Maintenance of broad neutralizing antibodies and memory B cells 1 year post-infection is predicted by SARS-CoV-2-specific CD4+ T cell responses

Abstract

Understanding the long-term maintenance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) immunity is critical for predicting protection against reinfection. In an age- and gender-matched cohort of 24 participants, the association of disease severity and early immune responses on the maintenance of humoral immunity 12 months post-infection is examined. All severely affected participants maintain a stable subset of SARS-CoV-2 receptor-binding domain (RBD)-specific memory B cells (MBCs) and good neutralizing antibody breadth against the majority of the variants of concern, including the Delta variant. Modeling these immune responses against vaccine efficacy data indicate a 45%-76% protection against symptomatic infection (variant dependent). Overall, these findings indicate durable humoral responses in most participants after infection, reasonable protection against reinfection, and implicate baseline antigen-specific CD4+ T cell responses as a predictor of maintenance of antibody neutralization breadth and RBD-specific MBC levels at 12 months post-infection.

Keywords: 12 months after SARS-CoV-2 infection; Delta variant; RBD-specific memory B cells longevity; SARS-CoV-2; SARS-CoV-2 OX40 CD4 T cell assay; SARS-CoV-2 antibody duration; SARS-CoV-2 neutralization; protection efficacy prediction; variants of concern.

Graphical abstract

Figure 1

SARS-CoV-2 antibody analysis (A–C) Longitudinal (A) anti-RBD IgG antibody titer calculated from technical duplicates, (B) ADCP activity, and (C) neutralization ID₅₀ trend for D614G variant at visits 1, 2, and 3 calculated from technical duplicates. (D–F) Impact of disease severity at visit 3 on (D) anti-RBD IgG antibody titer, (E) ADCP activity, and (F) neutralization ID₅₀ for D614G variant. (G) Neutralization ID₅₀ for all variants of concern calculated from technical duplicates across three visits. (H) Breadth of neutralization across three visits stratified by disease severity. (I and J) Estimated protective efficacy of convalescent serum against symptomatic reinfection based on (I) anti-RBD IgG antibody levels at visit 3 (dotted line indicates the estimated 60% vaccine efficacy against symptomatic infection) and (J) neutralization titers at visit 3 as a proportion of early convalescent titers plotted on a logistic model developed for comparing vaccine efficacy studies (purple line indicates best fit for the logistics model, and pink shading indicates the 95% prediction interval) (Mean and SD indicated; ^∗∗∗∗p < 0.0001; ^∗∗∗p between 0.0001 and 0.001; ^∗∗p between 0.001 and 0.01; ^∗p between 0.01 and 0.05). Also see Figures S1 and S2.

Figure 2

Class-switched RBD-specific MBC analysis (A and B) Longitudinal (A) classical MBC levels (dotted line indicates the healthy control threshold) and (B) trends of classical MBC levels across three visits stratified by disease severity. (C) Impact of disease severity on classical MBCs at visit 3. (D) Number of class-switched RBD-specific MBCs per million B cells in each subset per visit. (E) Percentages of MBC subsets within the class-switched RBD-specific MBC group across three visits stratified by disease severity (mean + SD of participants). AM, activated memory; RM, resting memory; IM, intermediate memory; TLM, tissue-like memory (^∗∗p between 0.001 and 0.01;^∗p between 0.01 and 0.05). Also see Figures S1–S3 and Table S1.

Figure 3

T cell subset analysis (A) Antigen-specific OX40+ CD4+ T cells in the 24 participants at visit 1. (B) Percentage of OX40+ CXCR5+ cTfh cells of total CD4+ T cells examined by disease severity and antigen: Flu, Spike (SN), RBD (SR), and nucleocapsid protein (NP). (C) Proportion of cTfh Th1 (CXCR3+ CCR6–), Th2(CXCR3– CCR6–), Th17 (CXCR3– CCR6+), and Th1/Th17 (CXCR3+ CCR6+) subsets by antigen. (D) Spearman's correlation matrix comparing immune cell frequencies. Boxed area indicates comparison with the visit 3 immune measurements. (E) Representative Spearman's correlation plots. (F) PCA comparing all immune components and certain clinical or immunological characteristics. (G–J) A loading plot indicates the impact of the different variables on the clustering. Column graph comparing the influence of (G) OX40 CD4+ T cells, (H) OX40+ CXCR5+ T cells, (I) cTfh Th2 CD4 T cells, and (J) age on maintenance versus decline of MBC levels at visit 3, significance was tested with a logistic regression (Mean and SD indicated; ^∗∗p between 0.001 and 0.01; ^∗p between 0.01 and 0.05). Also see Figures S2 and S3 and Table S1.