A multispecific antibody confers pan-reactive SARS-CoV-2 neutralization and prevents immune escape

Continued evolution of the SARS-CoV-2 spike poses a challenge to immune interventions. To develop antibodies that protect against evolving SARS-CoV-2 viruses, we combined antibodies that recognize different RBD sites to generate a trivalent antibody that potently neutralized all major variants, including the most recent Omicron lineages. Negative stain electron microscopy suggests that this multispecific achieves synergistic neutralization by engaging different epitopes in specific orientations that facilitate inter-spike binding. These interactions resulted in not only improved potency but also importantly prevented virus escape, a feature not seen with parental antibody cocktails or the most potent clinical antibody. Such multispecific antibodies simplify treatment, maximize coverage, decrease the likelihood of SARS-CoV-2 escape, and provide the basis for building universal SARS-CoV-2 antibody therapies that are more likely to maintain broad reactivity for future variants.

Synthesis, cloning and expression of multispecific antibodies After design of the amino acid sequences for each multispecific antibody, the four genes for each multispecific antibody were synthesized using human preferred codons (GenScript) and cloned into 615 eukaryotic expression vectors. For each multispecific antibody expression, equal amounts of the 4 plasmid DNAs were transfected into Expi293 cells (Life Technology) using Expi293 transfection reagent (Life Technology) as previously reported (26). The transfected cells were cultured in shaker incubator at 120 rpm, 37 °C, 9% CO2 for 4~5 days. Culture supernatants were harvested and filtered, the multispecific antibodies were purified over a Protein A (GE Health Science) column. Each multispecific antibody was 620 eluted with IgG elution buffer (Pierce), immediately buffer exchanged with PBS and concentrated using Centricon Plus-70 (Millipore Sigma) membrane filter unit. After concentration, each multispecific antibody was applied to a Superdex 200 16/600 size exclusion column (Cytiva) to remove aggregates and different species in the preparation. The fractions were then analyzed on reduced and non-reduced SDS-PAGE to identify the fractions that contained the monomeric multispecific antibody before combining 625 them. The pooled fractions were then further concentrated, aliquoted and analyzed by SDS-PAGE as well as an analytical SEC column (Superdex 200 16/600) to verify purity. Molecular weight, extinction coefficient and predicted pI were determined using Geneious Prime (Biomatters Ltd.) To make the CODV from multispecific IgG, the FabALACTICA protease (Genovis) was used to digest the IgG for 16 hrs at room temperature. The digestion mixture was then incubated with protein A 630 resin to remove Fc and undigested IgG, the flowthrough and PBS wash of the protein A column that contained the Fab and CODV fragments was collected, concentrated and further purified with size exclusion column (Superose 6 10/300, Cytiva)

635
Codon optimized cDNAs encoding full-length S from SARS-CoV-2 (GenBank ID: QHD43416.1) were synthesized, cloned into the mammalian expression vector VRC8400 (32, 33) and confirmed by sequencing. S containing D614G amino acid change was generated using the wt S sequence. Other variants were made by mutagenesis using QuickChange lightning Multi Site-Directed Mutagenesis Kit (cat # 210515, Agilent) or via synthesis and cloning (Genscript) as previously reported (20,34). The S variants

680
37°C for 1 hour prior to being added to Vero E6 cells. Virus replication was assessed 72hrs after infection in the presence of selected antibodies. Supernatants from the well with the highest concentration of antibody which showed evidence of viral replication (>20% cytopathic effect) was passaged into the subsequent rounds of selection. Infection, monitoring, and collection of supernatants was performed as in the initial round.

685
Expression and Purification of Soluble Spike Constructs The soluble S protein mutants were made in a background of the HexaPro stabilization of the spike (38), incorporating D614G/K444E/L452R and D614G/K444E/F486S, and the protein was produced as previously described (39). One liter of Freestyle cells was transfected with 1mg of SARS-690 CoV-2 spike DNA premixed with 3mL of Turbo293 Transfection Reagent. The cells were grown for 6 days at 37°C, after which the supernatant was collected by centrifugation and filtration. The supernatant was incubated with nickel resin for 1 hour at room temperature, and then the resin was washed with 1X PBS pH 7.4. The spike was eluted with 20mM HEPES pH 7.5, 200mM NaCl, 300mM imidazole and concentrated before loading onto a Superdex S-200 gel filtration column equilibrated in 1X PBS pH 7.4.

695
The trimer containing peak was collected, concentrated to 1mg/ml, flash frozen in liquid nitrogen, and stored at -80°C until use.

Negative Stain Electron Microscopy
SARS-CoV-2 spike proteins were mixed with CODV fragments at a molar ratio of 1:1.2 and 700 incubated at room temperature for 10 min and then diluted to a concentration of approximately 0.02 mg spike/ml with 10 mM HEPES, pH 7.4, 150 mM NaCl. To make a grid, 4.8-µl of the diluted sample was placed on a freshly glow-discharged carbon-coated copper grid for 15 s. The drop on grid was then wicked away with filter paper, and the grid was washed and wicked three times. Same volume of 0.75% uranyl formate was added to the grid to negatively stain protein molecules adsorbed to the carbon and 705 immediately wicked away. After three times staining, the grid was allowed to air-dry. Datasets were collected using a Thermo Scientific Talos F200C transmission electron microscope equipped with a Ceta camera at 200 kV. The nominal magnification was 57,000x, corresponding to a pixel size of 2.53 Å, and the defocus was set at -2 µm. Data was collected automatically using EPU. Single particle analysis was performed using CryoSPARC 3.0.

Figure S1. Production and purification of SARS-CoV-2 CODV antibodies
A. Purity of cross-over dual variable (CODV) immunoglobulin antibodies were evaluated under non-715 reducing (nr) and reducing (r) conditions on a Coomassie SDS-PAGE gel (representative gels shown). B. CODV immunoglobulin bispecific and trispecific antibody traces shown before (top row) and after size exclusion chromatography (SEC) (bottom row). The purple box indicates that fractions combined to make final preparations of the indicated multispecific antibodies. The bottom row shows analytic SEC traces for the purified multispecifics. Shown are representative traces from production runs of antibodies.

720
Antibody properties and yields are shown in Table S1.