A Combination of Structure-based Virtual Screening and Experimental Strategies to Identify the Potency of Caffeic Acid Ester Derivatives as SARS-CoV-2 3CLpro Inhibitor from an In-house Database

doi:10.21203/rs.3.rs-3125927/v1

Download PDF

Research Article

A Combination of Structure-based Virtual Screening and Experimental Strategies to Identify the Potency of Caffeic Acid Ester Derivatives as SARS-CoV-2 3CLpro Inhibitor from an In-house Database

https://doi.org/10.21203/rs.3.rs-3125927/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Drug development requires significant time and resources, and computer-aided drug discovery techniques that integrate chemical and biological spaces offer valuable tools for the process. This study focused on the field of COVID-19 therapeutics and aimed to identify new active non-covalent inhibitors for 3CL^pro, a key protein target. By combining in silico and in vitro approaches, an in-house database was utilized to identify potential inhibitors. The drug-likeness criteria was considered to pre-filter 553 compounds from 12 groups of natural products. Using structure-based virtual screening, 296 compounds were identified that matched the chemical features of SARS-CoV-2 3CL^pro peptidomimetic inhibitor pharmacophore models. Subsequent molecular docking resulted in 43 hits with high binding affinities. Among the hits, caffeic acid analogs showed significant interactions with the 3CL^pro active site, indicating their potential as promising candidates. To further evaluate their efficacy, enzyme-based assays were conducted, revealing that two ester derivatives of caffeic acid (4k and 4l) exhibited more than a 30% reduction in 3CL^pro activity. Overall, these findings suggest that the screening approach employed in this study holds promise for the discovery of novel anti-SARS-CoV-2 therapeutics. Furthermore, the methodology could be extended for optimization or retrospective evaluation to enhance molecular targeting and antiviral efficacy of potential drug candidates.

COVID-19 therapeutics

3CLpro inhibitors

caffeic acid analogs

structure-based virtual screening

enzyme inhibition assay

In December 2019, Chinese health authorities reported the emergence of a disease, COVID-19, later declared a pandemic by the World Health Organization [1]. It is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and has wide-ranging global social and economic impacts [2]. The duration of the epidemic was initially uncertain due to several factors, such as different modes of transmission, a large number of asymptomatic carriers, limited availability of affordable tests, lack of effective therapies, and the resistance of variants to vaccines. The presence of various modes of transmission poses a challenge in implementing prevention and control measures [3, 4]. Effective treatments are needed for individuals who are not vaccinated or may not respond sufficiently to vaccination, particularly immunocompromised patients [5]. These treatments should also address viral mutations that could potentially reduce vaccine effectiveness.

The SARS-CoV-2 3-chymotrypsin-like protease (3CL^pro) has been identified as an attractive target for resolving the issue, as it is known to play a vital role in viral replication and there are no human proteases with similar cleavage specificity [6]. The first target-based drugs developed are 3CL^pro peptidomimetic inhibitors (Fig. 1), for example, aldehyde inhibitor (11a) [7], α-ketoamide inhibitors (13b) [8], and Michael acceptor inhibitor (N3) [9] which are identified by PDB code 6Y2F, 6LU7, and 6LZE, respectively. Nirmatrelvir inhibiting the 3CL^pro activity by binding tightly to the active site has been authorized for emergency use or granted regulatory approvals in several countries for treating COVID-19 in certain patient populations [10]. Although nirmatrelvir has curative properties, it must be co-administered with ritonavir to achieve adequate plasma concentrations. Beyond that, administering ritonavir-boosted nirmatrelvir may not be the best choice for all patients due to the possibility of significant drug-drug interactions with concomitant medications [11]. Additionally, the presence of the single mutations E166M and H172Y in the SARS-CoV-2 3CL^pro significantly reduced its inhibitory activity, which affects drug resistance and treatment outcomes [12, 13].

Natural and synthetic compounds are vital in pharmaceutical studies due to their broad-spectrum activities, including anticancer, anti-inflammatory, antioxidant, and antiviral properties [14]. Natural compounds derived from plants, marine organisms, and microbes have a history of use in traditional medicine for their inherent antiviral properties. For example, Ang et al. highlighted the potential 3CL^pro inhibitors derived from indole and benzene derivatives, such as β-carbolines, benzoic acid esters, and quinazolinamines. They employed a combination of ligand-based and structural-based hybrid virtual screening and provided in vitro experimental evidence to confirm the inhibitory effects on the activity of SARS-CoV-2 3CL^pro [15]. Similarly, Wasilewicz et al. reported a plausible mode of action for sanggenons against SARS-CoV-2 3CL^pro based on an in silico study, which was experimentally validated for their inhibitory activity in enzyme-based assays and their antiviral effect in Calu-3 cells [16]. Meanwhile, synthetic compounds offer the advantage of customization, allowing scientists to design molecules tailored for specific antiviral applications. Hydroxy-xanthones, α-mangostin, and piperine have also been identified as potential inhibitors of the 3CL^pro [17–19]. By combining the strengths of natural and synthetic compounds, we expedite the development of effective antiviral therapies, contributing to the ongoing research on the treatment and prevention of SARS-CoV-2.

Virtual screening techniques aid in identifying potential antiviral molecules [20]. Structure-based techniques, which focus on candidate-target interactions, are particularly effective in discovering novel chemical structures [21, 22]. While computational approaches and simulations have advanced pharmacological research [23], it is crucial to validate findings with experimental evidence to ensure the reliability of conclusions. In this study, we employed combined in silico and in vitro approaches to identify effective inhibitors for SARS-CoV-2 3CL^pro from a series of natural and synthetic compounds. Initially, a screening of 553 compounds from our in-house database was conducted based on drug-likeness properties. Pharmacophore-based virtual screening and molecular docking techniques were utilized, using previously reported peptidomimetic inhibitors (11a, 13b, and N3) as references, to narrow the selection to promising molecules. Intermolecular interactions and center of mass (COM) calculations were performed to assess the binding capability of the compounds at the 3CL^pro active site. The inhibitory activity of the hit compounds was examined through an in vitro enzymatic assay, while ligand-protein binding analysis was employed to gain a better understanding of their potential effects.

The in-house database focused in this study contains 553 natural and synthetic compounds as classified into 12 primary groups: 4 α-mangostins [19], 18 anticancer agents [24], 9 avicequinones [25], 42 caffeic acid derivatives [26], 159 acid homodimers, 92 xanthones, 78 avenalumic acid derivatives, 9 quinonoids, 30 steroids [27], 6 piperic acid derivatives [18], 53 renieramycins [28], and 53 ureas [29]. The Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Chulalongkorn University developed and provided these compounds.

Drug-likeness analysis

The pharmacokinetics, drug-like properties, and medicinal chemistry friendliness of the 553 compounds from an in-house database were evaluated by SwissADME web-based interface [30]. The compounds were assessed based on Lipinski's rule of five criteria [31], physicochemical properties, and drug-likeness profiles [32]. A radar chart was constructed using the number of compounds meeting each criterion, as listed in Table S1 of the Supplemental Information (SI). The compounds that fulfilled all predicted properties were chosen for pharmacophore-based virtual screening.

Structure-based pharmacophore modeling

The previous studies provided the representative pharmacophore models (RPMs) of inhibitors 11a, 13b, and N3 in complex with 3CL^pro, along with their predicted chemical features and spatial arrangement required for a molecule to interact with the protein target [33]. These RPMs and drug-like compounds were defined as input datasets. The pharmacophore-based virtual screening was performed using the iScreen node of the KNIME program [34] with default parameters. The results of output compounds were then subjected to model validation to ensure the model’s reliability and accuracy and to facilitate the interpretation and prioritization of virtual screening results [35].

The active compounds were inserted into the DecoyFinder [36] to recognize decoy molecules based on datasets downloaded from the ZINC [37] and DrugBank [38]. KNIME extensions [34] were additionally utilized for model validation. Its input datasets comprise representative pharmacophore models, known active compounds, and a set of decoys. The validation results, including metrics such as receiver operating characteristic (ROC) curves or enrichment factors (EF), are analyzed using KNIME nodes for data analysis and visualization.

Molecular docking

The crystal structures of SARS-CoV-2 3CL^pro with bound peptidomimetic inhibitor 11a, 13b, and N3 were obtained from the Protein Data Bank (https://www.rcsb.org) with PDB codes 6Y2F, 6LU7, and 6LZE, respectively. Water molecules and other ligands in the complex were removed using Accelrys Discovery Studio Client 4.0 [39]. The structure of the protein and ligand were extracted and saved in a separate file. The protonation state of ionizable amino acids and ligands at neutral pH 7.4 was checked by PDB2PQR [40] and MarvinSketch [41], respectively.

To prepare for the redocking simulations, the protein and ligand files were converted to PDBQT format using AutoDock Tools software [42]. These files were then used as input for AutoDock VinaXB [43], with specific docking parameters configured. The grid box with a dimension of 20×20×20 Å was centered. The docking process generated output files containing docked poses, root mean square deviation (RMSD), and binding free energies (ΔG). From redocking, the RMSD values obtained for models 6Y2F, 6LU7, and 6LZE were 1.25 Å, 1.67 Å, and 1.57 Å, respectively. The same procedure was also applied to selected active compounds from pharmacophore-based virtual screening. Only hit compounds presenting ΔG lower than the known inhibitors (11a, 13b, and N3) were chosen for ligand-protein binding analysis using Accelrys Discovery Studio Client 4.0. To describe the compound binding strength, the distance measured from the center of mass of each classified group to the center of mass of H41 and C145 catalytic dyad (d_COM) was also taken into consideration using ChimeraX [44].

Preparation of ester derivatives of caffeic acid

Caffeic acid and reagents were purchased from Sigma-Aldrich (Missouri, USA). Analytical grade solvents were obtained from Carlo Erba (Michigan, USA) and Honeywell (North Carolin, USA) and distilled before use. Anhydrous solvents were dried over 4 Å molecular sieves. Reactions were carried out in oven-dried glassware and magnetically stirred under a nitrogen atmosphere using a small balloon unless otherwise noted. Room temperature was 25°C unless otherwise stated. Thin-layer chromatography (TLC) was used to monitor all reactions by utilizing aluminum silica gel 60 F254 from Merck (Darmstadt, Germany) and observed under ultraviolet (UV) light at 254 nm. Flash column chromatography was also used to purify all synthetic compounds using silica gel (60 Å, 230–400 mesh) as the stationary phase and the suitable mixtures of ethyl acetate, dichloromethane, and hexane as the mobile phases. Spectroscopic methods were used to elucidate the structures of all synthetic compounds. Nuclear magnetic resonance (NMR) spectra, both proton (¹H) and carbon (¹³C), were measured on Bruker Avance NEO 400 MHz spectrometer. ¹H-NMR chemical shifts are reported as δ values in ppm relative to residual CHCl₃ (7.27 ppm). ¹H-NMR coupling constants (J) are reported in Hertz (Hz). Unless otherwise indicated, deuterochloroform (CDCl₃) served as an internal standard (77.0 ppm) for all ¹³C spectra. Infrared (IR) spectra were recorded on Perkin Elmer Frontier Fourier-transform IR Spectrometer. High-resolution mass spectra were performed on Bruker Daltonics microTOF mass spectrometer.

Synthesis of caffeic acid methyl ester (4k) and 3-hydroxyl-4-methoxy cinnamic acid methyl ester (4l). Caffeic acid (40 mg, 0.22 mmol) was added to an oven-dried round-bottomed flask and dissolved in a 4 mL mixture of dry tetrahydrofuran: dry methanol (1:1) under the inert atmosphere. The reaction mixture was cooled at 0°C using an ice bath. Then trimethylsilyl diazomethane (0.6 M in hexane, 1.1 mL, 0.66 mmol) was slowly added dropwise to the reaction mixture until the yellow color persisted. The reaction was stirred at room temperature for 30 minutes. The reaction was monitored by TLC. After that, the volatile solvent was removed under reduced pressure to obtain the crude product. The synthesized compounds were purified by column chromatography with silica gel as the stationary phase and the mixture of ethyl acetate and hexane as the mobile phase. The chemical structures of the purified compounds were elucidated by spectroscopic techniques. Caffeic acid methyl ester (4k) was obtained as the brown oil at 15 mg (35%). ¹H-NMR (CDCl₃, 400 MHz) δ 7.58 (1H, d, J = 16.0 Hz, 7-H), 7.08 (1H, d, J = 2.0 Hz, 3-H), 7.00 (1H, dd, J = 8.4, 2.0 Hz, 5-H), 6.87 (1H, d, J = 8.0 Hz, 6-H), 6.26 (1H, d, J = 16.0 Hz, 8-H), 3.79 (3H, s, 10-H) ppm; IR (ATR, ν_max) 3361, 2927, 1698, 1634, 1601, 1511, 1263, 1159, 854, 807 cm^− 1 [45]. 3-Hydroxyl-4-methoxycinnamic acid methyl ester (4l) was obtained as the brown oil at 32 mg (70%) ¹H-NMR (CDCl₃, 400 MHz) δ 7.58 (1H, d, J = 16.0 Hz, 7-H), 7.00 (1H, dd, J = 8.4, 2.0 Hz, 5-H), 6.88 (1H, d, J = 8.4 Hz, 3-H), 6.81 (1H, d, J = 7.2 Hz, 6-H), 6.26 (1H, d, J = 16.0 Hz, 8-H), 3.87 (3H, s, 10-H), 3.76 (3H, s, 11-H) ppm; IR (ATR, ν_max) 3354, 2949, 1692, 1631, 1599, 1509, 1258, 1157, 852, 807 cm^− 1; HRMS–ESI m/z 231.0625 ([M + Na]⁺, calcd for C₁₁H₁₂NaO₄ 231.0628) and 209.0807 ([M + H]⁺, calcd for C₁₁H₁₃O₄ 209.0808).

Synthesis of 3,4-diacetyloxy cinnamic acid (4n). Caffeic acid (500 mg, 2.78 mmol), acetic anhydride (5 mL, 52.82 mmol), and pyridine (0.1 mL) were added to an oven-dried round-bottomed flask. Then, the reaction mixture was refluxed at 120°C for 5 h and monitored by TLC. After that, the reaction was cooled in the ice bath and quenched by adding water. The product was precipitated and filtered to obtain 4n as a pale light brown solid at 589.3 mg (80%). ¹H-NMR (CDCl₃, 400 MHz) δ 7.74 (1H, d, J = 16.0 Hz, 11-H), 7.46 (1H, dd, J = 8.4, 2.0 Hz, 7-H), 7.41 (1H, d, J = 2.0 Hz, 5-H), 7.27 (1H, d, J = 8.4 Hz, 8-H), 6.42 (1H, d, J = 16.0 Hz, 12-H), 2.34 (3H, s, 1-H), 2.33 (3H, s, 10-H) ppm; ¹³C-NMR (CDCl₃, 100 MHz) δ 170.8 (C-13), 168.1 (C-9), 168.0 (C-2), 145.0 (C-11), 143.9 (C-3), 142.5 (C-4), 132.9 (C-6), 126.7 (C-7), 124.0 (C-8), 123.0 (C-5), 118.3 (C-12), 20.7 (C-1), 20.6 (C-10); IR (ATR, ν_max) 2822, 2522, 1755, 1674, 1628, 1328, 1290, 1251, 878. 824 cm^− 1; HRMS–ESI m/z 265.0702 ([M + H]⁺, calcd for C₁₃H₁₃O₆ 265.0707).

Synthesis of 3,4-diacetyloxy cinnamic acid-2,2-dimethyl-1-propyl ester (4p). Caffeic acid (30 mg, 0.11 mmol) and 2,2-dimethyl-1-propanol (9.69 mg, 0.11 mmol) were added to an oven-dried round-bottomed flask and dissolved in 20 mL dry dichloromethane under the inert atmosphere. The reaction mixture was cooled at 0°C in an ice bath. Then, 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide (EDCI HCl: 24.9 mg, 0.13 mmol), hydroxybenzotriazole (HOBt: 17.6 mg, 0.13 mmol), and N,N-diisopropylethylamine (DIPEA: 0.038 mL, 0.22 mmol) were added to the reaction mixture. The reaction was stirred at room temperature for 4 h. TLC was used to monitor the reaction. After that, the volatile solvent was removed under reduced pressure to obtain the crude product. The synthesized compounds were purified by column chromatography with silica gel as the stationary phase and dichloromethane as the mobile phase to obtain 4p as a white amorphous solid at 3.9 mg (11%). ¹H-NMR (CDCl₃, 400 MHz) δ 7.65 (1H, d, J = 16.0 Hz, 11-H), 7.44 (1H, dd, J = 8.4, 2.0 Hz, 7-H), 7.40 (1H, d, J = 2.0 Hz, 5-H), 7.25 (1H, d, J = 8.4 Hz, 8-H), 6.44 (1H, d, J = 16.0 Hz, 12-H), 3.93 (2H, s, 14-H), 2.33 (3H, s, 1-H), 2.33 (3H, s, 10-H), 1.02 (9H, s, 16-H) ppm; ¹³C-NMR (CDCl₃, 100 MHz) δ 168.1 (C-9), 168.0 (C-2), 166.7 (C-13), 143.4 (C-3), 142.6 (C-11), 142.4 (C-4), 133.4 (C-6), 126.4 (C-7), 123.9 (C-8), 122.7 (C-5), 119.5 (C-12), 74.0 (C-14), 68.2 (C-15), 26.5 (C-16), 20.7 (C-1), 20.6 (C-10) ppm; IR (ATR, ν_max) 2921, 2851, 1763, 1658, 1632, 1368, 1252, 1209, 1172, 1146, 1106, 874, 836, 803 cm^− 1; HRMS–ESI m/z 357.1319 ([M + Na]⁺, calcd for C₁₈H₂₂NaO₆ 357.1309) and 335.1492 ([M + H]⁺, calcd for C₁₈H₂₃O₆ 335.1489).

Synthesis of 3,4-diacetyloxy cinnamic acid-2,5-bis(trifluoromethyl) phenyl ester (4u). Caffeic acid (30 mg, 0.11 mmol) and 2,5-bis(trifluoromethyl)phenol (25.3 mg, 0.11 mmol) were added to an oven-dried round-bottomed flask and dissolved in 20 mL dry dichloromethane under the inert atmosphere. The reaction mixture was cooled down to 0°C using an ice bath. Then, EDCI HCl (24.9 mg, 0.13 mmol), HOBt (17.6 mg, 0.13 mmol), and DIPEA (0.038 mL, 0.22 mmol) were added. The reaction was stirred at room temperature for 4 h and monitored by TLC. After that, the reaction was stopped by concentration under reduced pressure. The crude product was worked up by redissolving in ethyl acetate and washing with 0.5% HCl, 2.5% NaHCO₃, and brine, respectively, to remove the by-products. The organic layers were combined and dried over anhydrous sodium sulfate. Then, the organic filtrate was dried under reduced pressure to obtain the crude product. The synthesized compounds were purified by column chromatography with silica gel as the stationary phase and the mixture of ethyl acetate and hexane as the mobile phase. Ester 4u was obtained as a white amorphous solid at 7.0 mg (13%). ¹H-NMR (CDCl₃, 400 MHz) δ 7.88 (1H, d, J = 16.0 Hz, 11-H), 7.88 (1H, d, J = 8.5 Hz, 16-H), 7.68 (1H, s, 19-H), 7.66 (1H, d, J = 8.5 Hz, 17-H), 7.53 (1H, dd, J = 8.4, 2.0 Hz, 7-H), 7.48 (1H, d, J = 2.0 Hz, 5-H), 7.31 (1H, d, J = 8.4 Hz, 8-H), 6.60 (1H, d, J = 16.0 Hz, 12-H), 2.35 (3H, s, 1-H), 2.34 (3H, s, 10-H) ppm; ¹³C-NMR (CDCl₃, 100 MHz) δ 168.1 (C-9), 168.0 (C-2), 163.8 (C-13), 148.6 (q, J = 1.0 Hz, C-14), 146.4 (C-11), 144.3 (C-3), 142.6 (C-4), 135.2 (q, J = 34.0 Hz, C-15), 132.5 (C-6), 127.8 (q, J = 5.0 Hz, C-16), 126.9 (C-7), 126.3 (q, J = 33.0 Hz, C-18), 124.2 (C-8), 123.2 (C-5), 122.8 (q, J = 5.0 Hz, C-17), 121.9 (q, J = 4.0 Hz, C-19), 116.8 (C-12), 124.2 & 123.6 & 121.4 & 120.9 (J = 330.0 Hz, C-15a & C18a), 20.7 (C-1), 20.6 (C-10) ppm; IR (ATR, ν_max) 2921, 2851, 1753, 1212, 1176, 1125, 873, 835 cm^− 1; HRMS–ESI m/z 499.0591 ([M + Na]⁺, calcd for C₂₁H₁₄F₆NaO₆ 499.0587).

Synthesis of 3,4-diacetyloxy cinnamic acid-3,4-bis(trifluoromethyl) phenyl ester (4v). Caffeic acid (30 mg, 0.11 mmol) and 3,4-bis(trifluoromethyl)phenol (25.3 mg, 0.11 mmol) were employed under the same esterification protocol as described above. The ester 4v was obtained as a white amorphous solid at 7.5 mg (14%). ¹H-NMR (CDCl₃, 400 MHz) δ 7.93 (1H, d, J = 8.7 Hz, 18-H), 7.88 (1H, d, J = 16.0 Hz, 11-H), 7.71 (1H, d, J = 2.1 Hz, 15-H), 7.57 (1H, dd, J = 8.7, 2.1 Hz, 19-H), 7.51 (1H, dd, J = 8.4, 2.0 Hz, 7-H), 7.47 (1H, d, J = 2.0 Hz, 5-H), 7.31 (1H, d, J = 8.4 Hz, 8-H), 6.59 (1H, d, J = 16.0 Hz, 12-H), 2.35 (3H, s, 1-H), 2.35 (3H, s, 10-H) ppm; ¹³C-NMR (CDCl₃, 100 MHz) δ 168.1 (C-9), 168.0 (C-2), 163.9 (C-13), 153.1 (C-14), 146.2 (C-11), 144.3 (C-3), 142.6 (C-4), 132.5 (C-6), 130.1 (dd, J = 30.0, 1.0 Hz, C-17), 129.6 (q, J = 6.0 Hz, C-18), 126.9 (C-7), 125.5 (dd, J = 32.0, 1.5 Hz, C-16), 125.1 (C-19), 124.2 (C-8), 128.9 & 123.5 & 121.3 & 120.8 (J = 307.0 Hz, C-16a & C17a), 123.2 (C-5), 121.8 (q, J = 7.0 Hz, C-15), 117.1 (C-12), 20.7 (C-1), 20.6 (C-10) ppm; IR (KBr) 2918, 2849, 1763, 1275, 1181, 1141, 903, 841 cm^− 1; HRMS–ESI m/z 499.0595 ([M + Na]⁺, calcd for C₂₁H₁₄F₆NaO₆ 499.0587).

Evaluation of SARS-CoV-2 3CL^pro inhibitory activity

In order to evaluate the enzymatic activity of SARS-CoV-2 3CL^pro using an enzyme inhibition assay, the enzyme was expressed, purified, and stored following similar procedures described for SARS-CoV-1 3CL^pro [46]. A concentration of 0.2 µM of 3CL^pro was used in each experiment. The enzymatic activity was monitored by measuring the initial rate of cleavage of the fluorogenic substrate E(EDANS)TSAVLQSGFRK(DABCYL) (25 µM). The excitation and emission wavelengths were recorded at 340 and 490 nm, respectively. To assess inhibitory activity, the enzymatic activity was examined in the presence of 100 µM of caffeic acid compounds, with a group of control that did not include any inhibitor. The initial rate of the enzymatic reaction in the absence of inhibitors was used as a normalization factor, allowing for the relative assessment of inhibitory effects induced by the caffeic acid compounds.

Evaluation of drug-like properties

Evaluating compounds with drug-like properties served as a crucial filter in the drug discovery process, aiding in identifying potential and safe compounds for further development. The properties under consideration were divided into three principal areas, as shown in Table S1. The result of compounds that meet each drug-like characteristic is depicted in Fig. 3. Most of the compounds from our in-house database do not violate the following criteria as stated by Lipinski's rule: molecular weight ≤ 500 Da, number of H-bond donors ≤ 5, number of H-bond acceptors ≤ 10, and octanol-water partition coefficient (cLogP) ≤ 5 [31]. 490, 488, 549, and 515 compounds meet the criterion of molecular weight, number of H-bond donors, number of H-bond acceptors, and cLogP, respectively. These results indicated that they are more likely to have favorable drug-like properties, especially oral bioavailability and permeability probability [47]. However, it is essential to note that certain compounds fail to meet the criteria for physicochemical properties and drug-likeness profiles, which provide insights into a compound's behavior in biological systems, including solubility, stability, and safety. [30, 48, 49]. To increase the chance of identifying compounds with improved pharmacokinetics, bioavailability, and therapeutic potential for further development, compounds that fulfill all criteria should be selected. Therefore, a total of 358 compounds derived from 11 groups (1 α-mangostin, 14 anticancer agents, 4 avicequinones, 22 caffeic acid derivatives, 129 acid homodimers, 41 xanthones, 65 avenalumic acid derivatives, 5 quinonoids, 21 steroids, 5 piperic acid derivatives, and 51 ureas) were used for the subsequent studies. Their properties predicted by SwissADME are displayed in Table S2.

Pharmacophore-based virtual screening and model validation

In a structure-based virtual screening, drug-like compounds were computationally screened using 11a, 13b, and N3 pharmacophore models generated by K. Sanachai et al. [33] to identify molecules that possess similar chemical features and spatial arrangements. Fig. S1 shows the virtual screening results obtained using KNIME, with the pharmacophore fit score employed as the scoring function. It was demonstrated that 258, 230, and 157 active compounds were identified through screening with models 11a, 13b, and N3, respectively. These compounds were classified into 11 groups, including 1 α-mangostin, 12 anticancer agents, 2 avicequinones, 19 caffeic acid derivatives, 118 acid homodimers, 37 xanthones, 58 avenalumic acid derivatives, 4 quinonoids, 10 steroids, 5 piperic acid derivatives, and 31 ureas. The identified compounds aligned well with the key chemical features, including hydrogen bond donor, hydrogen bond acceptor, and hydrophobic interaction to its corresponding reference model, as previously reported [33]. These compounds were subsequently utilized to find decoys during the model validation step. Due to their higher likelihood of exhibiting the desired biological activity of 3CL^pro, they emerged as promising candidates for further exploration.

It is essential to validate the models following the pharmacophore-based virtual screening to comprehensively assess the model performance. In this study, we gathered the results in ROC curves. In Fig. 4, the area under the ROC curve (AUC) value was calculated by considering that the top 5%, 10%, and 100% of the ranked compounds from each model ranged between 0.5 and 1.0. This suggested that all models have a moderate to strong ability to classify compounds as active or inactive [50] correctly. We also examined the EF for the screened dataset at different percentages. The EF value denotes the degree of enrichment achieved by the screening method compared to a random selection [51]. When the EF value is greater than 1, the screening approach has a higher probability of detecting active compounds than random sampling [52]. Therefore, based on the AUC and EF values evaluation, the N3 model indicated superior discrimination and enrichment compared to the other two models. The 11a model showed moderate discrimination and enrichment, while the 13b model exhibited a lower performance. However, it is necessary to conduct additional analysis and validation to ensure the reliability and robustness of the findings.

Hits identification from molecular docking

The 296 active compounds from pharmacophore-based virtual screening were further studied using molecular docking to predict the most favorable binding mode and affinity between ligand and protein target. Figure 5A illustrates the predicted ΔG in kcal/mol of the best pose compound determined by three criteria: RMSD < 2 Å, low ΔG, and similar binding orientation to their reference compound. As a result, the 43 compounds exhibited a stronger binding affinity with the SARS-CoV-2 3CL^pro than known inhibitors. The observed variations in ΔG can be ascribed to the distinctive interactions between the protein and the ligands, especially hydrogen bonding, electrostatic interactions, hydrophobic interactions, and van der Waals forces. These interactions serve a crucial role in determining the strength and stability of the complex, influencing the overall binding energy. They belong to six groups: 8 caffeic acid derivatives, 10 acid homodimers, 14 xanthones, 6 avenalumic acid derivatives, 2 steroids, and 3 ureas as the number of intermolecular interactions with 3CL^pro plotted in Fig. 5B. Note that Table S3 provides data on the identification between the selected hit compounds as they correspond to the compound number in Fig. 5A. The caffeic acid phenethyl esters compounds can effectively bind to the substrate-binding pocket of SARS-CoV-2 3CL^pro with comparable binding efficacy and binding energies to the known N3 inhibitor [53]. Hengphasatporn et al. reported that xanthone derivatives, especially rubraxanthone, stably attached the 3CL^pro at the allosteric site and potentially the active site [54]. Furthermore, several studies have shown that steroids and ureas, such as glycyrrhizin, withanolides, curcumin, and quercetin, possess potential inhibitory activities against the same target protein [55–58]. These studies further support our computational results, indicating that the identified compounds have the potential to serve as promising hit candidates in inhibiting the activity of the 3CL^pro.

Compounds with sufficient qualitative and quantitative interactions tend to have higher binding ability and can produce more stable complexes [59]. Figure 5B depicts the proportion of intermolecular interactions between each compound and protein target consisting of hydrogen bonding, π-alkyl interaction, halogen interaction, and alkyl-alkyl interaction. The intermolecular hydrogen bonds are highly contributed for derivatives of caffeic acid, acid homodimer, avenalumic acid derivatives, and steroid binding to SARS-CoV-2 3CL^pro. In addition, π-alkyl interactions have a crucial effect on the complementary shape and fit of all screened compounds, particularly xanthones and ureas, within the enzyme active site. The 4 caffeic acid derivatives, 3 avenalumic acid derivatives, steroid, and urea could form the halogen interaction(s), while alkyl-alkyl interactions were likely found in some caffeic acid derivatives, xanthones, and avenalumic acid derivatives.

Figure S2 displays the aligned structure of the selected hit compounds, allowing for the identification of variations and similarities among the compounds within each group when binding to the target protein. The avenalumic acid derivatives and steroids predominantly occupied the S1 and S3 subsites, while the acid homodimer and ureas preferred binding to the S1, S2, and S3 subsites. The caffeic acid derivatives exhibited binding potential to all four subsites: S1, S2, S3, and S1'. In contrast, the xanthones demonstrated variations in their binding interactions with 3CLpro. To further assess how the hit compounds positioned within the active site of the SARS-CoV-2 3CL^pro, we calculated the center of mass for each compound group and measured its distance from the center of mass of the catalytic dyad (H41 and C145), denoted as d_COM. This analysis provided additional insights into the spatial arrangement of the compounds relative to the key catalytic residues. In Fig. 5C, the d_COM value for the focused caffeic acid derivatives is 4.7 Å, indicating significantly closer proximity to the catalytic site than the others (≥ 6.5 Å). These findings suggested that caffeic acid derivatives are well-positioned and likely to establish strong interactions with the 3CL^pro. As a result, we prioritize the synthesis and subsequent experimental testing of caffeic acid derivatives that demonstrate at least 10 ligand-protein interactions.

Synthesis of ester derivatives of caffeic acid

Ester derivatives of caffeic acid were prepared under mild synthetic conditions (Fig. 6). Methylation of caffeic acid (1) provided methyl ester 4k and 4l. Furthermore, the synthesis of diacetyl caffeic acid ester derivatives 4n, 4p, 4u, and 4v was completed through the two-step acetylation and esterification to obtain the corresponding ester derivatives in acceptable yields. Ester 4p, 4u, and 4v were afforded in small quantities according to their steric hindrance of alcohol reagents.

In vitro investigation of enzyme inhibition assay

The inhibitory activities of caffeic acid ester derivatives against the SARS-CoV-2 3CL^pro were evaluated using a well-established in vitro enzymatic activity assay [54, 60]. According to Fig. 7, compounds 4k and 4l had better inhibitory activity than rutin, a reference compound. Compounds 4k and 4l at 100 µM decreased the enzymatic acitivity to 68.8% and 58.0%, respectively. These findings indicate the potential of these two ester caffeic acid derivatives as promising inhibitors of the 3CL^pro, highlighting their suitability for continued investigation and development as antiviral agents targeting SARS-CoV-2.

Figure 8 illustrates intermolecular interactions between the two ester caffeic acid derivatives and the SARS-CoV-2 3CL^pro. Both 4k and 4l compounds exhibited interactions with significant residues H41, F140, L141, N142, S144, H164, and E166, the resemblance with the critical contact residues of the orally active SARS-CoV-2 3CL^pro inhibitors nirmatrelvir [61] and ensitrelvir [62]. Specifically, hydrogen bonds were observed between two hydroxyl groups of the phenyl moiety and residues F140, L141, N142, S144, and H163, while the ester carbonyl oxygen formed a hydrogen bond with H164 and additionally C145 in the case of 4k. Moreover, two π-alkyl interactions with H41 and E166 could stabilize the binding of caffeic acid ester derivatives at the methyl and cinnamyl catechol groups, respectively. The residues F181 and D187 favorably contacted with the ethyl ester of compound 4l via alkyl-alkyl interactions. These detailed interactions provide insights into the specific molecular interactions of two potent compounds with the SARS-CoV-2 3CL^pro active site. Although the potency of these compounds is not as pronounced as others reported from virtual screening [15, 33, 63], the data presented suggest that ester derivatives of caffeic acid could serve as a scaffold to develop more potent inhibitors further.

To enhance and expedite the drug development process, we employed a combination of computer-aided techniques and experimental studies. Our research focused on a collection of 553 natural product compounds and their derivatives, which were organized into 12 distinct groups. These groups included compounds such as α-mangostins, anticancer agents, avicequinones, caffeic acid derivatives, acid homodimers, xanthones, avenalumic acid derivatives, quinonoids, steroids, piperic acid derivatives, renieramycins, and ureas. Through applying drug-likeness criteria, we narrowed the selection to 358 compounds that showed potential. Utilizing pharmacophore-based virtual screening, we compared the chemical features of these 358 compounds to three specific pharmacophore models of peptidomimetic inhibitor of SARS-CoV 3CL^pro: 11a, 13b, and N3. As a result, we identified 296 compounds from the aforementioned groups that matched the characteristics of these pharmacophore models. Next, we conducted molecular docking studies, which revealed 43 hits from six groups (8 caffeic acid derivatives, 10 acid homodimers, 14 xanthones, 6 avenalumic acid derivatives, 2 steroids, and 3 ureas) with a stronger binding affinity towards the SARS-CoV 3CL^pro relative to known inhibitors. These hits underwent visual inspection and analysis of the binding between the compounds and the target protein. Significantly, the six derivatives of caffeic acid exhibited particularly noteworthy results, as they demonstrated a high number of interactions with key residues in the active site and appeared to be positioned closer to the catalytic dyad of the 3CL^pro enzyme. Experimental evaluation of these six synthesized caffeic acid compounds confirmed the favorable inhibitory activity of two ester derivatives, specifically compounds 4k and 4l, against the 3CL^pro enzyme. These findings provide a good starting point for the development of novel therapeutics targeting the SARS-CoV-2 3CL^pro.

Data Availability

The 3D structures of SARS-CoV-2 3CL^pro in complexes with peptidomimetic inhibitor was retrieved from the RCSB protein data bank (https://www.rcsb.org/). Drug-like properties were predicted using SwissADME. LigandScout 4.4.9 and KNIME 4.4.0, licensed programs, were employed for pharmacophore-based virtual screening and model validation. Protein and ligand protonation states were analyzed using PDB2PQR (https://server.poissonboltzmann.org/pdb2pqr) and Marvin 17.21.0 (https://chemaxon.com/marvin), respectively. Molecular docking was performed through AutoDock VinaXB (https://vina.scripps.edu). UCSF ChimeraX 1.4 (https://www.rbvi.ucsf.edu/chimerax/) was utilized for the visualization of ligand-protein binding. GraphPad Prism 9.5.0 (https://www.graphpad.com/) was used to generate all charts. The authors have provided scripts for data analysis and other purposes upon request. For any inquiries or data-related requests, please direct your correspondence to Prof. Thanyada Rungrotmongkol at the following email address: [email protected].

Acknowledgments

This project is funded by the National Research Council of Thailand (NRCT, grant number N42A650231). We gratefully thank the Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Chulalongkorn University, for providing the chemical databases. In vitro enzyme inhibition assay was supported by Thailand Science Research and Innovation Fund Chulalongkorn University (CU_FRB65_hea (71)_134_23_64) (Fundamental Fund 2565, CU) to KW. PaW was supported by the Second Century Fund (C2F), CU. The 90th Anniversary of the Chulalongkorn University Scholarship (GCUGR1125661076M) and the Overseas Academic Presentation Scholarship for Graduate Students (to PP). PaW, PD, and KW are supported by the CU grant to the Center of Excellence for Molecular Biology and Genomics of Shrimp and the Center of Excellence for Molecular Crop. PaW and PD are partially supported by the 90th Anniversary of the CU Scholarship. PD is supported by the Science Achievement Scholarship of Thailand (SAST). The ASEAN European Academic University Network (ASEA-UNINET) for a short research visit. We acknowledge the Vienna Scientific Cluster for providing us with their computing facilities. In addition, our gratitude is extended to the NSTDA Supercomputer Center for providing computing resources for this work.

World Health O (2020) WHO Director-General’s opening remarks at the media briefing on COVID-19.
Wu F et al (2020) A new coronavirus associated with human respiratory disease in China. Nature 579(7798):265–269
Tosta E (2020) Transmission of severe acute respiratory syndrome coronavirus 2 through asymptomatic carriers and aerosols: A major public health challenge. Rev Soc Bras Med Trop 53:e20200669
Jegan R et al (2023) 2019-nCoV effects, transmission and preventive measures: an overview. Z Gesundh Wiss 31(1):27–35
Pullen RL Jr (2022) Managing patients who are immunocompromised during the COVID-19 pandemic. Nursing 52(12):27–32
Zhao Y et al (2022) Structural basis for replicase polyprotein cleavage and substrate specificity of main protease from SARS-CoV-2. Proceedings of the National Academy of Sciences, 119
Dai W et al (2020) Structure-based design of antiviral drug candidates targeting the SARS-CoV-2 main protease. Science 368(6497):1331–1335
Liang J et al (2020) Interaction of the prototypical α-ketoamide inhibitor with the SARS-CoV-2 main protease active site in silico: Molecular dynamic simulations highlight the stability of the ligand-protein complex. Comput Biol Chem 87:107292
Yang H et al (2005) Design of wide-spectrum inhibitors targeting coronavirus main proteases. PLoS Biol 3(10):e324
Joyce RP, Hu VW, Wang J (2022) The history, mechanism, and perspectives of nirmatrelvir (PF-07321332): an orally bioavailable main protease inhibitor used in combination with ritonavir to reduce COVID-19-related hospitalizations. Med Chem Res 31(10):1637–1646
COVID-19 Treatment Guidelines Panel. Coronavirus Disease 2019 (COVID-19) Treatment Guidelines May 30, 2023]; Available from: https://www.covid19treatmentguidelines.nih.gov/
Sasi VM et al (2022) Predicting Antiviral Resistance Mutations in SARS-CoV-2 Main Protease with Computational and Experimental Screening. Biochemistry 61(22):2495–2505
Clayton J et al (2023) Integrative Approach to Dissect the Drug Resistance Mechanism of the H172Y Mutation of SARS-CoV-2 Main Protease. Journal of Chemical Information and Modeling
Chusri S, Singthong P, Kaewmanee T (2015) Antioxidant, anticancer, and cytotoxic effects of Thai traditional herbal preparations consumed as rejuvenators. CyTA - Journal of Food 13(1):40–48
Ang D, Kendall R, Atamian HS (2023) Virtual and In Vitro Screening of Natural Products Identifies Indole and Benzene Derivatives as Inhibitors of SARS-CoV-2 Main Protease (Mpro). Biology 12(4):519
Wasilewicz A et al (2023) Identification of Natural Products Inhibiting SARS-CoV-2 by Targeting Viral Proteases: A Combined in Silico and in Vitro Approach. J Nat Prod 86(2):264–275
Dean B et al (2023) Hydroxy-xanthones as promising antiviral agents: Synthesis and biological evaluation against human coronavirus OC43. Bioorg Med Chem Lett 84:129211
Wansri R et al (2022) Semi-Synthesis of N-Aryl Amide Analogs of Piperine from Piper nigrum and Evaluation of Their Antitrypanosomal, Antimalarial, and Anti-SARS-CoV-2 Main Protease Activities. Molecules 27(9):2841
N-Containing α-Mangostin Analogs via Smiles Rearrangement as the Promising Cytotoxic,Antitrypanosomal, and SARS-CoV-2 Main Protease Inhibitory Agents. Molecules, 2023. 28(3): p. 1104
Skariyachan S (2022) Scope of computational biology and bioinformatics toward the discovery of potential therapeutic agents against viral diseases. Future Virol, 17
Vázquez J et al (2020) Merging Ligand-Based and Structure-Based Methods in Drug Discovery: An Overview of Combined Virtual Screening Approaches. Molecules, 25(20)
Swift RV et al (2016) Knowledge-Based Methods To Train and Optimize Virtual Screening Ensembles. J Chem Inf Model 56(5):830–842
Tripathi A et al (2023) Artificial Intelligence in Accelerating Drug Discovery and Development. Recent Pat Biotechnol 17(1):9–23
Pithi C et al (2016) Potential Anti-metastasis Natural Compounds for Lung Cancer. Anticancer Res 36(11):5707
Karnsomwan W et al (2017) Synthesis, Biological Evaluation and Molecular Docking of Avicequinone C Analogues as Potential Steroid 5α-Reductase Inhibitors. Chem Pharm Bull 65(3):253–260
Lin ACK et al (2022) Caffeic acid N-[3,5-bis(trifluoromethyl)phenyl] amide as a non-steroidal inhibitor for steroid 5α-reductase type 1 using a human keratinocyte cell-based assay and molecular dynamics. Sci Rep 12(1):20858
Ke S et al (2017) Steroidal[17,16-d]pyrimidines derived from dehydroepiandrosterone: A convenient synthesis, antiproliferation activity, structure-activity relationships, and role of heterocyclic moiety. Sci Rep 7(1):44439
Yokoya M et al (2023) Transformation of Renieramycin M into Renieramycins T and S by Intramolecular Photoredox Reaction of 7-Methoxy-6-methyl-1,2,3,4-tetrahydroisoquinoline-5,8-dione Derivatives. Journal of Natural Products, 86(1): p. 222–231
Thongsom S et al (2023) N,N'-Diarylurea Derivatives (CTPPU) Inhibited NSCLC Cell Growth and Induced Cell Cycle Arrest through Akt/GSK-3β/c-Myc Signaling Pathway. Int J Mol Sci, 24(2)
Daina A, Michielin O, Zoete V (2017) SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci Rep 7(1):42717
Lipinski CA (2004) Lead- and Drug-like Compounds: The Rule-of-Five Revolution. Drug Discovery Today: Technologies 1:337–341
Alam MS, Ahmed JU, Lee D-U (2016) Biological features, drug-likeness, pharmacokinetic properties, and docking of 2-arylidenehydrazinyl-4-arylthiazole analogues. Appl Biol Chem 59(2):181–192
Sanachai K et al (2022) Identification of repurposing therapeutics toward SARS-CoV-2 main protease by virtual screening. PLoS ONE 17(6):e0269563
Berthold MR et al (2007) KNIME: The Konstanz Information Miner. In Studies in Classification, Data Analysis, and Knowledge Organization (GfKL 2007). University of Konstanz, Nycomed Chair for Bioinformatics and Information Mining. Box 712, 78457 Konstanz, Germany
Fei J et al (2013) Pharmacophore modeling, virtual screening, and molecular docking studies for discovery of novel Akt2 inhibitors. Int J Med Sci 10(3):265–275
Cereto-Massagué A et al (2012) DecoyFinder: an easy-to-use python GUI application for building target-specific decoy sets. Bioinformatics 28(12):1661–1662
Sterling T, Irwin JJ (2015) ZINC 15 – Ligand Discovery for Everyone. J Chem Inf Model 55(11):2324–2337
Wishart DS et al (2006) DrugBank: a comprehensive resource for in silico drug discovery and exploration. Nucleic Acids Res 34(Database issue):D668–D672
Discovery Studio Modeling Environment, Release 2.5.1. Accelrys Software Inc: San Diego, CA.
Dolinsky TJ et al (2004) PDB2PQR: an automated pipeline for the setup of Poisson–Boltzmann electrostatics calculations. Nucleic Acids Res 32(suppl2):W665–W667
Marvin 17.21.0, Chemaxon August 19, 2022]; Available from: https://www.chemaxon.com
El-Hachem N et al (2017) AutoDock and AutoDockTools for Protein-Ligand Docking: Beta-Site Amyloid Precursor Protein Cleaving Enzyme 1(BACE1) as a Case Study. Methods Mol Biol 1598:391–403
Trott O, Olson AJ (2010) AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 31(2):455–461
Pettersen EF et al (2021) UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Sci 30(1):70–82
Shin H-C et al (2021) Identification and Biological Activities of the Phenolic Compounds in Eisenia arborea. Am J Plant Sci 12:259–265
Xue X et al (2007) Production of Authentic SARS-CoV Mpro with Enhanced Activity: Application as a Novel Tag-cleavage Endopeptidase for Protein Overproduction. J Mol Biol 366(3):965–975
Benet LZ et al (2016) BDDCS, the Rule of 5 and drugability. Adv Drug Deliv Rev 101:89–98
Martin YC (2005) A bioavailability score. J Med Chem 48(9):3164–3170
Council NR (2014) A Framework to Guide Selection of Chemical Alternatives. June 20, 2023]; Available from: https://www.ncbi.nlm.nih.gov/books/NBK253956/
Parate S et al (2021) Investigating natural compounds against oncogenic RET tyrosine kinase using pharmacoinformatic approaches for cancer therapeutics. RSC Adv 12(2):1194–1207
Li H et al (2009) An effective docking strategy for virtual screening based on multi-objective optimization algorithm. BMC Bioinformatics 10(1):58
Gan J-h et al (2023) DrugRep: an automatic virtual screening server for drug repurposing. Acta Pharmacol Sin 44(4):888–896
Kumar V et al (2021) Withanone and caffeic acid phenethyl ester are predicted to interact with main protease (M(pro)) of SARS-CoV-2 and inhibit its activity. J Biomol Struct Dyn 39(11):3842–3854
Hengphasatporn K et al (2022) Promising SARS-CoV-2 main protease inhibitor ligand-binding modes evaluated using LB-PaCS-MD/FMO. Sci Rep 12(1):17984
van de Sand L et al (2021) Glycyrrhizin Effectively Inhibits SARS-CoV-2 Replication by Inhibiting the Viral Main Protease. Viruses, 13(4)
Dhanjal JK et al (2021) Molecular mechanism of anti-SARS-CoV2 activity of Ashwagandha-derived withanolides. Int J Biol Macromol 184:297–312
Alves DR et al (2022) Curcumins and its derivatives as potential inhibitors of New Coronavirus (COVID-19) main protease: an in silico strategy. Res Soc Dev 11(1):e6511124334
Bahun M et al (2022) Inhibition of the SARS-CoV-2 3CLpro main protease by plant polyphenols. Food Chem 373:131594
Suriya U, Mahalapbutr P, Rungrotmongkol T (2022) Integration of In Silico Strategies for Drug Repositioning towards P38α Mitogen-Activated Protein Kinase (MAPK) at the Allosteric Site. Pharmaceutics, 14(7)
Deetanya P et al (2021) Interaction of 8-anilinonaphthalene-1-sulfonate with SARS-CoV-2 main protease and its application as a fluorescent probe for inhibitor identification. Comput Struct Biotechnol J 19:3364–3371
Lee JT et al (2022) Genetic Surveillance of SARS-CoV-2 M(pro) Reveals High Sequence and Structural Conservation Prior to the Introduction of Protease Inhibitor Paxlovid. mBio 13(4):e0086922
Unoh Y et al (2022) Discovery of S-217622, a Noncovalent Oral SARS-CoV-2 3CL Protease Inhibitor Clinical Candidate for Treating COVID-19. J Med Chem 65(9):6499–6512
Onyango H et al (2022) In Silico Identification of New Anti-SARS-CoV-2 Main Protease (M < sup > pro) Molecules with Pharmacokinetic Properties from Natural Sources Using Molecular Dynamics (MD) Simulations and Hierarchical Virtual Screening. Journal of Tropical Medicine, 2022: p. 3697498

No competing interests reported.

SI9CaffeicAcidFinalversion.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

A Combination of Structure-based Virtual Screening and Experimental Strategies to Identify the Potency of Caffeic Acid Ester Derivatives as SARS-CoV-2 3CLpro Inhibitor from an In-house Database

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Drug-likeness analysis

Structure-based pharmacophore modeling

Molecular docking

Preparation of ester derivatives of caffeic acid

Evaluation of SARS-CoV-2 3CL^pro inhibitory activity

Results and discussion

Evaluation of drug-like properties

Pharmacophore-based virtual screening and model validation

Hits identification from molecular docking

Synthesis of ester derivatives of caffeic acid

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

A Combination of Structure-based Virtual Screening and Experimental Strategies to Identify the Potency of Caffeic Acid Ester Derivatives as SARS-CoV-2 3CLpro Inhibitor from an In-house Database

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Drug-likeness analysis

Structure-based pharmacophore modeling

Molecular docking

Preparation of ester derivatives of caffeic acid

Evaluation of SARS-CoV-2 3CLpro inhibitory activity

Results and discussion

Evaluation of drug-like properties

Pharmacophore-based virtual screening and model validation

Hits identification from molecular docking

Synthesis of ester derivatives of caffeic acid

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Evaluation of SARS-CoV-2 3CL^pro inhibitory activity