Identification of SARS-CoV-2 Mpro inhibitors containing P1' 4-fluorobenzothiazole moiety highly active against SARS-CoV-2

Abstract

COVID-19 caused by SARS-CoV-2 has continually been serious threat to public health worldwide. While a few anti-SARS-CoV-2 therapeutics are currently available, their antiviral potency is not sufficient. Here, we identify two orally available 4-fluoro-benzothiazole-containing small molecules, TKB245 and TKB248, which specifically inhibit the enzymatic activity of main protease (M^pro) of SARS-CoV-2 and significantly more potently block the infectivity and replication of various SARS-CoV-2 strains than nirmatrelvir, molnupiravir, and ensitrelvir in cell-based assays employing various target cells. Both compounds also block the replication of Delta and Omicron variants in human-ACE2-knocked-in mice. Native mass spectrometric analysis reveals that both compounds bind to dimer M^pro, apparently promoting M^pro dimerization. X-ray crystallographic analysis shows that both compounds bind to M^pro's active-site cavity, forming a covalent bond with the catalytic amino acid Cys-145 with the 4-fluorine of the benzothiazole moiety pointed to solvent. The data suggest that TKB245 and TKB248 might serve as potential therapeutics for COVID-19 and shed light upon further optimization to develop more potent and safer anti-SARS-CoV-2 therapeutics.

Fig. 1. Intracellular concentrations of TKB245 and TKB248.

VeroE6 or HeLa-ACE2-TMPRSS2 cells were incubated with 10, 50, and 100 μM of each compound for 6 hours, vigorously washed with PBS, and intracellular concentrations of each compound were determined using LC/MS. Bars indicate arithmetic means (n = 2). LOQ: Limit of Quantitation. Source data are provided as a Source Data file.

Fig. 2. In vivo efficacy of TKB245 and TKB248 against SARS-CoV-2_NC928-2N^Omicron_BA.1, SARS-CoV-2_UW-5250^Delta and SARS-CoV-2_NCD1288^Omicron_BA.2-infected hACE2-knocked-in mice.

a Five hACE2KI mice per group were challenged intranasally with SARS-CoV-2_NC928-2N^Omicron (5 × 10⁵ PFU) or SARS-CoV-2_UW-5250^Delta (5 × 10⁵ PFU). Two hours later, animals were intraperitoneally administered 100 mg/kg BID TKB245 or vehicle (placebo). Animals were euthanized on 2- and 3-days post infection and lungs collected for determination of virus titers using VeroE6^TMPRSS2 cells. Bars indicate mean values and error bars represent standard deviations. All p values are calculated using the exact Wilcoxon rank-sum test with two-sided, and no multiple adjustment was made. Source data are provided as a Source Data file. b Five hACE2K mice per group were challenged intranasally with SARS-CoV-2_NC928-2N^Omicron_BA.1 (5 × 10⁵ PFU) or SARS-CoV-2_NCD1288^Omicron_BA.2 (1 × 10⁵ PFU). Two hours later, animals were intraperitoneally administered 120 mg/ kg BID TKB248 or vehicle (placebo). Animals were euthanized on 2- and 3-days post infection and lungs collected for determination of virus titers using VeroE6^TMPRSS2 cells. Bars indicate mean values and error bars represent standard deviations. All p values are calculated using the exact Wilcoxon rank-sum test with two-sided, and no multiple adjustment was made. Source data are provided as a Source Data file.

Fig. 3. Native mass spectrometric analysis of SARS-CoV-2 M^pro-inhibitor interaction.

a Authentic M^pro (7.5 µM) was treated with 15 µM of each indicated compound. Relative native mass spectra of the M^pro with or without TKB245, TKB248, or nirmatrelvir are shown. Charge states 10⁺, 11⁺ and 12⁺ are annotated to mass spectra corresponding to M^pro monomer species and charge states 14⁺, 15 ⁺, 16 ⁺, 17 ⁺, 18⁺ and 19⁺ are annotated to mass spectra corresponding to M^pro dimer species. b Glycine-added M^pro (10 µM) was treated with 50 µM of each indicated compound. Relative native mass spectra of the M^pro with or without TKB245, TKB248, or nirmatrelvir are shown. Charge states 8⁺, 9⁺, 10⁺, 11⁺, 12⁺, and 13⁺ are annotated to mass spectra corresponding to M^pro monomer species and charge states 13⁺, 14⁺, 15⁺, and 16⁺ are annotated to mass spectra corresponding to M^pro dimer species.

Fig. 4. Co-crystal structures of TKB245 and TKB248 with SARS-CoV-2 M^pro.

a Overview of M^pro dimer in complex with TKB245. Molecular surface of protomer A colored in cyan and protomer B in orange. b Superimposition of TKB245 in green onto TKB248 in purple exhibits identical binding mode. M^pro binding pocket is colored according to hydrophobicity scale. Hydrophobicity of the binding pocket is represented by the intensities of orange color such as hydrophobic residues: Leu-27 and Phe-140 as shown in orange. Polar or charged residues such as Glu-166, Gln-189 are shown in light blue. c Binding mode and hydrogen bond network in M^pro complexed with TKB245. Cartoon representation of the crystal structure of M^pro is shown in orange complexed with TKB245 (green stick). Hydrogen bonds are indicated as black dashed lines. d Top view focused on trifluoromethyl (P4) group with potential fluorine-based interaction: Less than 4 Å are indicated as cyan dashed lines. Three fluorine atoms engaged in multi-directional halogen bond interactions with surrounding residues consist of Leu-167, Pro-168, Gln-192, and Met-165. e–g Comparison of nirmatrelvir (PBD ID: 7RFW) vs TKB245 (PBD ID: 8DOX) and TKB248 (PBD ID: 8DPR) inside the S1’ subsite including oxyanion hole interactions. While all inhibitors covalently bind to the catalytic residue Cys-145, the 4-fluorobenzothiazole ring with the fluorine atom points to solvent. h Side-by-side comparison of nirmatrelvir vs TKB-245 and TKB-248 (as shown in transparent surface) onto the polyprotein substrate. Blue color indicates the 4-fluorobenzothiazole ring of TKB245 and TKB248 that effectively fill the S1’subsite compared to nirmatrelvir.