Expression and characterization of SARS-CoV-2 spike proteins

Abstract

The severe acute respiratory syndrome coronavirus 2 spike protein is a critical component of coronavirus disease 2019 vaccines and diagnostics and is also a therapeutic target. However, the spike protein is difficult to produce recombinantly because it is a large trimeric class I fusion membrane protein that is metastable and heavily glycosylated. We recently developed a prefusion-stabilized spike variant, termed HexaPro for six stabilizing proline substitutions, that can be expressed with a yield of >30 mg/L in ExpiCHO cells. This protocol describes an optimized workflow for expressing and biophysically characterizing rationally engineered spike proteins in Freestyle 293 and ExpiCHO cell lines. Although we focus on HexaPro, this protocol has been used to purify over a hundred different spike variants in our laboratories. We also provide guidance on expression quality control, long-term storage, and uses in enzyme-linked immunosorbent assays. The entire protocol, from transfection to biophysical characterization, can be completed in 7 d by researchers with basic tissue cell culture and protein purification expertise.

Fig. 1 |. Overview of SARS-CoV-2 spike structure and antigenicity.

a, Schematic of the SARS-CoV-2 spike domains subdivided by color. NTD, N-terminal domain; RBD, receptor-binding domain; FP, fusion peptide; HR1, heptad repeat 1; CH, central helix; CD, connector domain. White region is unresolved in cryo-EM structures of the ectodomain. Glycan positions noted on top of the primary structure as previously reported. Scissors denote the S1/S2 and S2′ proteolytic cleavage sites. Location of stabilizing HexaPro prolines is denoted. b,c, Spike structures in the three RBD down (PDB:6VXX) (b) and two RBD up prefusion states (PDB:6X2B) (c). Left: front view, Right: top view. HexaPro prolines shown in black. d, Composite structure of a spike monomer bound to neutralizing antibodies CR3022 (PDB: 6W41) and 4A8 (PDB:7C2L).

Fig. 2 |. Purification and characterization of the SARS-CoV-2 spike.

a, Overview and timeline of the protocol. b, SDS-PAGE gel of recombinant spike after elution from the Strep-Tactin column. c, Superose-6 traces of S-2P and HexaPro. Note the higher yield for HexaPro. d, Differential scanning fluorimetry analysis. e, A negative-stain electron micrograph. f, 2D class averages of HexaPro.

Fig. 3 |. BLI analysis of spike-ACE2 binding.

a, Schematic of the BLI experiment. Recombinant spikes with a C-terminal TwinStrep epitope are immobilized on a SA-coated tip. Association and dissociation of ACE2 is measured over time. b, Stepwise preparation of the BLI tip. Highest response is visualized with pre-binding spike and challenging with ACE2. c, Determination of the association (k_on) and dissociation (k_off) constants of spike for ACE2.

Fig. 4 |. ELISA of monoclonal antibodies and patient serum.

a, An indirect ELISA used for spike detection. b, Binding profile of five purified mAbs against the spike trimer. c, Binding profile of five donor plasma samples against the spike trimer.