Naturally enhanced neutralizing breadth against SARS-CoV-2 one year after infection

Abstract

More than one year after its inception, the coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) remains difficult to control despite the availability of several working vaccines. Progress in controlling the pandemic is slowed by the emergence of variants that appear to be more transmissible and more resistant to antibodies^1,2. Here we report on a cohort of 63 individuals who have recovered from COVID-19 assessed at 1.3, 6.2 and 12 months after SARS-CoV-2 infection, 41% of whom also received mRNA vaccines^3,4. In the absence of vaccination, antibody reactivity to the receptor binding domain (RBD) of SARS-CoV-2, neutralizing activity and the number of RBD-specific memory B cells remain relatively stable between 6 and 12 months after infection. Vaccination increases all components of the humoral response and, as expected, results in serum neutralizing activities against variants of concern similar to or greater than the neutralizing activity against the original Wuhan Hu-1 strain achieved by vaccination of naive individuals^2,5-8. The mechanism underlying these broad-based responses involves ongoing antibody somatic mutation, memory B cell clonal turnover and development of monoclonal antibodies that are exceptionally resistant to SARS-CoV-2 RBD mutations, including those found in the variants of concern^4,9. In addition, B cell clones expressing broad and potent antibodies are selectively retained in the repertoire over time and expand markedly after vaccination. The data suggest that immunity in convalescent individuals will be very long lasting and that convalescent individuals who receive available mRNA vaccines will produce antibodies and memory B cells that should be protective against circulating SARS-CoV-2 variants.

Fig. 1. Plasma ELISAs and neutralizing activity.

a–d, Plasma IgG antibody binding to SARS-CoV-2 RBD (a) and N protein (b) shown as area under the curve (AUC; numbers in red are mean geometric AUC), and plasma neutralizing activity (NT₅₀) in unvaccinated (c) and vaccinated (vac) (d) individuals 12 months after SARS-CoV-2 infection (n = 63). n = 63 individuals, 37 convalescent unvaccinated (black) and 26 convalescent vaccinated (blue) individuals. a, b, Two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons. c, d, Lines connect longitudinal samples from the same individual. Two-sided Friedman test with subsequent Dunn’s multiple comparisons. Two individuals who received their first dose of vaccine 24–48 h before sample collection are represented in purple. e, Plasma neutralizing activity against indicated SARS-CoV-2 variants of concern (n = 30, 15 convalescent and 15 convalescent vaccinated individuals). The B.1.526 variant used here contains the E484K substitution. Substitutions, deletions and insertions in S variants used here are described in Methods. Two-tailed Mann–Whitney test. Red numbers in c–e indicate the geometric mean NT₅₀ at the indicated time point. All experiments were performed at least in duplicate.

Fig. 2. SARS-CoV-2 RBD-specific B cell memory.

a, Number of antigen-binding memory B cells per 2 × 10⁶ B cells (Extended Data Fig. 5b, c) obtained at 1.3, 6.2 and 12 months after infection from 40 randomly selected individuals (vaccinated, n = 20; non-vaccinated, n = 20). Each dot represents one individual. Red horizontal bars indicate geometric mean values. Two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons. WT, wild type. b, The distribution of antibody sequences from 6 individuals 1.3 (top) or 6.2 (middle) or 12 (bottom) months after infection^,. The number in the inner circle indicates the number of sequences analysed for the individual whose identifier is denoted above the circle. Pie slice size is proportional to the number of clonally related sequences. The outer black arc indicates the frequency of clonally expanded sequences detected in each participant. Coloured slices indicate persisting clones (same IGV and IGJ genes, with highly similar CDR3 sequences) found at both time points in the same participant. Grey slices indicate clones unique to the time point. White indicates sequences isolated once, and white slices indicate singlets found at both time points. c, Number of somatic nucleotide mutations (SHM) in the IGVH and IGVL genes (Supplementary Table 3) obtained after 1.3 or 6.2 or 12 months (1.3 month, n = 889; 6.2 month, n = 975; 12 month, n = 1,105 (unvaccinated, n = 417; vaccinated, n = 688)). Red horizontal bars indicate mean values. Two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons.

Fig. 3. Anti-SARS-CoV-2 RBD monoclonal antibodies.

a, EC₅₀ for SARS-CoV-2 RBD of antibodies isolated at 1.3 (n = 152) 6.2 (n = 153) and 12 (n = 174) months after infection^,, determined by ELISA. Two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons (1.3 months versus 6.2 months, P = 0.27; 1.3 months versus 12 months, P = 0.0075; 6.2 versus 12 months, P < 0.0001). b, SARS-CoV-2-neutralizing activity of monoclonal antibodies measured using a SARS-CoV-2 pseudovirus neutralization assay^,. IC₅₀ values for antibodies isolated at 1.3, 6.2 and 12 months after infection against wild-type SARS-CoV-2 (Wuhan-Hu-1 strain) are shown. Each dot represents one antibody. Pie charts illustrate the fraction of non-neutralizing (IC₅₀ > 1,000 ng ml⁻¹) antibodies (grey slices), inner circle shows the number of antibodies tested per group. Horizontal bars and red numbers indicate geometric mean values. Statistical significance was determined through the two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons.

Fig. 4. Epitope targeting and evolution of anti-SARS-CoV-2 RBD antibodies.

a, Schematic of the BLI experiment (left) and IC₅₀ values for randomly selected neutralizing and non-neutralizing antibodies (Ab1 and Ab2) isolated at 1.3 and 12 months after infection (n = 15 antibodies per group, n = 60 antibodies in total). Red horizontal bars indicate geometric mean. Two-sided Mann–Whitney test. b, Dissociation constants (K_D) of the n = 30 neutralizing (green) and n = 30 non-neutralizing (red) antibodies shown in a. Horizontal bars indicate geometric mean values. Two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons. BLI traces are shown in Extended Data Fig. 8. c, Heat map of relative inhibition of binding of a monoclonal antibody (Ab2) to preformed complexes of RBD with another monoclonal antibody (Ab1) (grey, no binding; orange, intermediate binding; red, high binding). Data are normalized by subtraction of the autologous antibody control. BLI traces are shown in Extended Data Fig. 9. d, Neutralization of the indicated mutant RBD proteins with antibodies shown in a–c. Pie charts illustrate the fraction of antibodies that are poorly or non-neutralizing (IC₅₀ of 100–1,000 ng ml⁻¹, red), intermediate neutralizing (IC₅₀ of 10–100 ng ml⁻¹, pink) and potently neutralizing (IC₅₀ of 0–10 ng ml⁻¹, white) for each mutant. The number in the inner circle shows the number of antibodies tested. e, Graphs show affinities (y-axis) plotted against neutralization activity (x-axis) for 18 clonal antibody pairs isolated 1.3 (top) and 12 months (bottom) after infection (n = 36 antibodies). Spearman correlation test. f, BLI affinity measurements for same n = 36 paired antibodies as in e. Two-tailed Wilcoxon test. g, IC₅₀ values for n = 30 paired neutralizing antibodies isolated at indicated time points versus indicated mutant SARS-CoV-2 pseudoviruses. Antibodies are divided into groups I, II and III (left), on the basis of neutralizing activity: I, potent clonal pairs that do not improve over time; II, clonal pairs that show increased activity over time; and III, clonal pairs showing decreased neutralization activity after 12 months. Antibody class assignment based on initial (1.3 month after infection) sensitivity to mutation is indicated on the right. Red stars indicate antibodies that neutralize all tested RBD mutants. Colour gradient indicates IC₅₀ values ranging from 0 (white) to 1,000 ng ml⁻¹ (red).

Extended Data Fig. 1. Clinical correlations.

a–d, Association of persistence of symptoms (Sx) 12 months after infection with various clinical and serological parameters in our cohort of individuals who recovered from COVID-19 (n = 63). a, b, Acute disease severity as assessed with the WHO Ordinal Scale of Clinical Improvement (a, P = 0.99) and duration of acute phase symptoms (b, P = 0.63) in individuals reporting persistent symptoms (+) compared to individuals who are symptom-free (−) 12 months post-infection. c, Proportion of individuals reporting persistent symptoms (black area) compared to individuals who are symptom-free (grey area) 12 months after infection grouped by vaccination status (P = 0.72). d, Anti-RBD IgG (P = 0.75), anti-N IgG (P = 0.15), the RBD/N IgG ratio (P = 0.73), and NT₅₀ titers (P = 0.38) at 12 months after infection in individuals reporting persistent symptoms (+) compared to individuals who are symptom-free (−) 12 months post-infection. Statistical significance was determined using the two-tailed Mann–Whitney test in a, b, d, and using the two-sided Fisher’s exact test in c.

Extended Data Fig. 2. Plasma activity.

a–h, ELISA results for plasma against SARS-CoV-2 RBD 12 months after infection (n = 63). Non-vaccinated individuals are depicted with black circles and lines, and vaccinated individuals are depicted in blue throughout. Two outlier individuals who received their first dose of vaccine 24–48 h before sample collection is depicted as purple circles. a–n, IgM (a–d) IgG (e–g) and IgA (h–k) antibody binding to SARS-CoV-2 RBD and IgG binding to N (l–n) 12 months after infection. a, e, h, i, ELISA curves from non-vaccinated (black lines) individuals, as well as individuals who received one or two doses (blue lines) of a COVID-19 mRNA vaccine (left panels). Area under the curve (AUC) over time in non-vaccinated (b, f, i, m) and vaccinated individuals (c, g, j, n). Lines connect longitudinal samples. d, k, Boxplots showing AUC values of all 63 individuals, as indicated. o, ranked average NT₅₀ at 1.3 months (light grey) and 6.2 months (dark grey), as well as at 12 months for non-vaccinated (orange) individuals, and individuals who received one or two doses (blue circles) of a COVID-19 mRNA vaccine, respectively. Two individuals who received their first dose of vaccine 24–48 h before sample collection is depicted in purple. p–r, Correlation of serological parameters in non-vaccinated (black circles and black statistics) and vaccinated (blue circles and blue statistics) individuals. Two individuals who received their first dose of vaccine 24-48 h before sample collection is depicted as purple circles. Correlation of 12-month titers of anti- RBD IgG and NT₅₀ (p), anti-RBD IgG and N IgG (q), and anti-N IgG and NT₅₀ (r). s, Plasma neutralizing activity against authentic virus isolates WA1/2020 and B.1.351, as indicated (n = 6). Statistical significance was determined using two-sided Friedman test with subsequent Dunn’s multiple comparisons (b, c, f, g, i, j, m, n), or two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons (d, k) or using the Spearman correlation test for the non-vaccinated and vaccinated subgroups independently (p–r) or using two-tailed Mann–Whitney test (s). Red numbers indicate the geometric mean NT₅₀ at the indicated time point. All experiments were performed at least in duplicate.

Extended Data Fig. 3. Flow cytometry.

a, Gating strategy. Gating was on singlets that were CD20⁺ and CD3-CD8-CD16-Ova-. Anti-IgG, IgM, and IgA antibodies were used for B cell phenotype analysis. Sorted cells were RBD-PE⁺ and RBD/KEN-AF647⁺. b, c, Flow cytometry showing the percentage of RBD-double positive (b) and 647-K417N/E484K/N501Y mutant RBD cross-reactive (c) memory B cells from 1.3 or 6- and 12-months post-infection in 10 selected participants. d, As in Fig. 2b, Pie charts show the distribution of antibody sequences from 4 individuals after 1.3 (upper panel) or 6.2 months (middle panel) or 12 months (lower panel). e, Circos plot depicts the relationship between antibodies that share V and J gene segment sequences at both IGH and IGL. Purple, green, and grey lines connect related clones, clones and singles, and singles to each other, respectively. f, Graph summarizes cell number (indicated in b, c) (per 2 million B cells) of immunoglobulin class of antigens binding memory B cells in samples obtained at 1.3, 6.2 and 12 months. Each dot is one individual. (Vaccinees, n = 20, and non-vaccinees, n = 20). Red horizontal bars indicate mean values. Statistical significance was determined using two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons.

Extended Data Fig. 4. Frequency distribution of human V genes.

Graph shows comparison of the frequency distributions of human V genes of anti-SARS-CoV-2 antibodies from donors at 1.3, 6.2, 12 months after infection. a, Graph shows relative abundance of human IGVH genes Sequence Read Archive accession SRP010970 (green), convalescent vaccinees (blue), and convalescent non-vaccinees (orange). Statistical significance was determined by two-sided binomial test. b, c, Same as in a, but showing comparison between antibodies from donors at 1.3 months (b), 6.2 month (c) and 12 months after infection. Two-sided binomial tests with unequal variance were used to compare the frequency distributions., significant differences are denoted with stars (* P < 0.05, ** P < 0.01, *** P < 0.001, **** = P < 0.0001).

Extended Data Fig. 5. Analysis of anti-RBD antibodies.

a, Number of clonally expanded B cells (per 10 million B cells) at indicated time points in 10 individuals. Colours indicate shared clones appearing at different time points. Statistical significance was determined using two-tailed Wilcoxon matched-pairs signed rank test. Vaccinees are marked in red. Statistical significance was determined using Wilcoxon matched-pairs signed rank tests. Vaccinees are marked in red. b, Number of somatic nucleotide mutations in the IGVH (top) and IGVL (bottom) in antibodies obtained after 1.3 or 6.2 or 12 months from the indicated individual. c, Same as b, but graphs show comparison between new clones and conserved clones in 6 vaccinated convalescent individuals at 12 months after infection. d, The amino acid length of the CDR3 s at the IGVH and IGVL for each individual. Right panel shows all antibodies combined. (1.3m: n = 889; 6.2m: n = 975; 12m: n = 1105, (non-vax: n = 417; vax: n = 688)). The horizontal bars indicate the mean. Statistical significance was determined using two-sided Kruskal–Wallis test with subsequent Dunn’s multiple comparisons (a, b, d), or two-tailed Mann–Whitney U-tests (c).

Extended Data Fig. 6. Evolution of anti-SARS-CoV-2 RBD antibody clone.

Clonal evolution of RBD-binding memory B cells from ten convalescent individuals, a, Phylogenetic tree graph shows clones from convalescent non-vaccinees, b, Same as a, but from convalescent vaccinees. Numbers refer to mutations compared to the preceding vertical node. Colours indicate time point; grey, orange and red represent 1.3, 6 and 12 months respectively, black dots indicate inferred nodes, and size is proportional to sequence copy number; GL = germline sequence.

Extended Data Fig. 7. WT RBD binding and pseudovirus neutralization.

a, b, Binding curves (a) and EC₅₀ dot plot (b) of mAbs isolated from non-vaccinated (black curves and dots) and from vaccinated (blue curves and dots) convalescents individuals 12 months after infection (P = 0.74). c, Avidity (dissociation rate) measuring plasma reactivity to RBD at the 1.3- and 12 month follow-up visit (n = 33). d–f, IC₅₀ values of mAbs isolated 12 months after infection from non-vaccinated and vaccinated individuals; all 12 month antibodies irrespective of clonality (d), singlets only (e), and only antibodies belonging to a clone or shared over time (f). Statistical significance in b, d–f was determined using the two-tailed Mann–Whitney test; two-tailed Wilcoxon test (c). The geometric mean EC₅₀ and IC₅₀ are indicated in red. g, Heat map shows the neutralizing activity of clonally related antibodies against wt-SARS-CoV-2 over time. White tiles indicate no clonal relative at the respective time point. Clones are ranked from left to right by the potency of the 12 month progeny antibodies which are denoted below the tiles. h, IC₅₀ values of shared clones of mAbs cloned from B cells from the initial 1.3- and 6.2-, as well as 12 month follow-up visit, divided by participant, as indicated. Lines connect clonal antibodies shared between time points. Antibodies with IC₅₀ > 1,000 ng/ml are plotted at 1,000 ng/ml in d–h. i, IC₅₀ values of 5 neutralizing antibody pairs against indicated authentic SARS-CoV-2 WA1/2020 and B.1.351 viruses (n = 10). Average EC₅₀ and IC₅₀ values of two independent experiments are shown.

Extended Data Fig. 8. Biolayer interferometry affinity measurements.

a, b, Graphs depict affinity measurements of neutralizing (green) and non-neutralizing (red) antibodies isolated 1.3 months (a) or 12 months (b) after infection.

Extended Data Fig. 9. Biolayer interferometry antibody competition experiment.

a, b, Anti-SARS-CoV-2 RBD antibodies isolated 1.3 (a) or 12 months (b) after infection were assayed for competition with structurally characterized anti-RBD antibodies by biolayer interferometry experiments as in Fig. 4a. Graphs represent the binding of the second antibody (2nd Ab) to preformed first antibody (1st Ab)–RBD complexes. Dotted line denotes when 1st Ab and 2nd Ab are the same. For each antibody group identified in Fig. 4c the left graphs represent the binding of the class-representative C144, C121, C135 or C105^, (2nd Ab) to the candidate antibody (1st Ab)-RBD complex. The right graphs represent the binding of the candidate antibody (2nd Ab) to the complex of C144-RBD, C121-RBD, C135-RBD or C105-RBD (1st Ab). Antibodies belonging to the same groups are indicated to the left of the respective curves.