Aging and Viral Evolution Impair Immunity Against Dominant Pan-Coronavirus-Reactive T Cell Epitope

Abstract

Immune evasion by escape mutations subverts immunity against SARS-CoV-2. A role of pan-coronavirus immunity for more durable protection is being discussed, but has remained understudied. We here investigated the effects of age, mutations, and homo-/heterologous vaccination regimens on the dominant pan-coronavirus-specific cellular and humoral epitope iCope after SARS-CoV-2 infection and vaccination in detail. In older individuals, the quantitatively and qualitatively reduced iCope-reactive CD4⁺ T cell responses with narrow TCR repertoires could not be enhanced by vaccination and were further compromised by emerging spike mutations. In contrast, pan-coronavirus-reactive humoral immunity was affected only by mutations and not by age. Our results reveal a distinct deficiency of the dichotomous layer of pan-coronavirus immunity in the older, critical for long-term protection against SARS-CoV-2 variants.

Keywords: SARS‐CoV‐2; T cells; aging; cross reactivity; pan‐coronavirus.

FIGURE 1

iCope mutation rate and sequence homology. (A) Mutations in 10.8 mio spike sequences per amino acid (GISAID, May 15, 2022). Amino acids 816–830 (highlighted in red) compared with other amino acids within the spike, including RBD binding domain covering aa 333–526 (highlighted in blue). (B) Numbers of 1aa shifting 15mers carrying a mutation throughout spike (GISAID, May 15, 2022). 15mers covering the iCope sequence are indicated in red. (C) Motif analysis of amino acid conservation of SARS‐CoV‐2 S816‐830 in endemic coronaviruses (229E, NL63, OC43, HKU1). (D) Number of sequences with indicated mutations that were reported in GISAID by September 2021 and additionally generated nonexisting mutations. Those replacing the most conserved amino acids are highlighted in red.

FIGURE 2

Mutations in S819‐826 impair iCope‐specific T cell responsiveness in unexposed and convalescents. Ex vivo stimulation of PBMCs from unexposed individuals (n = 17) and COVID‐19 convalescents (n = 8) with iCope WT or different mutated iCope peptides and control pools S‐I and S‐II, and CEFX as positive control. (A, C) The percentage of CD40L⁺4‐1BB⁺ CD4⁺ T cells among stimulated PBMC was divided by the percentage of these cells among unstimulated PBMC to determine the stimulation index (stim. index) shown on the y‐axis. Bars show the median. Dotted lines indicate a stim. index of 1.5 and 3. All values below 0.1 were set at 0.1 for display. (B, D) Bar plots show the proportions of individuals with the indicated peptide or peptide pool stimulations with a stim. index >3 (black), 1.5‐3 (grey) or <1.5 (white). (E) MHCII peptide binding was predicted for all combinations of the iCope/mutations peptides for all 96 donors from Table S1 using IEDB. The median of the best 20 prediction ranks was calculated for each allele and plotted against the stimulation index from the stimulations with iCope or its mutant peptides. Color codes indicate iCope wildtype peptide (red), mutated iCope peptides with no to weak negative effect on T cell activation (blue), and mutated peptides with strong negative effect on T cell activation (grey). Those with IEDB‐based, predicted negative effects are separated from the others by light grey dots. * indicates documented iCope mutations.

FIGURE 3

iCope responsiveness in vaccinated donors is age dependent. Ex vivo stimulation of PBMCs from young or older BNT/BNT (n = 18/19) vaccinated individuals with iCope WT or different mutated iCope peptides and control pools S‐I, S‐II, and CEFX. * indicates documented iCope mutations. (A, C, D) The percentage of CD40L⁺4‐1BB⁺ CD4⁺ T cells among stimulated PBMC was divided by the percentage of these cells among unstimulated PBMC to determine the stimulation index (stim. index) shown on the y‐axis. Bars show the median. Dotted lines indicate a stim. index of 1.5 and 3. All values below 0.1 were set at 0.1 for display. (B) Bar plots show the proportions of individuals with the indicated peptide or peptide pool stimulations with a stim. index >3, 1.5‐3 or <1.5. (C) Comparison of young and older BNT/BNT vaccinated donors under the indicated stimulation conditions. (D) Comparison of young BNT/BNT with young AZ/BNT vaccinated donors and older BNT/BNT with older AZ/AZ vaccinated donors under the indicated stimulation conditions. *p < 0.05, **p < 0.01, ***p < 0.001 and ns for p < 0.05 (Student's t‐test). BNT: BNT162b2 mRNA COVID‐19 vaccine (BioNTech); AZ: ChAdOx1 COVID‐19 vaccine (AstraZeneca).

FIGURE 4

The quality of iCope responsiveness is impaired in older individuals. Ex vivo stimulation of PBMCs from unexposed (n = 17), convalescent (n = 8), young BNT/BNT vaccinated (n = 18), and older BNT/BNT vaccinated (n = 19) individuals with iCope WT or different mutated iCope peptides and the control pools S‐I, S‐II, and CEFX. * indicates documented iCope mutations. (A, B) Frequencies of CD3^lo cells in CD40L⁺4‐1BB⁺ CD4⁺ T cells. CD3^lo frequencies are shown for T cell responses with a stim. index ≥1.5. Bars show the median. (C) Proportion of IFN‐γ and/or TNF‐α producing T cells among CD40L⁺4‐1BB⁺ CD4⁺ T cells in the indicated cohorts after stimulation with iCope WT, S‐I, or S‐II. (D) Optical density (OD) of anti‐S809‐826 wildtype (WT) peptide IgG (ELISA) in unexposed (n = 17), convalescents (n = 8), BNT/BNT young (n = 18), and older BNT/BNT (n = 19). Bars show the median. (E) Levels of anti‐S1‐ or anti‐S2‐IgG binding antibody intensity units in indicated cohorts. Dotted lines indicate lower cut‐off (set at 11) for definitive negative values and a range from 11–18 for values classified as positive with uncertainty. Bars show the median. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns for p < 0.05 (Student's t‐test).

FIGURE 5

Older people possess functional iCope‐reactive T cells but with limited clonal breadth. (A, B) Ex vivo stimulation of PBMCs from young (<40 years, blue, n = 507) or older (>70 years, red, n = 1267) individuals with S‐II (A) or iCope (B) prior to first vaccination (preVax) or at indicated times after the first, second, or third dose of vaccination (V1‐3). The percentage of CD40L⁺4‐1BB⁺ CD4⁺ T cells among stimulated PBMC was divided by the percentage of these cells among unstimulated PBMC to determine the stimulation index (stim. index) shown on the y‐axis. (C, D) Single‐cell gene expression of FACS‐purified iCope reactive CD40L⁺4‐1BB⁺ and CD40L⁻4‐1BB⁻ CD4⁺ T cells of n = 9 young (<43 years) and n = 8 older (>76 years) donors. tSNE visualization of CD40L⁺4‐1BB⁺ (C) and CD40L⁻4‐1BB⁻ (D) CD4⁺ T cell clustering of young (blue dots) and older (red dots) donors. (E) D50 diversity index indicating the number of clonotypes occupying the top 50% of the TCR repertoire. (F) Proportion of clonotypes with specific counts in indicated samples of the different donors (D1‐6). D1‐3 are the triple vaccinated young donors without, D4‐9 with recent infection, D10‐12 triple vaccinated older donors without, 13–17 with recent infection. (G) Frequencies of overlapping clonotypes between activated CD40L⁺4‐1BB⁺ and inactivated CD40L⁻4‐1BB⁻ CD4⁺ T cells in young and older donors. Bars show the median. (H) Heatmap of gene expression signatures of curated gene sets (as given in the Material and Methods section) for naïve and effector T cell characteristics as well as cytotoxicity, anergy, exhaustion and senescence in overlapping clonotypes between activated CD40L⁺4‐1BB⁺ (+/+) and inactivated CD40L⁻4‐1BB⁻ (−/−) CD4⁺ T cells. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns for p < 0.05 (Student's t test).