General experimental procedure

Chemicals were purchased from Sigma-Aldrich Chemicals Company and were used as was supplied. All solvents were reagent grade. Where necessary, solvents and starting materials were purified using standard procedures. Solvent removal was carried out under reduced pressure using a Buchi rotary evaporator at temperatures not greater than 60°C. Melting points were measured using a Mel-Temp II apparatus with the use of open capillaries and were uncorrected. The progress of all reactions was monitored by thin layer chromatography (TLC) on aluminum-backed silica gel 60 F254 plates obtained from Sigma-Aldrich; visualization was by UV light at 254 nm or by staining with iodine. The compounds were purified by medium-pressure liquid chromatography over silica gel 60-to-400 mesh, using the appropriate solvent systems.

High-resolution Fourier transform mass spectrometry electrospray ionization (FTMS-ESI) mass spectra were generated using an LTQ Orbitrap XL mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). A heated electrospray interface (H-ESI) was operated for ionization of the molecules at a spray voltage of 5 kV. Capillary voltage and tube lens voltages were then adjusted to 20 and 100 V, respectively. The vaporizer temperature was set at 250°C and the ion transfer capillary temperature was set to 200°C. Measurements were carried out in the positive ion mode in a mass range of m/z 100–600 at a mass resolution of 60 000 at m/z 200. MS/MS experiments were performed using argon as collision gas in collision-induced dissociation (CID) mode, with collision energies measured at 15, 25 and 35 eV.

Nuclear magnetic resonance (NMR) spectra were obtained using a Bruker Avance III spectrometer operating at 600 MHz (H1) and 150 MHz (13C). Spectra were recorded in deuterated solvents and referenced to residual solvent signals. Chemical shifts (δ) were measured in parts per million. Hydrogen and carbon assignments were done using gradient correlation spectroscopy (gCOSY), gradient heteronuclear single quantum correlation (gHSQC) spectroscopy, and heteronuclear multiple bond correlation (gHMBC) techniques. Multiplicities are reported as singlet (s), doublet (d), doublet of doublets (dd), doublet of triplets (dt), triplet (t), triplet of doublets (td) and multiplet (m). Coupling constants (J) are reported in Hertz. For biological evaluation, all compounds were converted to the corresponding hydrochlorides by treatment of the free bases with methanolic HCl. All compounds are greater than 95% pure by high-performance liquid chromatography (HPLC) analysis.

Synthesis of additional spirooxindoles

Synthesis of 5,7-dibromo-6′,7′-dihydroxy-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one (12c) following Method A.

The compound was synthesized via the phenolic Pictet–Spengler reaction, as reported [50]. To a solution of 5,7-dibromo isatin (1.5 g, 5.1 mmol) in absolute ethanol (10 ml) was added dopamine (1 g, 5.1 mmol) and triethylamine (1 ml). The reaction mixture was stirred and heated under reflux for 7–10 h, and subsequently concentrated under reduced pressure to remove the solvent. Distilled water was added to the resulting viscous mass and the product, which precipitated out was extracted into ethyl acetate (3 × 30 ml). The combined organic extracts were dried over anhydrous sodium sulfate and concentrated to a minimum volume. The crude product was further purified by column chromatography (hexane: ethyl acetate – 60:40). The final product was re-crystallized from absolute ethanol. Yield, 1.7 g, 76% (brown solid). M.p. 256–258°C (HCl salt).

1H NMR (CD3OD, 700 MHz): δ ppm 2.74 (dt, J = 16.1, 4.6 Hz, 1H, H4′a), 2.90 (ddd, J = 15.9, 5.4 Hz, 1H, H4′b), 3.10–3.15 (m, H3′a), 3.71–3.77 (m, H3′b), 5.91 (s, 1H, H8′), 6.62 (s, 1H, H5′), 7.27 (d, J = 1.8 Hz, 1H, H4), 7.64 (d, J = 1.8 Hz, 1H, H6). 13C NMR (CD3OD, 175 MHz): δ ppm 27.2 (C4′), 38.4 (C3′), 64.7 (C3/C1′), 102.8 (C7), 111.9 (C8′), 114.9 (C5), 115.4 (C5′), 123.5 (C8′a), 126.7 (C4), 127.3 (C4′a), 133.6 (C6), 138.7 (C3a), 140.9 (C7a), 143.8 (C7′), 145.0 (C6′), 180.1 (C2). MS(ESI): cald for C16H12Br2N2O3 [M + H]+ 440.09, found 440.93; LC(ESI): tR 8.77 min, purity 90%.

General method for the synthesis of 6-methoxy- & 6′,7′-dimethoxy-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-ones (13l &14h) following Method B.

A mixture of the appropriate isatin (1 equiv), methoxyphenethylamine (1.2 equiv), and polyphosphoric acid (2 g) was heated in an oil bath (bath temperature at 100°C) while stirring mechanically for 5 h. Upon completion of the reaction, as revealed by TLC, the reaction mixture was allowed to cool to about 50°C and quenched by slow addition of water. To this mixture, a saturated solution of sodium carbonate to adjust the pH to 11. The floating product obtained was extracted into ethyl acetate (3 × 30 ml). The combined organic extracts were dried over anhydrous sodium sulfate and concentrated under reduced pressure to obtain the crude product. The latter was purified using suitable solvent systems by flash chromatography on silica gel. Yields ranged between 60 and 98%.

6′-methoxy-5-methyl-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one (13l) following Method B. This was prepared from 5-methylisatin (2.8 g, 17 mmol), 3-methoxyphenethylamine (2.6 g 17 mmol), and polyphosphoric acid (3 g). The crude product was purified by flash chromatography (hexane: ethyl acetate – 80:20). Yield, 4.6 g, 92% (brown solid), M.p. 208–209°C.

1H NMR (DMSO-d6, 600 MHz): δ ppm 1.46 (s, 3H, 5-CH3), 2.07 -2.13 (m, 1H, H4′a), 2.23 (ddd, J = 16.5, 8.7, 5.3 Hz, 1H, H4′b), 2.39 (dt, J = 12.8, 5.2 Hz, 1H, H3′a), 2.96 (d, J = 5.1 Hz, 4H, H3′b, m, 4H, 7′-OCH3), 5.64 (d, J = 8.6 Hz, 1H, H8′), 5.80 (dd, J = 8.6, 2.7 Hz, 1H, H7′), 5.95 (d, J = 2.7 Hz, 1H, H5′), 6.07 (d, J = 7.87 Hz, 1H, H7), 6.14–6.17 (m, 1H, H4), 6.29 (ddd, J = 7.9, 1.7, 0.8 Hz, 1H, H6). 13C NMR (DMSO-d6, 150 MHz): δ ppm 18.9 (5-CH3), 27.6 (C4′), 37.5 (C3′), 53.4 (6′- OCH3), 62.9 (C3/C1′), 108.7 (C7), 111.8 (C7′), 112.6 (C5′), 124.3 (C4), 125.4 (C8′a), 126.4 (C8′), 128.3 (C6), 131.5 (C3a), 134.5 (C7a), 136.4 (C4′a), 138.4 (C5), 158.0 (C6′), 180.3 (C2). MS(ESI): cald for C18H18N2O2 [M + H]+ 294.35, found 294.15; LC(ESI): tR 9.52 min, purity 93%.

1-(4-fluorobenzyl)-6′,7′-dimethoxy-5-methyl-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one (14h) following Method B.

This was prepared from 5-methyl-1-(4-fluorobenzyl)indoline-2,3-dione (1 g, 3.7 mmol), 3,4- dimethoxyphenethylamine (0.8 g, 4.4 mmol) and polyphosphoric acid (3 g). The crude product was purified by flash chromatography (hexane: ethyl acetate – 60:40). Yield, 1.4 g, 90% (brown solid), M.p. 99–101°C

1H NMR (DMSO-d6, 600 MHz): δ ppm 2.19 (s, 3H, 5-CH3), 2.71 (dt, J = 15.9, 4.1 Hz, 1H, H4′a), 2.88 (ddd, J = 15.1, 9.3, 5.4 Hz, 1H, H4′b), 3.05 (ddd, J = 12.5, 5.4, 4.1 Hz, 1H, H3′a), 3.29(s, 3H, 7′-OCH3), 3.65 (ddd, J = 12.4, 9.3, 4.3 Hz, 1H, H3′b), 3.74 (s, 3H, 6′-OCH3), 4.76 (d, J = 15.6 Hz, 1H, CH2-Ar), 4.96 (d, J = 15.6 Hz, 1H, CH2-Ar), 5.72 (s, 1H, H8′), 6.76 (s, 1H, H5′), 6.92 (dd, J = 4.8, 3.1 Hz, 2H, H4, H7), 7.05 (ddd, J = 8.0, 1.8, 0.9 Hz, 1H, H6), 7.16–7.18 (m, H3″, H5″), 7.42 (dd, J = 8.6, 5.5 Hz, 2H, H2″, H6″). 13C NMR (DMSO-d6, 150 MHz): δ ppm 21.0 (5-CH3), 28.7 (C4′), 38.9 (C3′), 42.2 (CH2-Ar), 55.8 (7′-OCH3), 56.0 (6′-OCH3), 63.6 (C3/C1′), 109.2 (C7), 109.4 (C8′), 113.0 (C5′), 115.8 (C3″, C5″), 125.5 (C4), 127.2 (C8′a), 129.3 (C6), 129.5 (C4′a), 129.9 (C2″, C6″), 132.2 (C3a), 133.5 (C1″), 135.5 (C5), 140.4 (C7a), 148.5 (C7′), 148.5 (C6′), 161.2 (C4″), 178.8 (C2). MS(ESI): cald for C26H25FN2O3 [M + H]+ 432.49, found 432.19; LC(ESI): tR 17.50 min, purity 70%.

Synthesis of N-ethyl-6′,7′-dimethoxy-1-(4-methylbenzyl)-2-oxo-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinoline]-2′-carboxamide (17c) following Method C.

The target compound was prepared from the previously described 6′,7′-dimethoxy-1-(4-methylbenzyl)-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one 14e (1 g, 2.4 mmol) and ethyl isocyanate (0.21 g, 0.23 mL, 2.9 mmol, 1.2 eq). An acetonitrile solution of 14e and ethyl isocyanate was heated to 60°C for 2 h. Upon completion of the reaction, the mixture was allowed to cool to room temperature, made basic by the slow addition of aqueous sodium bicarbonate to pH 10. The product was extracted into ethyl acetate (30 mL x 2), and the combined organic extracts were dried over anhydrous sodium sulfate and concentrated under reduced pressure. The crude product was purified by flash chromatography (hexane: ethyl acetate – 70:30). Yield, 0.6 g, 50% (white solid). M.p. 193–194°C.

1H NMR (DMSO-d6, 700 MHz): δ ppm 0.96 (t, J = 7.17 Hz, 3H, N1″′-CH2CH3), 2.27 (s, 3H, 4″-CH3), 2.90 (ddd, J = 15.4, 4.8, 3.4 Hz, 1H, H4′a), 2.92 - 3.01 (m, 3H, 1H, H4′b, N1″′-CH2CH3), 3.11 (s, 3H, 7′-OCH3), 3.55–3.60 (m, 1H, H3′a), 3.72 (s, 3H, 6′-OCH3), 3.97 (td, J = 12.2, 4.6 Hz, 1H, H3′b), 4.62 (d, J = 15.5 Hz, 1H, N1-CH2), 4.96 (d, J = 15.5 Hz, 1H, N1-CH2), 5.76 (s, 1H, H8′), 6.84 (s, 1H, H5′), 6.89 (td, J = 7.5, 1.0 Hz, 1H, H5), 6.93 (dt, J = 7.9, 0.7 Hz, 1H, H7), 7.03 (dd, J = 7.3, 1.25 Hz, 1H, H4), 7.10 - 7.13 (d, J = 7.8 Hz, 2H, H3″, H5″), 7.17 (td, J = 7.7, 1.3 Hz, 1H, H6), 7.31–7.34 (m, 2H, H2″, H6″). 13C NMR (DMSO-d6, 175 MHz): δ ppm 15.8 (N1″′-CH2CH3), 21.1 (4″-CH3), 30.0 (C4′), 35.4 (N1″′-CH2CH3), 42.2 (C3′), 43.4 (N1-CH2), 55.4 (7′-OCH3), 55.9 (6′-OCH3), 65.5 (C3/C1′), 108.9 (C7), 109.1 (C8′), 112.3 (C5′), 122.3 (C4), 122.7 (C5), 126.5 (C8′a), 128.2 (C4′a), 128.3 (2C, C2″, C6″), 128.9 (C6), 129.5 (2C, C3″, C5″), 134.5 (C1″), 135.7 (C3a)137.0 (C4″), 143.5 (C7a), 147.7 (C7′), 148.3 (C6′), 156.7 (C2″′), 177.3 (C2). MS(ESI): cald for C29H31N3O4 [M + H]+ 485.57, found 485.24; LC(ESI): tR 22.58 min, purity 95%.

Synthesis 6′,7′-dimethoxy-2′-methyl-1-(4-methylbenzyl)-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one (17d) following Method C.

This compound was prepared from previously synthesized 6′,7′-dimethoxy-1-(4-methylbenzyl)-3′,4′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one (14e) [43] (1 g, 2.4 mmol) and formaldehyde (0.3 mL of 37% formalin, 3.6 mmol, 1.5 eq). To a formic acid solution of 14e formaldehyde was added dropwise. The resulting mixture was heated at 60°C for 3 h, allowed to cool to room temperature, and made basic by slowly adding 2 M aqueous sodium hydroxide. The product was extracted into ethyl acetate (30 mL x 3), and the combined organic extracts were dried over anhydrous sodium sulfate and concentrated under reduced pressure. The crude product was purified by flash chromatography (hexane: ethyl acetate – 50:50). Yield, 0.8 g, 78% (yellow oil).

1H NMR (DMSO-d6, 700 MHz): δ ppm 2.06 (s, 3H, N2′-CH3), 2.26 (s, 3H, 4″-CH3), 2.78 (dt, J = 15.8, 3.6 Hz, 1H, H4′a), 2.88 (ddd, J = 11.3, 5.7, 2.9 Hz, 1H, H3′a), 3.03 (ddd, J = 16.0, 10.5, 5.6 Hz, 1H, 1H, H4′b), 3.22 (s, 3H, 7′-OCH3), 3.63 (td, J = 10.9, 4.1 Hz, 1H, H3′b), 3.73 (s, 3H, 6′-OCH3), 4.77 (d, J = 15.4 Hz, 1H, N1-CH2), 4.97 (d, J = 15.5 Hz, 1H, N1-CH2), 5.65 (s, 1H, H8′), 6.77 (s, 1H, H5′), 6.99–7.01 (m, 2H, H4, H5), 7.06 (d, J = 7.9 Hz, 1H, H7), 7.14 (d, J = 7.8 Hz, 2H, H3″, H5″), 7.26–7.29 (m, 3H, H6, H2″, H6″). 13C NMR (DMSO-d6, 175 MHz): δ ppm 21.0 (4″-CH3), 28.7 (C4′), 39.6 (N2′-CH3), 42.7 (N1-CH2), 46.9 (C3′), 55.6 (7′-OCH3), 55.9 (6′-OCH3), 69.0 (C3/C1′), 109.6 (C8′), 109.6 (C7), 112.4 (C5′), 123.5 (C5), 124.8 (C4), 126.7 (C8′a), 128.0 (C4′a), 128.1 (2C, C2″, C6″), 129.4 (2C, C3″, C5″), 129.7 (C6), 133.2 (C3a), 134.2 (C1″), 137.3 (C4″), 143.5 (C7a), 147.4 (C7′), 148.5 (C6′), 177.3 (C2). MS(ESI): cald for C27H28N2O3 [M + H]+ 428.52, found 428.21; LC(ESI): tR 21.05 min, purity 90%.

Description of biological screening (AlphaScreen assay) procedure

SARS-CoV-2 spike-RBD binding to ACE2 was determined using AlphaScreen technology-based assay as described previously [51]. For RBD-ACE2 assays, 2 nM of ACE2-Fc (Sino Biological, Chesterbrook, PA, USA) was incubated with 5 nM HIS-tagged SARS-CoV-2 Spike-RBDs representing the parental USA-WA/2020 (“Wild-type” (WT)) sequence (SinoBiological) in the presence of 5 μg/mL nickel chelate donor bead in a total of 10 μL of 20 mM Tris (pH 7.4), 150 mM KCl, and 0.05% CHAPS. Test compounds were diluted to 100× final concentration in DMSO. 5 μL of ACE2-Fc/Protein A acceptor bead was first added to the reaction, followed by 100 nL test compound and then 5 μL of RBD-HIS/Nickel chelates donor beads. All conditions were performed in duplicate. Following incubation at room temperature for 2 h, luminescence signals were measured using a ClarioStar plate reader (BMC Labtech, Cary, NC, USA). Data were then normalized to percent inhibition, where 100% equaled the AlphaScreen signal in the absence of RBD-HIS, and 0% denoted the AlphaScreen signal in the presence of both protein and DMSO vehicle control.

Molecular modeling proceduresTarget proteins for docking

Since the emergence of the COVID-19 pandemic, several computational methodologies have been employed in an attempt to identify lead compounds for drug discovery and development [52, 53] Molecular modeling protocols were performed, as previously reported [7,8,9, 51, 54, 55]. The docking evaluated protein-ligand interaction of the small molecules towards the ACE2 protein, as we had previously demonstrated that compounds that bind into the angiotensin II site of the ACE2 receptor would prevent recognition of the two proteins and, hence, binding of the viral to the ACE2 receptor [9, 55]. The protein structures (ID: 6M0J) for SARS-CoV-2 spike/ACE2, corresponding to the Wuhan strain were retrieved from the Protein Data Bank (PDB) [56,57,58] and used for the entire study.

Protein preparation

All water molecules were deleted using the Molecular Operating Environment (MOE) software [59]. The Protein Preparation Wizard integrated into the Schrödinger package software [60, 61] was used to prepare the protein by adding the missing hydrogen bonds, assigning bond orders, and filling the missing side chains using PRIME. After this, the protein structures were energy minimized to reduce atomic clashes and optimized their interactions with the ligands during docking. From the Schrödinger software, the commercialized Maestro package’s Epik-tool was used to predict the protonation states at a pH of 7.0 [62, 63]. Finally, a restrained energy minimization step was carried out using the Optimized Potentials for Liquid Simulations 2005 (OPLS2005) forcefield [64] on both proteins. During the protein optimization step, the root mean square deviation (RMSD) of the displacement of the atoms was set to end with the minimization at 0.3 Å.

Ligand preparation

The MOE [60] builder module was used to generate the 3D models of the library of synthesized spirooxindoles. For consistency, only the R stereoisomers were prepared for docking, as these addressed the voluminous hydrophobic regions in the ACE2 site more appropriately during trial docking. The generated 3D structures were then energy minimized using the MMFF94 force field [65,66,67,68,69,70,71]. The ligands were further prepared for docking using the LigPrep tool to generate all the plausible tautomers of each ligand as implemented in Schrödinger’s Maestro software package [63]. Using the incorporated OPLS2005 force field [64], the spirooxindole 3D structure library was further energy minimized. The ConfGen tool (implemented in the Schrödinger package) was then used to compute 60 conformers per ligand in the 3D library, by setting all other options to default except for the minimization of the output [60].

Docking and scoring

Docking was carried out using the Glide program incorporated in the Maestro package distributed by Schrödinger [61, 62] as shown in our recent publications [7,8,9, 51,52,53,54,55], with some modifications. Docking validation results on this protein have already been reported in our previously reported studies [57,58,59]. After the protein preparation phase, a docking grid box was generated for the spike/ACE2 complex to investigate how the ligands will bind around the following amino acid residues; Asp597, Thr598, Lys516, Val321, Gln121, Lys578, Ala283, Ser91, Asn746, Gln68, Pro744, Glu518 and Thr610. The ligand size for each of these grid boxes, which is the area where all the generated 3D structures were docked, was set to a maximum ligand size of 36 Å. While writing 10 poses per ligand conformer, 20 poses were included for each ligand conformer, and taking into consideration the input of ring conformation, all other settings were allowed to default. The outputs were scored using standard precision (SP) GlideScore as the scoring function [72].

Selection of binding modes

After the extraction of the results and the computation of carefully selected descriptors, the specific area ligands bound with the protein in the receptor binding domain (RBD) of both the Spike/ACE2, the binding modes, and the residues taking part in the interaction during binding were observed using MOE [60]. Browsing through the docking results and establishing the ligand interactions of each docked protein-ligand complex made it possible to establish structure-activity relationships (SAR) in the RBD in both cases and to identify some ligand moieties important for activity and selectivity. The ligands in both protein RBD were then superimposed to highlight their preferred binding modes.

Re-scoring of docked poses by MM-GBSA

To properly explain the observed biological activities, the Molecular Mechanics Generalized Born Solvation Area (MM-GBSA) model was employed as a means of re-scoring the docked protein-ligand poses. The PRIME tool incorporated in the Maestro package from Schrödinger (2017) was used to do this [61]. The free energy of the binding (ΔGbind) for each ligand towards the spike/ACE2 complex was calculated by using the Prime MM-GBSA algorithm (using default parameters). Each docked pose was retrieved from the Glide docking output and input on the PRIME program for calculating several thermodynamic properties including the binding free energy (ΔGbind) and solvation free energy (ΔGsolv) values in kcal/mol. The binding pose of the complex structures was visually inspected by using the ligand interaction tool in MOE to gain insight into the binding mode (see additional notes on this method under Fig. S3, Supplementary Data).

Prediction of pharmacokinetic properties of the active compounds

An initial assessment of the risk of further developing the active molecules into lead compounds for the discovery of next-generation antiviral agents was conducted by prediction of the drug metabolism and pharmacokinetics properties of the active and moderately active compounds. The pharmacokinetic properties were conducted by the computation of parameters related to drug absorption, distribution, metabolism, and elimination by using the SwissADME web server [73]. Each chemical structure was converted to a simple molecular input line entry system (SMILES) and uploaded onto the SwissADME web server platform (http://www.swissadme.ch) [73]. This then enabled the computation of 46 descriptors often used to predict the DMPK (drug metabolism and pharmacokinetic) profiles. The computed descriptors were, amongst others, molecular weight, molar refractivity, number of rotatable bonds, Lipinski violations, aqueous solubility, Veber violations, Ghose violations, Egan violations, gastro-intestinal absorption, blood-brain-barrier permeability, cytochrome inhibition, synthetic accessibility, P-glycoprotein binding, skin permeability, Bioavailability score, Muegge violations, PAINS alerts, Lead-likeness violations, etc. Additional DMPK- and toxicity-related parameters were computed using the pkCSM web server (https://biosig.lab.uq.edu.au/pkcsm/) [74]. The pkCSM signatures are then applied across different pharmacokinetic properties to develop predictive regression and classification models to predict absorption, distribution, metabolism, excretion, and toxicity (ADMET). The additional parameters include the steady-state volume of distribution, central nervous system (CNS) permeability, blood−brain barrier permeability, total clearance, and toxicity parameters like maximum recommended tolerated dose (MRTD), oral rat acute toxicity (LD50), oral rat chronic, lowest observed adverse effect (LOAEL), as well as toxicity against fish species Tetrahymena pyriformis and fathead minnow toxicity (LC50).