A newly identified linear epitope on non-RBD region of SARS-CoV-2 spike protein improves the serological detection rate of COVID-19 patients

Abstract

Background: Serological test is helpful in confirming and tracking infectious diseases in large population with the advantage of fast and convenience. Using the specific epitope peptides identified from the whole antigen as the detection antigen is sensitive and relatively economical. The development of epitope peptide-based detection kits for COVID-19 patients requires comprehensive information about epitope peptides. But the data on B cell epitope of SARS-CoV-2 spike protein is still limited. More importantly, there is a lack of serological data on the peptides in the population. In this study, we aimed to identify the B cell epitope peptides of spike protein and detect the reactivity in serum samples, for further providing data support for their subsequent serological applications.

Results: Two B cell linear epitopes, P104 and P82, located in non-RBD region of SARS-CoV-2 S protein were identified by indirect ELISA screening of an overlapping peptide library of the S protein with COVID-19 patients' convalescent serum. And the peptides were verified by testing with 165 serum samples. P104 has not been reported previously; P82 is contained in peptide S21P2 reported before. The positive reaction rates of epitope peptides S14P5 and S21P2, the two non-RBD region epitopes identified by Poh et al., and P82 and P104 were 77.0%, 73.9%, 61.2% and 30.3%, respectively, for 165 convalescent sera, including 30 asymptomatic patients. Although P104 had the lowest positive rate for total patients (30.3%), it exhibited slight advantage for detection of asymptomatic infections (36.7%). Combination of epitopes significantly improved the positive reaction rate. Among all combination patterns, (S14P5 + S21P2 + P104) pattern exhibited the highest positive reaction rate for all patients (92.7%), as well as for asymptomatic infections (86.7%), confirming the feasibility of P104 as supplementary antigen for serological detection. In addition, we analyzed the correlation between epitopes with neutralizing antibody, but only S14P5 had a medium positive correlation with neutralizing antibody titre (r_s = 0.510, P < 0.01).

Conclusion: Our research proved that epitopes on non-RBD region are of value in serological detection especially when combination more than one epitope, thus providing serological reaction information about the four epitopes, which has valuable references for their usage.

Keywords: Antibody; Epitopes; Humoral immunity; SARS-CoV-2; Spike protein.

Fig. 1

RBD-IgG, RBD-Ab and neutralizing antibody titre of COVID-19 discharged patients’ convalescent sera. a-c RBD-IgG (COI) (a), RBD-Ab (COI) (b) and neutralizing antibody titre (c) of patients’ sera with different clinical types. Cutoff index (COI) value = the sample detection value/cutoff value. Each set of data was represented by the median (interquartile range). Each black dot represented the antibody detection value of each serum sample. d Correlation analysis among the RBD-IgG, RBD-Ab, neutralizing antibody titre, the disease severity and age groups. The figures showed the scatter plots, Spearman correlation coefficients and the red lines were trend lines

Fig. 2

The identification of B cell linear epitopes in SARS-CoV-2 spike protein. a Each peptide group was tested with the test sera from twenty patients by indirect ELISA to screen the positive peptide groups (G9 and G11). b-c Performing indirect ELISA tests for each peptide contained in the selected positive peptide groups G9 and G11 with test sera to screen for the positive peptides (P82 and P104). Twenty convalescent serum samples randomly selected from the cohort that included fourteen moderate patients, three mild patients and three asymptomatic infections were pooled as test sera. Twenty serum samples of healthy people were pooled as negative sera. RBD of spike protein served as positive antigen control (data were not shown). The cutoff value was calculated as the mean + 3 × standard deviations (SD) of negative sera. The S/CO value (sample average OD_{450 nm–630 nm} value/cutoff value) of test sera was greater than 1 was considered to be a positive response. The red dots represented the S/CO values of test sera to peptides. The blue squares represented the S/CO values of negative sera to peptides. The grey dotted line indicated 1. Red dots above the grey line were considered to be a positive response. d The localization of two positive epitope peptides in spike protein. The 3D structure of SARS-CoV-2 spike protein was obtained from the NCBI website (PDB ID:6VXX). The image was processed through the NCBI website. The three monomers of spike protein were represented respectively as blue, pink, and grey. The positions of RBD, P82, P104 and S2 subunit in one monomer were distinguished by red, purple, green, brown and indicated by arrows

Fig. 3

The positive reaction rates of epitope peptide S21P2, S14P5, P82 and P104 in convalescent sera of COVID-19 patients. Each convalescent serum sample of COVID-19 patients was tested with each epitope peptide by indirect ELISA. a The positive reaction rates of peptides S21P2, S14P5, P82 and P104 in total COVID-19 patients and asymptomatic infections. b Analyses of the positive reaction rates of different peptide combinations. S14P5, S14P5 + S21P2 and S14P5 + S21P2 + P104 were the best combinations with the highest overall positive reaction rate in peptide combinations respectively contained the single, two or the three epitope peptides. As long as the serum sample was positive for any one of the epitope peptides in combination, the sample was considered to be positive for the epitope peptide combination. c The positive reaction rates in different clinical types detected by different peptide combinations

Fig. 4

IgG reactive level (S/CO values) of epitope peptides. a The positive IgG reactive level (S/CO values) in convalescent serum samples and the S/CO values of negative sera to four epitope peptides respectively. b Positive IgG S/CO values of four epitope peptides with different clinical type patients. Each set of data was represented by the median (interquartile range)