The impact of SARS-CoV-2 nucleocapsid protein mutations on the performance of rapid antigen tests

In a recent study published on bioRxiv* preprint server, researchers designed a lentivirus-based mammalian surface display platform combined with deep mutation scanning (DMS) to assess the impact of mutations in the coronavirus nucleocapsid (N) protein of the severe acute respiratory syndrome 2 (SARS-CoV-2) on rapid performance antigen trials.

Frequent and widespread testing for SARS-CoV-2 is essential to curbing the coronavirus disease 2019 (COVID-19) pandemic. Rapid antigen tests that detect SARS-CoV-2 N proteins are the diagnostic tests of choice in several settings due to their ease of use and short turnaround time. However, the unprecedented emergence of SARS-CoV-2 variants has raised major concerns about test performance against mutant variants.

About the study

In the present study, researchers developed a mammalian surface display platform that allows quantitative and direct measurement of antibody binding with SARS-CoV-2N. The platform was used with the DMS to assess the vulnerability of rapid antigen tests to SARS-CoV-2 mutations N.

For the analysis, 17 antibodies with Emergency Use Authorization (EUA) from the Food and Drug Administration (FDA) used in current SARS-CoV-2 rapid antigen tests were used. Two DMS libraries containing nearly all N protein mutations were then packaged into lentiviral particles for transduction of human embryonic kidney 293 (HEK293) cells.

An expression construct was generated, which contained an N protein flanked by an N-terminal (SP) signal peptide and a C-terminal transmembrane (TM) region, which were derived from immunoglobulin G4 (IgG4) and the receptor platelet-derived growth factor (PDGFR), respectively. A master regulator of cell cycle entry and proliferative metabolism (Myc)-tag was inserted between N and SP protein to control protein expression differences.

The construct was cloned into a plasmid expressing a lentivirus followed by a green fluorescent protein (GFP). Myc+ and GFP+ cells were analyzed by flow cytometry. In addition, biolayer interferometry (BLI) experiments were performed. In DMS, barcodes from each sample were counted to determine weighted escape scores (E_w), which reflect the extent to which antibody binding has been reduced by a particular mutation.

For further DMS validation, cloning of escape mutations of three antibodies (C706, R040, and 3C3) having different epitope types and locations was performed. Antibody binding to point mutations was assessed by antibody titration experiments.

Results

Of 7,942 mutations of the entire N protein sequence, DMS libraries #1 and #2 included 7,893 and 7,901 mutations, respectively. All 17 antibodies showed distinctive escape mutation patterns in the N protein dimerization domain (NDD), and most N mutations did not affect antibody binding.

The linear epitope of R040 was confined to a continuous stretch of amino acids between amino acid residues L394 and F403, and the D399N mutation exhibited a high E_w for the antibody. On the other hand, the 3D epitopes of the 3C3 and C706 antibodies were characterized by discontinuous amino acid stretches with varying degrees of mutational escape. All escape mutations tested for antibodies C706 (F110S, G85K, R149D, and G116R) and R040 (D399N, D402V, and L395V) abolished antibody binding, while mutations outside the epitope had no no effect. For the 3C3 antibody. V324E and E323V mutations abolished binding, while P326A and T329G mutations reduced binding affinity. The escape sites were located in the hydrophobic core of the SARS-CoV-2 N receptor binding domain (RBD) and likely deployed this domain.

Titration experiments with 3C3 and N protein mutants showed that escape scores were affected by reduced availability of conformational epitopes and by reductions in antibody binding affinities and that although the total binding signals of antibodies were reduced, the binding affinity remained unchanged.

N-RBD epitopes have been grouped into four classes. Class I epitopes were identified for C524, MM08, 1A7, and RC17604 and were linked to loop sites located between residues K143 and Y123 and on the NDD opposite the beta (β) hairpin. MM05 defined class II epitopes and its binding region was located in the β-hairpin loop at positions D98, M101 and K95. Escape mutations of class III epitopes and class IV epitopes were located in the C-terminal loop regions between V158 and A173 for class III and between R149 and P151 for class IV epitopes.

Secondary escape sites were also observed for a few antibodies (N-Ab3, Ab166, 1C1, 1A7 and R004) at the N-terminus of the N protein in residues S2 to P6. Additionally, monoclonal antibody-1 (mAb-1) and mAb-2 used in the commercial Clip COVID rapid antigen test bound to N-RBD with comparable affinity to MM08 and R004, respectively. Detection of secondary escape sites and comparable results with commercial rapid antigen tests highlight the accuracy of DMS.

The D399N mutation exhibited a high E_w, while all other mutations exhibited low E_w, indicating that SARS-CoV-2 variants containing these mutations did not affect the performance of rapid antigen tests using the antibodies. Besides D399N, D3L, present in variant B.1.1.7, was the only mutation with a slightly elevated E_w for Ab166, 1C1 and 3C3 antibodies and could affect their assay performance.

Conclusion

Overall, the study results showed that rapid antigen testing with the DMS approach could accurately identify antibody escape mutations in the SARS-CoV-2 N protein. The approach was able to identify locations of escape mutations and determining how the mutations affected antibody identification. DMS provided a comprehensive map of antibody epitopes and their susceptibility to mutational escape.

*Important Notice

bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be considered conclusive, guide clinical practice/health-related behaviors, or treated as established information.