SARS-CoV-2 infected children form early immune memory responses dominated by nucleocapsid-specific CD8+ T cells and antibodies

Abstract

This is the third year of the SARS-CoV-2 pandemic, and yet most children remain unvaccinated. COVID-19 in children manifests as mostly mild or asymptomatic, however high viral titers and strong cellular and humoral responses are observed upon acute infection. It is still unclear how long these responses persist, and if they can protect from re-infection and/or disease severity. Here, we analyzed immune memory responses in a cohort of children and adults with COVID-19. Important differences between children and adults are evident in kinetics and profile of memory responses. Children develop early N-specific cytotoxic T cell responses, that rapidly expand and dominate their immune memory to the virus. Children's anti-N, but not anti-S, antibody titers increase over time. Neutralization titers correlate with N-specific antibodies and CD8⁺T cells. However, antibodies generated by infection do not efficiently cross-neutralize variants Gamma or Delta. Our results indicate that mechanisms that protect from disease severity are possibly different from those that protect from reinfection, bringing novel insights for pediatric vaccine design. They also underline the importance of vaccination in children, who remain at risk for COVID-19 despite having been previously infected.

Keywords: COVID-19; Memory T cell; N protein; antibodies; neutralizing antibodies; variants of concern.

Figure  1

Study design and longitudinal Analysis of Antibodies. (A) Intervals of second and third blood draw from cohort patients, followed by cryopreservation and analysis. (B, C) Plasma IgG antibody binding to SARS-CoV-2 RBD and N protein, respectively, determined by enzyme-linked immunosorbent assay (ELISA), (n=70). Longitudinal analysis through connecting lines, color-coded, represented as area under the curve (AUC). Dotted lines in black highlight median values. Individuals colored in black were vaccinated between the second and the third collection point (Mild with Astrazeneca and Severe with Sinovac). (D, E) Median titers plotted over time. Differences among group profiles were analyzed using Generalized Estimating Equations (GEE) and sequential Sidák’s adjustment for pairwise multiple comparisons, colored asterisk (*) indicate statistical difference compared to previous point of collection. (F, G) Individual AUC 6 months after acute infection, in the three groups of patients; The horizontal bar is the median, P values are displayed over brackets. Graphical study design was created with biorender.com.

Figure 2

Longitudinal analysis of neutralizing antibody titers by PRNT. (A) Individual PRNT50 titers over time. (B) Comparison of mean neutralization curves at each collection point, per group. (C) Antibody potency analysis between groups through analysis of ratio between S RBD antibody/PRNT 50 antibody neutralization. Differences between patients’ groups at the same collection point and differences between collection points were assessed by Kruskall Wallis’ and Dunn’s post hoc test. ns, non-significant.

Figure 3

Generation dynamics of specific memory T cell responses at 2^nd and 3^rd collection points. An unbiased tSNE analysis was performed for analysis of memory T cells populations, stimulated with peptide pools of the S, N, and M proteins of SARS-CoV-2, gated from live cells. Clusters were characterized by the expression of surface markers CD4, CD8, CD45RA, and CCR7. (A) Clusters at the second time point, from a total of 36 samples (8 children, 17 mild, 11 severe); (B) Clusters at the third time point, from a total of 29 samples (7 children, 12 mild, 10 severe). Bar graphs show frequencies of events in each of the seven most frequent clusters identified in each time point. Color legends identify clusters as interpreted based on their surface markers.

Figure 4

Longitudinal analysis of functional specific memory T cells. (A) Individual frequencies of TNF-α production in memory T CD8⁺ cells collected at the 3 and 6-months post-acute infection, responding to in vitro challenge with different SARS-CoV-2 peptide pools to proteins S, N, or M, percentages of responders are indicated above columns. Differences between patients’ groups at the same collection point were assessed by Kruskal-Wallis and Dunn’s post hoc test; differences between collection points were assessed by Wilcoxon test. (B) Heatmap illustrating CD8⁺ and CD4⁺ responses of 8 individual children to the different SARS-CoV-2 peptide pools. Data is shown as z-score, where patients who did not respond to stimuli are considered in light gray. C – individual children. NR – Non-responsive. CM – Central memory; EM – Effector memory; EMRA – Terminally differentiated effector memory cells.

Figure 5

Spearman correlation analysis of antibody and N-protein specific TNF-α+ T cell responses 6-months after acute infection. The matrix represents the correlation analysis of N-specific effector T cells responses (in percentages of CD4⁺ and CD8⁺ cells expressing cytokine TNF-α in response to the peptide pools), neutralizing antibodies to the B original variant, as well as the antibody response to the N protein (represented as values for the AUC – area under the curve). Total of 32 samples (children = 9, mild= 16, severe = 10).

Figure 6

Impact of variant emergence on memory antibody recognition 6 months after acute infection. (A) Total individual IgG binding to the RBD compared to trimeric B and variants spike protein, by enzyme-linked immunosorbent assay (ELISA); (B–D) PRNT analysis and fold reduction of recognition of neutralizing antibodies. Differences between patients’ groups were assessed by Kruskal Wallis’ and Dunn’s post hoc test.