SARS-CoV-2 variants preferentially emerge at intrinsically disordered protein sites helping immune evasion

Abstract

The SARS-CoV-2 pandemic is maintained by the emergence of successive variants, highlighting the flexibility of the protein sequences of the virus. We show that experimentally determined intrinsically disordered regions (IDRs) are abundant in the SARS-CoV-2 viral proteins, making up to 28% of disorder content for the S1 subunit of spike and up to 51% for the nucleoprotein, with the vast majority of mutations occurring in the 13 major variants mapped to these IDRs. Strikingly, antigenic sites are enriched in IDRs, in the receptor-binding domain (RBD) and in the N-terminal domain (NTD), suggesting a key role of structural flexibility in the antigenicity of the SARS-CoV-2 protein surface. Mutations occurring in the S1 subunit and nucleoprotein (N) IDRs are critical for immune evasion and antibody escape, suggesting potential additional implications for vaccines and monoclonal therapeutic strategies. Overall, this suggests the presence of variable regions on S1 and N protein surfaces, which confer sequence and antigenic flexibility to the virus without altering its protein functions.

Keywords: DisProt; IDPs; SARS-CoV-2; ViralZone; biocuration; immune escape; intrinsically disordered proteins; mutations; variants.

Fig. 1

Amino acid changes in the spike protein of Variants of Concern (VOC) Alpha, Beta, Gamma, Delta, Omicron BA.1 and BA.2; Variants of Interest (VOI) Lambda, Mu, Epsilon, Zeta, Eta, Theta, Iota and Kappa. Disordered regions are indicated by cyan columns, and variants in disordered regions are coloured in red. Parentheses indicate variants whose prevalence is < 80% but > 10%. The main regions are annotated: S1 with N‐terminal domain (NTD) and receptor‐binding domain (RBD); S2 with fusion peptides, heptad repeat 1 (HR1) and 2 (HR2) and the transmembrane domain (TM) [73]. Major antigenic sites are shown below with the NTD supersite [56], and monoclonal antibody‐binding regions for sotrovimab [74], casirivimab and imdevimab [75, 76].

Fig. 2

Amino acid changes in the nucleoprotein of Variants of Concern Alpha, Beta, Gamma, Delta, Omicron BA.1 and BA.2; Variants of Interest (VOI) Lambda, Mu, Epsilon, Zeta, Eta, Theta, Iota and Kappa. Disordered regions are indicated by cyan columns, and variants in disordered regions are coloured in red. Parentheses indicate variants whose prevalence is < 80% but > 10%.

Fig. 3

Immune escape‐related mutations mapped on the IDRs of the spike protein (structure in closed conformation) [61]. The disordered regions—according to the DisProt database (protein S, DisProt: DP02772) ‐ are coloured in light brown on the structure, while mutations are highlighted in dark brown. Molecular graphics were performed using UCSF Chimera [71].

Fig. 4

Mutations in the VOCs and VOI lineages mapped on the sequences of ORF3a, E protein and ORF7a. VOCs and VOIs lineages are represented, along with the mutations falling inside (red) and outside (black) IDRs. IDRs are represented as cyan columns while transmembrane regions are in grey.

Fig. 5

Mutations in the VOCs and VOI lineages mapped on the sequences of ORF1ab polyprotein. VOCs and VOIs lineages are represented, along with the mutations falling inside (red) and outside (black) IDRs. IDRs are represented as cyan columns.