Fig 2: ACE2P1D1 and ACE2P2D1 do not affect the enzymatic activity of ACE2. (a,b) Assays were run at 37 °C in ACE2 reaction buffer containing 10 µM Mca-YVADAPK(Dnp) with 0.1 µg of ACE2 protein and various concentrations of ACE2P1D1 and ACE2P2D1. Fluorescence (excitation 320 nm and emission 405 nm) was measured in a microplate reader. Cleavage of the substrate by ACE2 enzyme produces fluorescence, which indicates ACE2 enzymatic activity. (c–h) Human cell lines NCI-H1299, Caco-2, and MCF-7 were treated with various concentrations of ACE2P1D1 or ACE2P2D1, in the presence of cOmplete™ Protease Inhibitor Cocktail (40 µL/mL) and 10 µM captopril. ACE2 activity was measured as described in (a,b), using 40 µg of cell lysate per each cell line. Each bar represents the mean + standard deviation (n = 3/group).

Fig 3: ACE2P1 and ACE2P2 dimers blocked the binding of spike protein to ACE2. (a–c) Recombinant ACE2 (0.1 µg/sample) and recombinant GST-tagged SARS-CoV2 spike proteins (0.1 µg/sample) were mixed together. ACE2P1 and ACE2P2 monomers or dimers (D1) were premixed with ACE2 for 1 h before the mixture was applied to the GST pull-down assay to determine if these compounds can block the interaction between ACE2 and spike protein. ACE2 in the pull-down assay was detected by western blot. (d) The interaction of recombinant ACE2 and D614G spike proteins (0.1 µg/sample) was determined by pull-down assay in the presence or absence of various concentrations of the peptoids.

Fig 4: SMD simulations. (A) SMD force-extension curve for pulling on SARS-CoV-2 RBD:ACE2 interaction without additional linker. Inset (blue frame) shows a rendering of the interface opening causing the peak in the trajectory. (B) SMD force-extension curve for pulling on tethered ligand SARS-CoV-2 RBD:ACE2 construct (same complex as investigated with the AFM and MT). Force and pathway of interface opening (indicated with blue circle) are equivalent to pulling on the interaction without linker as shown in A. Inset (red frame) shows rendering at the end of the simulation, when interface is dissociated and ACE2 and RBD are partially unfolded. (C) Force distribution of interface opening for SARS-CoV-1 (most probable rupture force: 499 pN), SARS-CoV-2 (most probable rupture force: 628 pN), and chimera (most probable rupture force: 597 pN) tethered ligand constructs. Chimera is a tethered ligand construct with the SARS-CoV-1 RBD with 12 aa replacements according to the SARS-CoV-2 RBD. These 12 aa replacements in the SARS-CoV-1 RBD nearly reproduce force stability of SARS-CoV-2. Distributions are from 40 replicas for each construct. (D) Crystal structure of the SARS-CoV-1 RBD:ACE2 interaction (PDB-ID: 2ajf). Residues changed for chimera are indicated in orange. Black lines quantify correlation between residues from the RBD and ACE2 shortly before interface rupture for the chimera system. The overall correlations between RBD and ACE2 residues determine the force propagation pathways through the interface and thus the stability of the interaction. (E and F) Heatmaps of the change in correlation of individual spike residues with ACE2 residues comparing SARS-CoV-1 with the chimera (E) and SARS-CoV-1 with SARS-CoV-2 (F). Residues were selected based on proximity to the interface. Modified residues in the chimera in close proximity with ACE2 are indicated in the y axis labels. Heatmaps indicate a change in mechanostability pattern throughout the interface. Difference in the interaction pattern under force load between SARS-CoV-1 and SARS-CoV-2 RBD:ACE2 interface helps to rationalize their difference in stability.

Fig 5: Binding of ACE2 to the SARS-CoV-2 Spike variants. (a) Substrate-Binding plots for original [D614] S1 spike (red), a/ß/? [D614G] S1 spike (blue), and original [D614] RBD (green) proteins. (b–d) Eadie–Hoffstee plots of data in Part (a) for original [D614] S1 spike (b), a/ß/? [D614G] S1 spike (c), and original [D614] RBD from the S1 spike (d). Note concave-down structure at low binding levels. Each point is the average ± SE for N = 5–6 independent experiments.