Synthesis, Structure–Activity Relationships, and Antiviral Profiling of 1-Heteroaryl-2-Alkoxyphenyl Analogs as Inhibitors of SARS-CoV-2 Replication

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19, has led to a pandemic, that continues to be a huge public health burden. Despite the availability of vaccines, there is still a need for small-molecule antiviral drugs. In an effort to identify novel and drug-like hit matter that can be used for subsequent hit-to-lead optimization campaigns, we conducted a high-throughput screening of a 160 K compound library against SARS-CoV-2, yielding a 1-heteroaryl-2-alkoxyphenyl analog as a promising hit. Antiviral profiling revealed this compound was active against various beta-coronaviruses and preliminary mode-of-action experiments demonstrated that it interfered with viral entry. A systematic structure–activity relationship (SAR) study demonstrated that a 3- or 4-pyridyl moiety on the oxadiazole moiety is optimal, whereas the oxadiazole can be replaced by various other heteroaromatic cycles. In addition, the alkoxy group tolerates some structural diversity.


Introduction
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of the human coronavirus disease 2019 . SARS-CoV-2 is a newly discovered coronavirus that was first identified in December 2019 in Wuhan, China, and spread quickly throughout the world, infecting and causing death to millions of people [1,2]. On 11 March 2020, the World Health Organization (WHO) officially declared COVID-19 a global pandemic. As of October 2021, more than 235 million people were infected with SARS-CoV-2 and almost 5 million people have died due to COVID-19 [3].
Major research efforts were carried out to control this pandemic. Using a number of different technology platforms (such as messenger RNA and vector-based vaccines), the development of various vaccines moved forward at an unparalleled speed and several of them received marketing approval [4]. Despite the success of these SARS-CoV-2 vaccines, there is still an urgent need for small molecule therapeutics. As a drug discovery program aiming at the development of SARS-CoV-2-specific antiviral drugs from scratch would take a long time, the initial focus has been on the repurposing of known drugs and drug candidates. Several large-scale phenotypic screening campaigns of different repurposing compound libraries have been described in the literature. With different SARS-CoV-2infected cell lines and different readouts being used, the outcome has been heterogeneous and a wide variety of compounds has been discovered [5][6][7][8].
Drug discovery programs focusing on viral proteins that are essential for SARS-CoV-2 replication are also being pursued. The viral genome of SARS-CoV-2 encodes for 16 nonstructural proteins (nsp1-nsp16). These are highly conserved among various coronaviruses and many of them are excellent candidates as targets for the discovery of antiviral agents [9]. Examples include the cap guanine N7-methyltransferase and 3 -5 exonuclease activity of nsp14 [10], the nsp15 endoribonuclease [11], and the papain-like protease (PLpro) activity of nsp3 [12]. Targets that are most intensively studied are nsp5, also known as the chymotrypsin-like protease (3CLpro) or the viral main protease (Mpro) [13] and the RNAdependent RNA polymerase (RdRp, encoded by nsp12) [14,15]. Drug discovery programs on the latter two enzymes have afforded drugs and drug candidates for the treatment of SARS-CoV-2 viral infections. PF-00835231 is a potent SARS-CoV-2 Mpro inhibitor, when evaluated in a biochemical assay and shows potent SARS-CoV-2 antiviral activity in a cellbased antiviral assay. Its corresponding phosphate prodrug, also known as PF-07304814, is currently undergoing clinical trials for COVID-19 treatment, after intravenous administration [16]. More recently, Pfizer disclosed PF-07321332, an orally bioavailable SARS-CoV-2 Mpro inhibitor, displaying excellent in vitro potency in biochemical and cell-based antiviral assays. Moreover, it is endowed with good activity in SARS-CoV-2 mouse and hamster models, after oral administration [17]. Pfizer recently reported interim data of a phase II/III clinical trial (unpublished data) demonstrating a marked reduction in progression to severe COVID-19 or hospitalization if treatment in high-risk patients is initiated during the first 5 days of onset of symptoms.
The viral RNA-dependent RNA polymerase (RdRp) is another attractive target for the treatment of COVID-19 patients. Remdesivir is a phosphoramidate prodrug of the C-nucleoside GS-441524 that received FDA approval for the treatment of adult and pediatric (aged > 12 years) SARS-CoV-2-infected patients. Since it requires intravenous administration, remdesivir usage is limited to hospital settings [18]. Yet, Gilead reported (unpublished) that early intravenous treatment of at risk-patients markedly reduced progression to severe COVID-19. Molnupiravir, an ester prodrug of N-4-hydroxycytidine, is orally bioavailable and has shown promising activity in various SARS-CoV-2 preclinical animal models [19] and was licensed for medical use in the United Kingdom and the EU for use in patients suffering from mild to moderate COVID-19 and who have at least one risk factor for developing severe illness.
Cell-based phenotypic based screening campaigns of structurally diverse, drug-like compound libraries are another important method to discover novel antiviral hit compounds which serve as a basis for further development. Advantages of this approach are that such screens are physiologically more relevant than target-based screens and that there is a priori no bias towards a particular target and hence allows to select antiviral agents targeting different viral, as well as cellular, factors [20]. To the best of our knowledge, this type of screening for the discovery of novel hit matter against SARS-CoV-2 has not been described in the literature. In this manuscript, the discovery of a novel SARS-CoV-2 hit compound, its preliminary structure-activity relationship (SAR) study, and its antiviral profile are discussed.

Hit Identification
In order to identify potential hits, the CD3 (Centre for Drug Design and Discovery, KU Leuven) small molecule library was screened against SARS-CoV-2 in VeroE6 cells, which constitutively expresses the enhanced green fluorescent protein (eGFP) [21]. A reduced eGFP expression correlates with the cytopathogenic effect (CPE) of SARS-CoV-2 and, hence, the measurement of the fluorescence by high-content imaging can be used as an indication for the antiviral potency. In the presence of a non-toxic antivirally active agent, the cytopathogenicity is inhibited and the fluorescent signal is maintained. In the first round of screening, compounds were tested at a single concentration of 10 µM in a 384-well plate format. Initial hits were followed up by full-dose response analysis allowing to express the antiviral activity as the concentration producing 50% antiviral effect (EC 50 ). In parallel, the cellular toxicity of the compounds, which was expressed as the 50% cytotoxic concentration (CC 50 ), was determined using mock-infected VeroE6 cells. Among the various hits, 5-(2-((1-phenethylpyrrolidin-3-yl)oxy)phenyl)-3-(pyridin-4-yl)-1,2,4-oxadiazole 1 (Figure 1) displayed an EC 50 value of 4.7 µM and a CC 50 value of 21 µM, and was considered a promising starting point for further chemistry.
In order to access a small library of compounds bearing various groups on the oxadiazole, a convergent route was designed to generate several compounds in one step from a common intermediate (Scheme 2). Briefly, the key building block 19 was prepared in three steps from phenol 15 and N-Boc-hydroxy-pyrrolidine 16 by Mitsunobu reaction, followed by Boc deprotection and N-alkylation by reductive amination with cyclopentanone. The ester 19 or the lithium salt of the corresponding carboxylic acid 20 were then converted into the desired oxadiazoles 23-31 by condensation with various amidoximes 22, which were either commercially available or prepared by nucleophilic attack of hydroxylamine on appropriate nitriles 21.
In order to study the importance of the 1,2,4-oxadiazole ring for antiviral activity, several synthetic routes were developed to obtain analogs bearing other 5-membered heteroaryls (Schemes 3-6). The synthesis of an analog having an isomeric 1,2,4-oxadiazole ring is shown in Scheme 3. The required amidoxime 34 was synthesized in three steps from intermediate 19. Treatment of the ester with aqueous ammonia yielded the primary amide 32. Dehydration with propylphosphonic anhydride (T3P) [24], followed by treatment with hydroxylamine yielded amidoxime 34. Oxadiazole ring formation via an O-acylamidoxime intermediate using EDC/HOBt as coupling agents provided the desired compound 35.
To access compound 46, the 1,2,4-thiadiazole ring was not synthesized since thiadiazole 39 is a commercially available building block (Scheme 5) [25]. A first selective Suzuki-Miyaura coupling using ortho-methoxyphenylboronic acid occurred on the chlorine and was followed by a second coupling with pyridin-4-ylboronic acid furnishing intermediate 42. Demethylation with LiBr led to the formation of the phenol 43, which underwent a Mitsunobu reaction with N-Boc-3-hydroxypyrrolidine to provide intermediate 44. Removal of the Boc protecting group under acidic conditions, followed by alkylation of the pyrrolidine moiety by reductive amination, furnished the desired compound 46.
The isoxazole analog 49 was obtained via a cyclization of oxime 47 with ester 19 in the presence of LDA, followed by treatment with 12N HCl at 70 • C. All compounds were evaluated for anti-SARS-CoV-2 activity in VeroE6 cells, which allowed to establish first SAR trends. As these cells have a high expression of the efflux transporter P-glycoprotein (P-gp), the assays were performed in the presence of CP-100356, a well-known P-gp inhibitor [12]. In all experiments, GS-441524 (the parent nucleoside of remdesivir) [26] was included as a positive control, with an average EC 50 and CC 50 value of 0.82 µM and 72.9 µM, respectively.
As compound 1 presents a chiral center, the two enantiomers 1a and 1b were synthesized using optically pure pyrrolidine building blocks and tested against SARS-CoV-2. The three compounds 1, 1a, and 1b showed similar biological profiles, indicating a limited influence of the stereochemistry on the antiviral activity (Table 1). All the other compounds were then synthesized and evaluated as racemic mixtures. The SAR was explored around three distinct structural subunits of hit compound 1: The alkoxy group on the phenyl moiety, the substituent on the oxadiazole ring, and the oxadiazole ring itself. We first investigated the alkoxy group modifications (Table 2), taking into account that an amino group was required for antiviral activity (data not shown). Several structural modifications around this alkoxy group were well tolerated: Replacement of the phenethyl substituent on the pyrrolidine of hit compound 1 by a cyclopentyl (compound 8) or a methylcyclopentyl (compound 9) moiety, the insertion of an additional methylene linker between the oxygen and the pyrrolidine (compound 10) or replacement of the pyrrolidine by a piperidine (compound 12). A cyclic amine was not essential for the antiviral potency as shown by compounds 13 and 14 with a linear secondary or tertiary amine. On the other hand, the introduction of a methylene group between the pyrrolidine and the phenoxy ring of compound 10 yielded analog 11, which was 20-fold less active than compound 10. We next investigated modifications around the 1,2,4-oxadiazole subunit (Table 3). To probe the effect of substituents on the 1,2,4-oxadiazole moiety on the antiviral activity, a number of derivatives were synthesized (compounds 23-31). Albeit the 3-pyridinyl analog 23 was equipotent to 8, the introduction of a second nitrogen into the ring led to potency loss (3-fold for the pyrimidine 24 and more than 10-fold for the pyridazine 25). The phenyl analog 26 showed only cytotoxicity. Since the 3-or 4-pyridyl derivatives seemed to be the optimal 6-membered ring, the addition of substituents was explored. Several groups with various sizes and electronic properties were well tolerated at different positions of the 3-pyridyl moiety, such as a methoxy (compound 27), a trifluoromethyl (compounds 28-29), or a methyl group (compound 30). The positive effect of the methyl group in ortho position to the oxadiazole was confirmed also with the 4-pyridyl analog 31. The importance of the 1,2,4-oxadiazole moiety for antiviral activity was investigated by the synthesis of a number of alternative 5-membered heteroaryl derivatives (Table 4). Except for compound 38 with a 1,3,4-oxadiazole moiety, which was 3-fold less active, other derivatives containing a 1,2,4-oxadiazole (35), a 1,2,4-thiadiazole (46) or an isoxazole (49) showed a similar biological profile as compound 8, indicating that the 1,2,4-oxadiazole ring is not involved in crucial interactions for antiviral activity, but can be considered as a linker.

Broad-Spectrum Antiviral Activity
The primary SARS-CoV-2 screening focused on the B.1 (Wuhan) lineage. However, several variants of SARS-CoV-2 have emerged and are circulating worldwide, and more will emerge in the future. Therefore, it is important to determine the activity of antiviral agents against the variants of concerns (VoCs). Interestingly, compound 1 showed a similar activity against the B.1 lineage and the B.1.1.7 (UK or Alpha variant) and B.1.617.2 (Indian or Delta variant) lineages in Vero E6 cells (Table 5). In addition, compound 1 is equally active against other β-coronaviruses, such as SARS-CoV-1 and the Middle East respiratory syndrome coronavirus (MERS-CoV).

Mode of Action Investigation
In order to exclude a non-specific antiviral effect, we tested compound 1 against the SARS-CoV-2 unrelated Chikungunya virus in Vero A cells [27]. The complete lack of antiviral activity of compound 1 (EC 50 > 100 µM) in this assay demonstrates a specific antiviral effect for coronavirus. However, since compounds were evaluated in a cell-based assay, the exact molecular target responsible for the antiviral activity remains elusive. In an effort to pinpoint the exact target, the inhibitory effect of compound 1 against various coronavirus enzymes such as SARS-CoV-2 Mpro, SARS-CoV-2 nsp14 (N7-methyltransferase activity), and the SARS-CoV-2 replication-transcription complex was studied. However, even at the highest tested concentration of 100 µM, compound 1 did not show any inhibitory activity on these enzymes.
To evaluate if compound 1 acted at the level of the virus entry, a SARS-CoV-2 and MERS pseudovirus neutralization assay was performed. In both assays, compound 1 was able to block the virus entry with an EC 50 value of 1.25 µM and 3.4 µM, respectively. Compound 1 was found to be inactive against human proteases involved in SARS-CoV-2 entry such as cathepsin L, cathepsin B, TMPRSS2, and furin. Altogether, this first set of data demonstrated that hit 1 exerts its antiviral activity via interference with the viral entry mechanism. More research is required to determine if this virus entry inhibition is mediated by a viral or a host target protein.

Materials and Methods
3.1. Chemistry 3.1.1. General Methods All reagents and solvents were purchased from Sigma-Aldrich (Saint Louis, MO, USA), TCI (TCI Europe N.V., Zwijndrecht, Belgium), Combi-Blocks (San Diego, CA, USA) and used without further purification. All the reactions were monitored by thin-layer chromatography (TLC) or liquid chromatography-mass spectrometry (LCMS). TLC was carried out with Sigma-Aldrich silica gel (254 nm) plates (cat ref 99571-25EA) and TLC plates were revealed with UV light, KMnO 4 , p-anisaldehyde, or ninhydrin solutions. LCMS analysis was performed on an Agilent 1260 Infinity II UPLC machine (Agilent Technologies, Santa Clara, CA, USA) with a YMC-Triart C18 column (YMC CO., Kyoto, Japan). Flash chromatography purifications were performed on Biotage prepacked silica gel columns using Biotage Isolera instruments (Biotage, Upsala, Sweden). Reversed-phase preparative highpressure liquid chromatography (HPLC) purification of final analogs was performed on a Waters Autopurification instrument (Waters Corporation, Milford, MA, USA) with MS-and UV-triggered collection operating at ambient temperature and at a flow rate of 16 mL/min. The identity of all compounds with reported biological activity was confirmed by NMR spectroscopy and low-resolution mass spectrometry (MS). The purity of all compounds with reported biological activity was ≥95% as determined by NMR and ultraperformance liquid chromatography (UPLC). Proton NMR spectra were recorded on a 300 or 400 MHz Bruker spectrometer (Bruker Corporation, Billerica, MA, USA) using TMS as internal standard, whereas carbon NMR spectra were recorded on a Bruker Avance 600 MHz instrument (Bruker Corporation, Billerica, MA, USA) ( 13 C NMR, 150 MHz) using DMSO-d 6 (39.5 ppm) as internal standard. Proton and carbon chemical shifts (δ) are reported in parts per million (ppm). Abbreviations used are: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad. Coupling constants are expressed in Hz. Low-resolution mass spectral data were obtained using a Waters H-Class UPLC (Waters Corporation, Milford, MA, USA) with a Waters Acquity UPLC BEH C18 1.7 µm, 2.1 mm × 30 mm column (Waters Corporation, Milford, MA, USA), equipped with an Acquity UPLC PDA Detector (Waters Corporation, Milford, MA, USA), an Acquity UPLC ELS Detector (Waters Corporation, Milford, MA, USA)and an Acquity TQ Detector (ESI/ESCi) (Waters Corporation, Milford, MA, USA).
3.1.14. tert-Butyl 3-(2-(methoxycarbonyl)phenoxy)pyrrolidine-1-carboxylate (17) To a solution of DIAD (8.8 mL, 44.9 mmol, 1.2 eq) in THF (20 mL) cooled at 0 • C was added PPh 3 (10.5 g, 41.1 mmol, 1.1 eq). The reaction mixture was stirred at 0 • C for 30 min. tert-Butyl 3-hydroxypyrrolidine-1-carboxylate (7.0 g, 37.4 mmol, 1 eq) was added and the reaction mixture was stirred at 0 • C for 15 min. Methyl 2-hydroxybenzoate (5.7 g, 37.4 mmol, 1 eq) was added and the reaction mixture was stirred at room temperature overnight. The reaction mixture was concentrated under reduced pressure. The residue was partitioned between water and EtOAc. The phases were separated. The aqueous phase was extracted twice with EtOAc. The organic phases were combined, dried over Na 2 SO 4 , filtered, and concentrated under reduced pressure. The residue was purified by flash chromatography on silica gel using a gradient of EtOAc (0-50%) in hexane to afford 4.0 g (33%) of the title compound as a colorless liquid. 1 (18) To a solution of 17 (12.8 g, 39.7 mmol, 1 eq) in CH 2 Cl 2 (130 mL) cooled at 0 • C was added 4N HCl in dioxane (30 mL, 120 mmol, 3 eq). The reaction mixture was stirred at room temperature overnight and was concentrated under reduced pressure to afford quantitatively the title compound as a brown oil. 1 (19) To a solution of 18 (11.4 g, 44.2 mmol, 1 eq) in CH 2 Cl 2 (120 mL) was added cyclopentanone (5.1 mL, 57.5 mmol, 1.3 eq) and the reaction mixture was stirred at room temperature for 3 h. NaBH(OAc) 3 (14.1 g, 66.4 mmol, 1.5 eq) was added portion-wise and the reaction mixture was stirred at room temperature overnight. The reaction mixture was washed with water and the phases were separated. The aqueous phase was extracted with 10% MeOH in CH 2 Cl 2 (3×). The organic phases were combined, dried over Na 2 SO 4 , filtered, and concentrated under reduced pressure. The residue was purified by flash chromatography on silica gel using a gradient of MeOH (0-10%) in CH 2 Cl 2 to afford 9.9 g (77%) of the title compound as a brown oil. 1 (20) To a solution of 19 (1.0 g, 3.46 mmol, 1 eq) in THF (30 mL), MeOH (10 mL) and water (20 mL) was added LiOH (0.138 g, 3.28 mmol, 0.9 eq). The reaction mixture was heated at 60 • C for 6 h and was concentrated under reduced pressure. The residue was dissolved in water (30 mL) and was washed twice with CH 2 Cl 2 . The aqueous phase was lyophilized to afford 0.800 g (82%) of lithium 2-((1-cyclopentylpyrrolidin-3-yl)oxy)benzoate as an offwhite solid. 1 (23) To a solution of N-hydroxynicotinimidamide (0.150 g, 1.09 mmol, 1 eq) in DMSO (2 mL) were added 19 (0.474 g, 1.64 mmol, 1.5 eq) and NaOH (0.066 g, 1.64 mmol, 1.5 eq). The reaction mixture was stirred at room temperature overnight. The reaction mixture was diluted with water and extracted with EtOAc. The organic phase was dried over Na 2 SO 4 , filtered, and concentrated under reduced pressure. The residue was purified by flash chromatography on silica gel using a gradient of MeOH
The reaction mixture was concentrated under reduced pressure. The residue was taken up with sat. NaHCO 3 and was extracted twice with 5% MeOH in CH 2 Cl 2 . The organic phases were combined, washed with brine, dried over Na 2 SO 4 , filtered, and concentrated under reduced pressure. The residue was purified by reversed-phase HPLC (YMC-Actus C18 column, 250 × 20 mm, 5µ; 30 to 75% ACN in 20 mM ammonium bicarbonate in 20 min) to afford 0. A mixture of 3-bromo-5-chloro-1,2,4-thiadiazole (1.0 g, 5.01 mmol, 1 eq), (2-methoxyphenyl) boronic acid (0.381 g, 2.51 mmol, 0.5 eq) and CsF (1.52 g, 10.0 mmol, 2 eq) in dioxane (20 mL) and water (5 mL) was degassed with Ar for 30 min. Pd(dppf)Cl 2 (0.367 g, 0.501 mmol, 0.1 eq) was added and the reaction mixture was heated at 85 • C overnight. After cooling to room temperature, the reaction mixture was diluted with EtOAc and filtered. The phases of the filtrate were separated. The organic phase was concentrated under reduced pressure. The residue was purified by flash chromatography on silica gel using a gradient of EtOAc (0-20%) in hexane to afford 0.620 g (46%) of the title compound as a white solid. 1

Cells and Viruses
The SARS-CoV-2 isolate used was derived from the BetaCov/Belgium/GHB-03021/2020 (EPI ISL407976|2020-02-03), which was isolated from a Belgian patient returning from Wuhan in February 2020. The isolate was passaged 7 times on VeroE6 cells which introduced two series of amino acid deletions in the spike protein [29]. The infectious content of the virus stock was determined by titration on Vero E6 cells.

SARS-CoV-1 and SARS-CoV-2 Screening
The SARS-CoV-1 and SARS-CoV-2 antiviral assay in Vero E6 cells that was used in this study was derived from a previously established SARS-CoV-1 assay [21]. Stock solutions of the various compounds in DMSO (10 mM) were prepared. On day −1, the test compounds were serially diluted in assay medium and the plates were incubated overnight (37 • C, 5% CO 2 , and 95% relative humidity). After, VeroE6-eGFP cells were plated corresponding to a final density of 25,000 cells per well in black 96-well plates (Greiner Bio-One, Vilvoorde, Belgium; Catalog 655090). On day 0, the cells with compound were infected with SARS-CoV-1 (at 20 CCID 50 per well) or SARS-CoV-2 (at 20 CCID 50 per well). The plates were incubated in a humidified incubator at 37 • C and 5% CO 2 . At 4 days p.i., the wells were examined for eGFP expression using an argon laser-scanning microscope. The microscope settings were excitation at 488 nm and emission at 510 nm, and the fluorescence images of the wells were converted into signal values. The antiviral activity was expressed as EC 50 defined as the concentration of compound achieving 50% inhibition of the virus-reduced eGFP signals as compared to the untreated virus-infected control cells.
The same methodology was applied for the VoC, in which cells were infected with the SARS-CoV-2 at a final MOI of approximately 0.05 TCID50/cell. The final dilution of the different strains was adapted in order to obtain a similar MOI between all variants of interest.
The cytotoxicity of the compounds for VeroE6 cells in the absence of virus was evaluated in a standard MTS assay as previously described [31].
MERS-CoV cytopathic effect reduction assays were performed in Huh-7 cells (1.5 × 104 cells/well) using 96-well plates. Cells, seeded on the day prior to infection, were incubated with 100 µL volumes of 2-fold serial dilutions of compounds in infection medium, followed by infection with 225 PFU of MERS-CoV in 50 µL, yielding a total assay volume of 150 µL. Non-infected cells were treated in parallel with the same dilution series of compounds to determine cytotoxicity. After incubation for 42 h at 37 • C, cell viability was quantified with the CellTiter-96 Aqueous Non-radioactive Cell Proliferation Kit (Promega, Madison, WI, USA) and absorption at 495 nm was measured with an EnVision multilabel plate reader (PerkinElmer) and after normalization to uninfected (and untreated cells), EC50 and CC50 values were determined by nonlinear regression using GraphPad Prism v8.0 (San Diego, CA, USA).
In the virus neutralization assay, compounds were threefold serially diluted at two times the desired final concentration in DMEM supplemented with 1% fetal bovine serum, 100 U/mL Penicillin, and 100 µg/mL Streptomycin (Lonza, Basel, Switzerland). Monoclonal antibodies against MERS-CoV spike (7.7G6) [32] or SARS2-CoV-2 spike (REGN10933) [34] were included as a positive control. Diluted compounds and mAbs were incubated with an equal volume of pseudotyped VSV particles for 1 h at room temperature, inoculated on confluent VeroE6 monolayers in a 96-well plate, and further incubated at 37 • C for 24 h. Cells were lysed with Luciferase Cell Culture Lysis 5× Reagent (Promega, Madison, Wisconsin, USA) at room temperature for 30 min. Luciferase activity was measured on a Promega GloMax ® Explorer luminometer using D-luciferin as a substrate (Promega, Madison, Wisconsin, USA). The half-maximal inhibitory concentrations (IC 50 ) were determined using 4-parameter logistic regression (GraphPad Prism version 8, San Diego, CA, USA).

Nsp14 RapidFire MS Screening Assay
An endpoint 384-well plate assay was developed to assess nsp14 activity [35]. Briefly, enzyme and substrates, SAM (S-(5 -adenosyl)-l-methionine chloride hydrochloride [Cayman Chemical, Ann Arbor, MI, USA] and cap G(5 )ppp(5 )G sodium salt [New England Biolabs, Ipswich, MA, USA]) were incubated to allow the reaction to take place, and then the product was quantified using MS. Assays were performed in 384-well, clear, flat-bottom plates (Greiner 781101) to a final volume of 20 µL. Components were diluted in buffer (20 mM Tris, pH 8.0, 50 mM NaCl) containing 1 mM TCEP (Thermo Scientific, Waltham, MA, USA), 0.1 mg/mL bovine serum albumin, 0.005% Nonidet P40 (Roche), and 3 mM MgCl2. In the assay, 5 nM nsp14 was incubated with 1 µM SAM and 0.7 µM cap (FAC). Following a 60 min incubation, the reaction was quenched using 1% formic acid (VWR, Radnor, PA, USA) containing 0.03 µg/mL S-adenosylhomocysteine-d4 (d4SAH; Cambridge Bioscience, Cambridge, UK) and loaded on the RapidFire system by aspiration for 600 ms using the Agilent RapidFire 365 high-throughput system with integrated solid-phase extraction (SPE) interfaced with the Agilent 6740 triple quadrupole mass spectrometer. The sample was then automatically loaded onto a C18 Type C SPE cartridge (Agilent Technologies), and buffer salts and protein matrix were removed from the sample by washing the cartridge with the load solution (water containing 0.1% trifluoroacetic acid [TFA]) at a flow rate of 1.5 mL/min for 5000 ms. The retained and purified analytes were eluted from the cartridge with the elution solution (acetonitrile: water [9:1, v/v] containing 0.1% TFA) at 1.25 mL/min for 5000 ms and directed to the mass spectrometer. The cartridge was re-equilibrated with load solution at 1.5 mL/min for 500 ms. Both S-adenosylhomocysteine (SAH) and d4SAH were assessed using multiple selected reaction monitoring (MRM) transitions of 385.1/134 for SAH and 389.2/135.9 for d4SAH. The dwell time was 50 ms for each transition. The fragmentor voltage was set to 120 V for SAH and 100 for d4SAH, the collision energy to 12 V for SAH and 14 V for d4SAH, the cell accelerator voltage to 5 V, and the delta electron multiplier voltage to 200 V. The mass spectrometer was operated with a gas temperature of 350 • C, gas flow rate of 7 L/min, nebulizer pressure of 40 psi, and capillary voltage of 3000 V. The areas under the daughter ion peaks of SAH and d4SAH were integrated using RapidFire QQQ Quantitative Analysis software (Agilent Technologies), and the area ratios of the SAH to the internal standard d4SAH were used for quantitation.

SARS-CoV-2 Polymerase Assay
The compound concentrations leading to 50% inhibition of polymerase-mediated RNA synthesis was determined in IC 50 buffer (50 mM HEPES pH 8.0, 10 mM KCl, 2 mM MnCl 2 , 2 mM MgCl 2 , 10 mM DTT) containing 350 nM of Poly(A) template, seven various concentrations of compound (from 1 to 100 µM) and 150 nM of nsp2 in complex with 450 nM of nsp7L8 and 450 nM of nsp8.
Reactions were conducted in 40 µL volume on a 96-well Nunc plate. All experiments were robotized by using a BioMek 4000 automate (Beckman). Two microliters of each compound diluted in 100% DMSO was added in wells to the chosen concentration (5% DMSO final concentration). For each assay, the enzyme mix was distributed in wells after a 5 min incubation at room temperature to form the active complex. Reactions were started by the addition of the UTP mix and were incubated at 30 • C for 20 min. Reaction assays were stopped by the addition of 20 µL EDTA 100 mM. Positive and negative controls consisted, respectively, of a reaction mix with 5% DMSO final concentration or EDTA 100 mM instead of compounds. Reaction mixes were then transferred to Greiner plate using a Biomek I5 automate (Beckman). Picogreen ® fluorescent reagent was diluted to 1/800 • in TE buffer according to the manufacturer and 60 µL of reagent was distributed into each well of the Greiner plate. The plate was incubated for 5 min in the dark at room temperature and the fluorescence signal was then read at 480 nm (excitation) and 530 nm (emission) using a TecanSafire2.
IC 50 was determined using the equation: % of active enzyme = 100/(1 + (I)2/IC 50 ), where I is the concentration of inhibitor and 100% of activity is the fluorescence intensity without inhibitor. IC 50 was determined from curve fitting using Prism software. For each value, results were obtained using triplicate in a single experiment. 3 dUTP was used as an inhibitor control.

Furin and TMPRSS2 Enzymatic Assays
Recombinant furin was purchased from BioLegend (#719406), human recombinant TRMPSS2 from Cliniscience (ref LS-G57269-100), and the fluorogenic DABSYL/Glu-TNSPRRAR↓SVAS-EDANS-labeled peptides encompassing the S1/S2 cleavage site were purchased from Genscript. Briefly, the reactions were performed at room temperature in black 384-well 553 polystyrene low-volume plates (CELLSTAR-Greiner Bio-One # 784476) at a final volume of 15 µL. The fluorescent peptides were used at 5 µM and the reactions were performed in 50 mM Tris buffer (pH 7.5), 0.2% Triton X-100, 1 mM CaCl 2 in the presence of 2 nM of furin. The inhibitors solubilized in DMSO were tested at 50 µM with 5% DMSO final concentration in the enzymatic assay.
For the TMPRSS2 assay, the fluorescent peptides were used at 5 µM and the reactions were performed in 50 mM Tris buffer (pH 8), and 150 mM NaCl and TMPRSS2 were added at final concentrations of 50 nM.
The cleavage of the synthetic peptides was quantitated by determining the increase in EDANS (562 nM) fluorescence following the release of the DABCYL quencher. The EDANS was excited at 335 nM using a Safire 2 Tecan fluorimeter. Each reaction was performed in triplicate.

SARS-CoV-2 Main Protease (Mpro) Assay
SARS-CoV-2 main protease (Mpro) was recombinantly produced as described before [36]. Compounds were tested in a Förster resonance energy transfer (FRET) assay for inhibition of the SARS-CoV-2 Mpro. The peptide Dabcyl-KTSAVLQ↓SGFRKM-E(Edans)-NH2 (Biosyntan, Berlin, Germany) (↓ indicates the cleavage site), corresponding to the P7-P6 residues of the nsp4-nsp5 processing site of the viral polyprotein pp1a/pp1ab, was used as the substrate. Quenching of the Edans fluorescence by the Dabcyl residue was eliminated after cleavage of the scissile bond between P1-Gln and P1 -Ser and the difference in fluorescence emission was measured with a Tecan Spark fluorescence plate reader, operated at an excitation wavelength of 360 nm and an emission wavelength of 460 nm.
Candidate compounds were kept in a DMSO stock solution; the Mpro was in a buffer containing 20 mM HEPES, 120 mM NaCl, 0.4 mM EDTA, 20% glycerol, pH 7.0. DTT (4 mM) was added to the buffer prior to running the assay. Following incubation of 50 nM SARS-CoV-2 Mpro with 100 µM of the candidate compound for 10 min at 37 • C, the reaction was initiated by the addition of 10 µM of the FRET substrate to each well. The final DMSO concentration was <2%.

Conclusions
A high-throughput screening of a drug-like compound library led to the discovery of a hit compound 1, showing low micromolar activity against SARS-CoV-2 with a selectivity index of 5. Exploration of the SAR yielded compounds that were equally active, but showed a decreased cytotoxicity and hence displayed an improved SI. Hit compound 1 was profiled more extensively and was shown to be active against the VoCs of SARS-CoV-2, as well as against various betacoronaviruses. Furthermore, pseudovirus neutralization assays revealed that the compound exerts its antiviral effect via interference with viral entry. Overall, this study highlights the possibility to discover new hit compounds with promising activities against SARS-CoV-2 from a phenotypic approach. More investigations are, however, needed to improve the antiviral potency and to study its mechanism of action.