Customizably designed multibodies neutralize SARS-CoV-2 in a variant-insensitive manner

Abstract

The COVID-19 pandemic evolves constantly, requiring adaptable solutions to combat emerging SARS-CoV-2 variants. To address this, we created a pentameric scaffold based on a mammalian protein, which can be customized with up to 10 protein binding modules. This molecular scaffold spans roughly 20 nm and can simultaneously neutralize SARS-CoV-2 Spike proteins from one or multiple viral particles. Using only two different modules targeting the Spike's RBD domain, this construct outcompetes human antibodies from vaccinated individuals' serum and blocks in vitro cell attachment and pseudotyped virus entry. Additionally, the multibodies inhibit viral replication at low picomolar concentrations, regardless of the variant. This customizable multibody can be easily produced in procaryote systems, providing a new avenue for therapeutic development and detection devices, and contributing to preparedness against rapidly evolving pathogens.

Keywords: Omicron; antibody; avidity; coronavirus; entry inhibitor; multi-target; neutralization; protein design.

Figure 1

Molecular representation of the construct and RBD binding capacity. (A) Left: Molecular representation of the soluble region of the Spike protein as determined in the PDB structure 7JZL. One of the monomers is shown in cartoon representation (green), with the epitopes on the RBD highlighted in blue (epitopes A/B) and red (epitope C). Right: Molecular model of a Xiang-Liu 1 molecule featuring binders A and B at the extremities (blue and light blue, respectively). Both figures are drawn at the same scale to emphasize the possibility of Xiang-Liu binding more than one Spike trimer. The inset shows the cysteine residues forming the disulfide bridges at the extremity of the five-helix bundle. (B) Cartoon representation of an RBD molecule (green) with binders A (blue), B (light blue), and C (red). The semitransparent surface indicates the position of hACE2. Black spheres indicate the position of mutations reported for variants before Omicron, while those in orange correspond to Omicron. (C) Schematic representation of a Xiang-Liu construct illustrating possible interaction modes with different trimeric spikes on the same virion or neighboring viral particles. (D) Dose-response cell surface binding inhibition curves of Wuhan RBD in the presence of isolated binders and Xiang-Liu molecules. Calculated IC50 and IC90 are shown in the table below. (E) Dose-response cell surface binding inhibition curves in the presence of different RBDs from Wuhan, Delta, and Omicron. Analogous curves for isolated binders are shown in Figure S3 . Calculated IC50 and IC90 are shown in the table below. (F) RBD variant recognition by antibodies from SARS-CoV-2 post-immunization human serum. Immunoassay immobilizes different RBDs VOCs and revealing the binding antibodies with HRP-conjugated anti-human IgG. (G) Percentage of antibodies displacement upon increasing concentrations of Xiang-Liu molecules (from 1.6 nM to 136 nM) calculated from ELISA experiments immobilizing different RBDs VOCs and revealing the binding antibodies from human post-immunization SARS-CoV-2 serum with HRP-conjugated anti-human IgG. We show the average value among two replicates and their standard deviation in all cases.

Figure 2

SARS-CoV-2 Neutralization and Inhibition. (A) Neutralization assay on pseudotyped SARS-CoV-2 particle variants adding XL-1 and XL-2 multibodies. Additionally, equivalent concentrations of binders A and C were added to estimate the avidity gain provided by the pentameric constructs. Calculated IC50 and IC90 are shown in the tables below. (B, C) Infection inhibition of SARS-CoV-2 infective viruses with XL-1 and XL-2. (B) Representative image of infected Huh7-hACE2 cell, the nucleus in red and SARS-CoV2 nucleocapsid in green. (C) Virus infection inhibition with different concentrations of XL-1 and XL-2 molecules.