Structure-based analyses of neutralization antibodies interacting with naturally occurring SARS-CoV-2 RBD variants

Dear


Mutational abundance analysis
Spike protein sequences were obtained from GISAID 1,2 and NCBI 3 databases. Metadata was from 2019nCoVR database developed and maintained by CNCB-NGDC 4 . To remove redundancies and filter sequences, SeqKit 5 was used to parse sequence ids, and filter sequences according to the metadata. We use MAFFT 7 6 to align all other sequences to the reference 7 along with the '--auto' option to identify mutation. Abundance and frequency were calculated using Python packages pandas 8 and NumPy 9 and visualized using matplotlib 10 and seaborn. R package UpSetR 11 was used to generate UpSet 12 plots.

Protein expression and purification
The receptor binding domain, as well as its mutants, of SARS-CoV-2 spike protein (residue 319-541) were cloned into pcDNA3.1 vector with an IL2 signal peptide at the N-terminal and a C-terminal 8×His-tag. Plasmids were transfected into HEK293F cells (cell density between 1~1.5 million cells/mL) using polyethylenimine (Polysciences). The conditioned media were collected after four days, and the proteins were purified using a Ni-NTA affinity column (GE Life Sciences), and Superdex 200 column (GE Life Sciences) in the final buffer: 20 mM HEPES, pH 7.2, 150 mM NaCl. For crystallization, the SARS-CoV-2 RBD and its mutants with an N-terminal 6×His-tag were cloned into pFastBac vector with a gp67 signal peptide, then bacmids were generated by using the Bac-to-Bac system (Invitrogen), SF-21 insect cells were then used to generate baculoviruses. Hi-5 insect cells were used to express proteins infected by recombinant baculoviruses when cell density reached 1.8 million cells/mL. The conditioned media were collected after 48 hours, concentrated and exchanged into 25 mM Tris, pH 8.0, 150 mM NaCl, then the proteins were purified as previously described.
The Fabs of BD-218, BD503, BD-508, BD-515, BD-604 and BD-623 were expressed using HEK293F cells. The heavy chains and light chains were cloned respectively into pcDNA3.1 vector with signal peptide and a C-terminal 6×His-tag. Plasmids of the heavy chain and light chain were admixed at a 1:1 ratio and then transfected into HEK293F cells by using polyethylenimine. The conditioned media were collected after 4 days, concentrated and exchanged into the binding buffer contained 25 mM Tris, pH 8.0, 150 mM NaCl. Then proteins were purified by Ni-NTA affinity column, and by gel filtration column Superdex 200 into the final buffer. The BD-368-2 Fab was obtained as previously described 13 .

Surface plasmon resonance
A Biacore T200 (GE Healthcare, USA) was used to measure and compare the dissociation coefficients between antibodies and SARS-CoV-2 RBD as well as its mutants. Fabs of the antibodies were captured to 200~300 RU on a Series S Sensor CM5 Chip (GE Healthcare). Then serial dilutions of SARS-CoV-2 RBD and mutants were injected, with concentrations from 20 to 0.63 nM (2-fold dilutions). All proteins were exchanged into running buffer containing 10 mM HEPES, pH 7.4, 150 mM NaCl, 3 mM EDTA and 0.005% (v/v) P20. The final data were processed by the Biacore Evaluation Software and fit to a 1:1 binding model.

Crystallization
The complexes of Fabs with SARS-CoV-2 RBD or the mutants were obtained by mixing the proteins at equimolar ratios and incubated at 4 ℃ for one hour, the complexes were then purified by Superdex 200 column in the final buffer. Purified complexes were concentrated to 10 mg/ml, crystallization then performed using the sitting-drop vapor diffusion method at 18 ℃.

Data collection and structure determination
All crystals were picked by cryo-loops and placed in the corresponding reservoir solution, containing 20% (v/v) glycerol before flash-frozen in liquid nitrogen. All data were collected at beamline BL19U of the National Facility for Protein Science Shanghai, beamline BL17U of the Shanghai Synchrotron Radiation Facility (SSRF) and beamline BL1A of the KEK Photon Factory. The data were integrated and scaled using HKL2000 (HKL Research, USA). Further calculation was performed using molecular replacement in the suite Phenix 13 . All structural models were refined using Phenix and manually corrected in Coot 14 .

Pseudovirus neutralization assay
The pseudovirus neutralization assays were performed using Huh-7 cell lines. Pseudovirus were prepared as previously described 16 . Various concentrations of antibodies (3-fold serial dilution using DMEM) were mixed with the same volume of SARS-CoV-2 pseudovirus in a 96 well-plate. The mixture was incubated for 1 h at 37 ℃ and supplied with 5% CO 2 . Pre-mixed Huh-7 cells were added to all wells and incubated for 24 h at 37 ℃ and supplied with 5% CO 2 . After incubation, the supernatants were removed, and D-luciferin reagent (Invitrogen) was added to each well and measured luciferase activity using a microplate spectrophotometer (PerkinElmer EnSight). The inhibition rate is calculated by comparing the OD value to the negative and positive control wells. IC 50 were determined by a four-parameter