key: cord-0788415-7gfwmcps
authors: Tietjen, Ian; Cassel, Joel; Register, Emery T.; Zhou, Xiang Yang; Messick, Troy E.; Keeney, Frederick; Lu, Lily D.; Beattie, Karren D.; Rali, Topul; Ertl, Hildegund C. J.; Salvino, Joseph M.; Davis, Rohan A.; Montaner, Luis J.
title: The natural stilbenoid (–)-hopeaphenol inhibits cellular entry of SARS-CoV-2 USA-WA1/2020, B.1.1.7 and B.1.351 variants
date: 2021-04-29
journal: bioRxiv
DOI: 10.1101/2021.04.29.442010
sha: 0de9221cc9a565f841e1cafcb861aae044d7eea3
doc_id: 788415
cord_uid: 7gfwmcps

Antivirals are urgently needed to combat the global SARS-CoV-2/COVID-19 pandemic, supplement existing vaccine efforts, and target emerging SARS-CoV-2 variants of concern. Small molecules that interfere with binding of the viral spike receptor binding domain (RBD) to the host ACE2 receptor may be effective inhibitors of SARS-CoV-2 cell entry. Here we screened 512 pure compounds derived from natural products using a high-throughput RBD/ACE2 binding assay and identified (–)-hopeaphenol, a resveratrol tetramer, in addition to vatalbinoside A and vaticanol B, as potent and selective inhibitors of RBD/ACE2 binding and viral entry. For example, (–)-hopeaphenol disrupted RBD/ACE2 binding with a 50% inhibitory concentration (IC50) of 0.11 μM in contrast to an IC50 of 28.3 μM against the unrelated host ligand/receptor binding pair PD-1/PD-L1 (selectivity index = 257.3). When assessed against the USA-WA1/2020 variant, (–)-hopeaphenol also inhibited entry of a VSVΔG-GFP reporter pseudovirus expressing SARS-CoV-2 spike into ACE2-expressing Vero-E6 cells and in vitro replication of infectious virus in cytopathic effect assays (IC50 = 10.2 μM) without cytotoxicity. Notably, (–)- hopeaphenol also inhibited two emerging variants of concern originating from the United Kingdom (B.1.1.7) and South Africa (B.1.351) in both cytopathic effect and spike-containing pseudovirus assays with similar (B.1.1.7) or improved (B.1.351) efficacies over the USA- WA1/2020 variant. These results identify (–)-hopeaphenol and related stilbenoid analogues as potent and selective inhibitors of viral entry across multiple SARS-CoV-2 variants including those with increased infectivity and/or reduced susceptibility to existing vaccines. Importance SARS-CoV-2 antivirals are needed to supplement existing vaccine efforts and target emerging viral variants with increased infectivity or reduced susceptibility to existing vaccines. Here we show that (–)-hopeaphenol, a naturally-occurring stilbenoid compound, in addition to its analogues vatalbinoside A and vaticanol B, inhibit SARS-CoV-2 by blocking the interaction of the viral spike protein with the cellular ACE2 entry receptor. Importantly, in addition to inhibiting the early USA-WA1/2020 SARS-CoV-2 variant, hopeaphenol also inhibits emerging variants of concern including B.1.1.7 (“United Kingdom variant”) and B.1.351 (“South Africa variant”), with improved efficacy against B.1.351. (–)-Hopeaphenol therefore represents a new antiviral lead against infection from multiple SARS-CoV-2 variants.

To identify potential inhibitors of SARS-CoV-2 entry, we used AlphaScreen technology 97 (13) to develop a high-throughput, 384 well plate-based assay to monitor the interaction of 98 SARS-CoV-2 Spike RBD with host ACE2 (Figure 2A) . Briefly, a SARS-CoV-2 RBD protein 99 derived from USA-WA1/2020 and containing a C-terminal His tag, in addition to a full-length 100 ACE2 peptide with a C-terminal Fc tag, were pre-bound to respective acceptor and donor beads 101 and co-incubated for 3 hours at room temperature. When a ligand/receptor binding event occurs, 102 excitation at 680 nm results in a singlet oxygen transfer between donor and receptor beads, 103 which results in luminescence at 615 nm. Compounds that inhibit binding of RBD to ACE2 104 should therefore inhibit luminescence. In the absence of compounds, we observed that 105 luminescence was highly dependent on concentrations of both RBD and ACE2 (Figure 2A) with 106 a > 200-fold signal to noise ratio and high reproducibility across experiments (Z' > 0.8) (14) . 107 Using this assay, we then screened 512 pure compounds obtained from natural products 108 and semisynthetic derivatives available from Compounds Australia at Griffith University at 1 DABCYL quencher and detection of fluorescence at 490 nm. Compounds that inhibit M pro 143 activity are therefore expected to inhibit fluorescence. This assay also exhibited a > 10-fold 144 signal to noise ratio and Z' > 0.6 ( Figure 3A) and was adaptable to 384-well screening format. 145 Using this assay, we observed that the control M pro inhibitor GC-376 blocked enzymatic 146 activity with an IC50 of 0.0052 μM, consistent with previous observations (Figure 3B ) (19) (20) . 147 In contrast, we observed that hopeaphenol, vatalbinoside A, and vaticanol B inhibited M pro 148 activity with IC50s of 42.5, 68.7 and 47.6 μM, respectively ( Figure 3B; Table 1 ), suggesting 149 that these stilbenoids, while potentially capable of targeting M pro , are to a first approximation 150 more effective against RBD binding to ACE2. 184 To confirm cellular antiviral activity of hopeaphenol and analogues, we next used a 185 cytopathic effect (CPE)-based assay with infectious virus in Vero-E6 cells (25) (26) . Briefly, 186 Vero-E6 cells were treated with compounds for 2 hours in 8-fold replicates in 96-well format 187 before infection with 50x median tissue culture infectious dose (TCID50) of SARS-CoV-2 (USA-WA1/2020 variant). Cells were then incubated for 4 days with daily scoring of CPE across 189 all wells by a user blinded to experimental conditions. Using this approach, we observed the To determine if hopeaphenol maintained activity against emerging SARS-CoV-2 variants 212 with accumulated mutations in the spike RBD, we repeated the CPE assay using two SARS-

CoV-2 variants of concern including B.1.1.7 (England/204820464/2020; "UK variant") and Table 1) . 226 To confirm that the antiviral activities of hopephanol against these variants of concern 227 corresponded to inhibition of viral entry, we generated VSVΔG-S-GFP-based pseudoviruses More recently, stilbenoids have been proposed as potential disruptors of SARS-CoV-2 271 spike protein with ACE2 in molecular docking studies (46) . Another stilbenoid, kobophenol A, 272 was also recently reported to inhibit binding of RBD with ACE2 (IC50 = 1.8 μM) and SARS-

CoV-2 replication (EC50 = 71.6 μM) (47) , and our data are consistent with these observations.

However, we observed no antiviral activity by resveratrol at up to 100 μM, which contrasts with 275 another recent study reporting an EC50 of 10.7 μM in SARS-CoV-2-infected Vero-E6 cells 276 (BetaCov isolate) (48) . However, this latter study measured supernatant viral RNA levels by 277 quantitative PCR after 48 hours' infection, and so disparate results could reflect differences in 278 sensitivity between quantitative PCR and CPE-based assays. Another consideration is that the three hit stilbenoid compounds, as polyphenolics, represent a structure class that has been given a 280 PAINS (Pan Assay Interference compoundS) designation (49) . Although we observed that these 281 compounds selectively disrupted RBD/ACE2 binding over an unrelated PD-1/PD-L1 282 ligand/receptor pair, for example with 257.3-fold selectivity for hopeaphenol, caution must still 283 be taken when considering these compounds for further therapeutic development. However, 284 generation or isolation of stilbenoid analogues with potentially improved selectivity as well as 285 assessment of chemical leads in primary cell models remain warranted. Table 1) . Assessment of additional 297 analogues and derivatives may also mitigate this concern. 298 A recent report also describes use of a virtual screening approach which identified 299 hopeaphenol as a potential inhibitor of SARS-CoV-2 M pro by interacting within its active site 300 (18). In contrast, we observed only weak inhibitory activity of hopeaphenol and analogues 301 against M pro (e.g., IC50s = ~40 -70 μM, compared to 0.0052 μM for GC-376; Nevertheless, these combined results do raise the intriguing possibility of identifying stilbenoid 308 derivatives that target both SARS-CoV-2 entry and M pro , which in turn may improve antiviral 309 efficacy and/or reduce the risk of eventual viral drug resistance.

There are currently no licensed antivirals that reliably protect against COVID-19. Recent 311 reports of SARS-CoV-2 variants with accumulated spike mutations and reduced susceptibility or 312 escape from neutralizing sera from convalescent and vaccine-treated patients (7-8) also raise the 313 concern of emerging variants with resistance to existing vaccines. In contrast, we observe that 

The 600 SARS-CoV-2 Spike protein has a broad tropism for mammalian ACE2 proteins

Analysis of ACE2 in polarized epithelial cells: surface 605 expression and function as receptor for severe acute respiratory syndrome-associated 606 coronavirus

Growth, detection, 609 quantification, and inactivation of SARS-CoV-2

The SARS-CoV-2 cytopathic effect is blocked 614 with autophagy modulators

CoV-2 in human lung cells and chimeric SARS-CoV expressing the SARS-CoV-2 RNA 621 polymerase in mice

Anti-inflammatory effects of vitisinol A 645 and four other oligostilbenes from Ampelopsis brevipedunculata var

Morikawa 649 T. 2017. Quantitative determination of stilbenoids and dihydroisocoumarins in Shorea 650 roxburghii and evaluation of their hepatoprotective activity

Antibacterial 653 activities of metabolities from Vitis rotundifolia (Muscadine) roots against fish pathogenic 654 bacteria

Anti-inflammatory effect and mechanism of action 658 of ellagic acid-3,3',4-trimethoxy-4'-O-α-L-rhamnopyranoside isolated from Hopea parviflora in 659 lipopolysaccharide-stimulated RAW 264.7 macrophages

Screening of 663 natural stilbene oligomers from Vitis vinifera for anticancer activity on human hepatocellular 44

Anti-hyperlipidemic 691 constituents from the bark of Shorea roxburghii

Antidiabetogenic oligostilbenoids and 3-ethyl-4-phenyl-3,r-dihydroisocoumarins from the bark 696 of Shorea roxburghii

Stilbene-based natural compounds as promising 699 drug candidates against COVID-19

Kobophenol A inhibits binding of host 703 ACE2 receptor with spike RBD domain of SARS-CoV-2, a lead compound for blocking 704 COVID-19

Resveratrol inhibits HCoV-229E and SARS-CoV-2 708 coronavirus replication in vitro

Feeling Nature's PAINS: Natural products, natural product drugs, and pan 711 assay interference compounds (PAINS)

Chemical and molecular mechanisms of antioxidants: 714 Experimental approaches and model systems

Bioavailability of the 717 polyphenols: Status and controversites

Bioavailability of dietary polyphenols 720 and gut microbiota metabolism: Antimicrobial properties

Protein production by auto-induction in high density shaking cultures. Figure 7. Effects of hopeaphenol on entry of pseudoviruses containing SARS-CoV-2 spike 770 variants. A, Representative images of Vero-E6 cells infected with VSVΔG-S-GFP pseudovirus 771 containing SARS-CoV-2

Images are organized as described in Figure 773

Level of GFP-positive (i.e., pseudovirus-infected) cells, relative to total cell nuclei

 474 We are indebted to Dr. Stephen J. Elledge for providing SARS-CoV-2 spike cDNA for