Fig. 4: AMK-1057 is a cyclic peptide that binds human-derived Spike RBD in vitro.

a Schematic of the construct used to express AMK-1057. Leader and core amino acids are shown and predicted modified core residues highlighted in light blue. TEV protease sites are annotated as TEVp sites. b The expression, modification, cleavage, and purification steps are shown. TEV protease cleavage removes the SUMO tag/leader peptide and HPLC purification produces the final product; it leaves a single G, shown as a darker amino acid in the core. c High-resolution MS of unmodified AMK-1057 (black trace) and singly modified AMK-1057 (orange trace). d High-resolution MS/MS of modified AMK-1057 and fragment mapping to the amino acid sequence. Numbered peaks correspond to fragment ions observed and represented as lettered amino acids next to the MS/MS spectrum. e Structural annotation of AMK-1057. Thioether modification is colored to highlight the macrocycle. See Methods for details. f Binding of purified peptides (modified AMK-1057, blue; unmodified AMK-1057, gray; ACE2α1, orange) to human cell line-derived Spike RBD_296-531. Three replicates performed on different days are shown. g Space-filling models of the Spike RBD (PDB ID: 6M17) colored to indicate annotated binding regions of ACE2 (pink)⁶⁶, as well as antibodies CR3022 (orange)⁶⁷ and B38 (blue)⁶⁸. Images were generated using ChimeraX⁸². The model is rotated 180°. h–j Competition BLI experiments measuring binding of purified, modified AMK-1057 to human cell-derived Spike RBD pre-incubated with h soluble hACE2 (ΔΔnm = 0.99 ± 0.02), i antibody B38 (ΔΔnm = 0.81 ± 0.03), and j antibody CR3022 (ΔΔnm = 0.71 ± 0.03). Δnm was measured at 600 s and normalized to no AMK-1057 control to account for competitor off-rates (ΔΔnm). Dashed lines separate association reactions from dissociation reactions (Methods). For all experiments, three replicates were performed on different days and a representative trace is shown.