CEBIT screening for inhibitors of the interaction between SARS-CoV-2 spike and ACE2

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, causing COVID-19, is the most challenging pandemic of the modern era. It has resulted in over 5 million deaths worldwide. To quickly explore therapeutics for COVID-19, we utilized a previously-established system, namely CEBIT. We performed a high-throughput screening of FDA-approved drugs to inhibit the interaction between the receptor-binding domain (RBD) of SARS-CoV-2 spike protein and its obligate receptor ACE2. This interaction is essential for viral entry and therefore represents a promising therapeutic target. Based on the recruitment of interacting molecules into phase-separated condensates as a readout, we identified six positive candidates from a library of 2572 compounds, most of which have been reported to inhibit the entry of SARS-CoV-2 into host cells. Our surface plasmon resonance (SPR) and molecular docking analyses revealed the possible mechanisms via which these compounds interfere with the interaction between RBD and ACE2. Hence, our results indicate that CEBIT is highly versatile for identifying drugs against SARS-CoV-2 entry, and targeting CoV-2 entry by small molecule drugs is a viable therapeutic option to treat COVID-19 in addition to commonly used monoclonal antibodies.

Keywords: ACE2; High-throughput screening; Protein-protein interaction; SARS-CoV-2; Spike protein.

Graphical abstract

Fig. 1

Validation of CEBIT-based system for detecting RBD-ACE2 interaction. (a) Schematic diagram showing the strategy of SmF-based phase separation system. The condensates formed by SmF-mCherry-SIM and SmF-GFP-SUMO3 serve as the “reactor” for PPI recruitment (top left). With the addition of (SUMO3)₅-SA and RBD-Strep-tag in turn (top right and bottom right), ACE2-His was recruited into the condensates via its interaction with RBD-Strep-tag. ACE2-His is fluorescently tagged to monitor its recruitment (bottom right). Disruption of the RBD-ACE2 interaction by an inhibitor or competitor evicts ACE2 from the condensates (bottom left). (b) Microscopy analysis of the recruitment of ACE2 labeled by Alexa 647 into the condensates formed by SMF-mCherry-SIM and SmF-GFP-SUMO3. (c) Relative fluorescence intensity ratio of ACE2 (Alexa 647) versus condensates (GFP).

Fig. 2

Optimization of a multivalent recruitment system driven by polySUMO-polySIM. (a) Schematic diagram showing the strategy for the phase separation system mediated by polySUMO-polySIM. The scaffolds used here to mediate the formation of condensates are (SUMO3)₅-SA and SIM₁₁. The colocalization of the client protein ACE2-His with the condensates was observed in the presence of the bridging protein RBD-Strep-tag. (b) Microscopy analysis of the recruitment of ACE2 labeled by Alexa 647 into the condensates formed by (SUMO3)₅-SA and SIM₁₁ (Alexa 488). (c) Relative fluorescence intensity ratio of ACE2 (Alexa 647) versus condensates (Alexa 488). (d-e) Competitive binding assay between RBD-His and RBD-Strep-tag for ACE2 protein.

Fig. 3

Optimization of a multivalent recruitment system driven by HP1α. (a) Schematic diagram showing the strategy for the HP1α-driven phase separation system. Partitioning of ACE2-His into the HP1α-PDZ-driven condensates is mediated by its interaction with RBD- KKETPV. (b) Microscopy analysis of the recruitment of ACE2 labeled by Alexa 546 into the condensates of HP1α-PDZ (Alexa 647), and the competitive binding assay by RBD-His protein. (c) Analysis of the relative fluorescence intensity ratio of ACE2 (Alexa 546) versus condensates (Alexa 647).

Fig. 4

High throughput screening for inhibitors of the RBD-ACE2 interaction. (a) Microscopy analysis of the inhibition activity of positive compounds. (b) Quantitative analysis of the relative ratio of fluorescence intensities of ACE2 (Alexa 546) and HP1α-PDZ (Alexa 647). (c) Schematic draws of the two-dimensional structures of the positive compounds.

Fig. 5

SPR analysis of the binding of the positive compounds to RBD. (a, c, e, g, i, k) Sensorgrams of the multicycle interaction between the indicated compounds and RBD. (b, d, f, h, j, l) Fitting analysis of the concentration-dependent response in a 1:1 equilibrium model.

Fig. 6

Molecular docking analysis of RBD complexed with selected compounds. (a) Molecular docking results of RBD in complex with Varenicline tartrate. The binding energy is -4.89 Kcal/mol and the predicted inhibition constant is 260 μM. (b) Molecular docking results of RBD in complex with Sennoside A. The binding energy is -4.51 Kcal/mol and the predicted inhibition constant of 494 μM. (c) Molecular docking results of RBD in complex with Methylene blue. The binding energy is -4.42 Kcal/mol and the predicted inhibition constant of 577 μM. The RBD and ACE2 proteins are depicted using ribbon structure while the compounds are depicted with stick models.