, Kingdom of Saudi Arabia (KSA, grant number TURSP-2020/137).