key: cord-0698251-thxwoh9o
authors: Paul, K.; Sibbertsen, F.; Weiskopf, D.; Luetgehetmann, M.; Barroso, M.; Danecka, M. K.; Glau, L.; Hecher, L.; Hermann, K.; Kohl, A.; Oh, J.; Schulze zur Wiesch, J.; Sette, A.; Tolosa, E.; Vettorazzi, E.; Woidy, M.; Zapf, A.; Zazara, D. E.; Mir, T. S.; Muntau, A. C.; Gersting, S. W.; Dunay, G. A.
title: Evaluation of the CD4+ T cell response to SARS-CoV-2 infection and cross reactivity to beta variant in children of all ages
date: 2022-01-28
journal: nan
DOI: 10.1101/2022.01.27.22269976
sha: a7d2e9189b6f513bc6864f2ab17bdfb8670b1efb
doc_id: 698251
cord_uid: thxwoh9o

SARS-CoV-2 is still a major burden for global health despite effective vaccines. With the reduction of social distancing measures, infection rates are increasing in children, while data on the pediatric immune response to SARS-CoV-2 infection is still lacking. Although the typ-ical disease course in children has been mild, emerging variants may present new challenges in this age group. Peripheral blood mononuclear cells (PBMC) from 51 convalescent children, 24 seronegative siblings from early 2020, and 51 unexposed controls were stimulated with SARS-CoV-2-derived peptide MegaPools from the ancestral and beta variants. Flow cytometric determina-tion of activation-induced markers and secreted cytokines were used to quantify the CD4+ T cell response. The average time after infection was over 80 days. CD4+ T cell responses were detected in 61% of convalescent children and were markedly reduced in preschool children. Cross-reactive T cells for the SARS-CoV-2 beta variant were identified in 45% of cases after infec-tion with an ancestral SARS-CoV-2 variant. The CD4+ T cell response was accompanied most predominantly by IFN-g; and Granzyme B secretion. An antiviral CD4+ T cell response was present in children after ancestral SARS-CoV-2 infec-tion, which was reduced in the youngest age group. We detected significant cross-reactivity of CD4+ T cell responses to the more recently evolved immune-escaping beta variant. Our find-ings have epidemiologic relevance for children regarding novel viral variants of concern and vaccination efforts.

The SARS-CoV-2 virus appeared in 2019 1 causing a pandemic which, despite effective vac-55 cination, is still a major threat to global health 2 . Children in particular are now facing increasing 56 infection rates due to a reduction of social distancing measures while vaccination of younger 57 age groups has just begun or is not yet available. So far, the disease course in the younger 58 population appears to be mild 3,4 , however emerging variants may present new challenges. With 59 the current emerging omicron variant hospitalization rates of young children, in particular, ap-60 pear to be on the rise 5 . Information on the pediatric immune response after infection or vac-61 cination is of great importance for planning protective strategies in the future. However, data 62 on T cell-mediated immunity in children is still lacking. The identification of antigen-specific 63 T cells via stimulation of patient PBMC with viral peptide pools followed by detection of reac-64 tive T cells through activation-induced markers allows the identification and simultaneous phe-65 notyping of these cells using limited available patient material 6 and has been broadly used to 66 identify SARS-CoV-2 reactive T cells in adults 7-14 , and children 15, 16 . The COVID-19 Child 67

Health Investigation of Latent Disease (C19.CHILD) Hamburg Study recruited children from 68 all age groups after the spring 2020 wave of SARS-CoV-2 infections in Hamburg, Germany. 69

Here, PBMC from over fifty SARS-CoV-2 convalescent children, their exposed siblings and 70 unexposed age-matched controls from the C19.CHILD cohort were stimulated with peptide 71 MegaPools (MP) spanning the entire SARS-CoV-2 Spike Glycoprotein of the Wuhan-Hu-1 72 strain and beta variant as well as predicted peptides representing the remaining entire SARS-73

CoV-2 Wuhan-Hu-1 strain proteome 17 to detect and characterize virus-specific CD4+ T cell 74 responses. 75 76 Methods 77 SARS-CoV-2 convalescent children, exposed seronegative siblings as well as unexposed con-79 trols were identified from the COVID-19 Child Health Investigation of Latent Disease 80 (C19.CHILD) Hamburg Study cohort, registered at clinicaltrials.gov (NCT04534608). Briefly, 81 6113 children (<18 years) who presented voluntarily or were recruited while receiving care in 82 one of the five pediatric hospitals of Hamburg, Germany, were invited for a screening for an 83 acute or recent SARS-CoV-2 infection via PCR and serum antibody testing. 84

Patients who tested positive in the PCR and/or the antibody screening were invited with all 85 household members for a follow-up appointment, where detailed history was obtained and PCR 86 and serologic SARS-CoV-2 testing were repeated. PBMC samples were obtained from all fam-87 ily members under 18 years. Pediatric unexposed and healthy volunteers, with no known SARS-88

CoV-2 contact, were welcomed to enrol through the C19.CHILD Study Clinic. 89

Recruiting was conducted from May 11 th until June 30 th 2020 after the first infection wave, 90 during and after the first lockdown in Germany. Parents or legal guardians provided written 91 informed consent in all cases. From children over 7 years, consent in writing was obtained 92 whenever possible but also consent in spoken word was accepted. The study was approved by 93 the local ethical committee of Hamburg (reference number: PV7336). 94

For screening purposes, serum samples were tested for SARS-CoV-2 specific antibodies di-96 rected against the viral nucleocapsid (IgA/IgM/IgG) using Elecsys® Anti-SARS-CoV-2 Ig as-97 say (Roche) on the cobas e411 system (Roche). Additionally, serum samples were tested for 98 SARS-CoV-2 specific antibodies against the S1 and S2 subunits of the viral Spike protein using 99 LIAISON® SARS-CoV-2 IgG serology test (DiaSorin). 100

To evaluate serostatus for "common cold" coronaviruses (HCoV) and to further confirm SARS-7 Thawed PBMC were incubated in 5ml RPMI + human serum (Pan Biotech) 5% + Benzonase 128 (Sigma-Aldrich) 50 U/ml for 1 hour (37°C 5% CO2), followed by a washing step with 15ml 129 RPMI + human serum (HS) 5%. Afterwards all available cells were equally divided to be stim-130 ulated for 24 hours (37°C 5% CO2) in 200µl RPMI + HS 5% in 96-well U-bottom plates with 131 mentioned peptide MegaPools (1µg/ml/peptide), PHA-L (Invitrogen) (1µg/ml) as positive con-132 trol and an equimolar amount of DMSO to serve as negative control. After 24h of stimulation 133 cell culture supernatant was carefully removed and stored at -20°C for later multiplex cytokine 134 analysis. Incubation was stopped by washing cells in PBS. Expression of activation-induced 135 markers (CD69 and OX40) in response to specific peptide stimulation, as well as their memory 136 phenotype were measured by flow cytometry (flow cytometry antibodies are listed in Supple-137 mentary Detection of cytokines in the cell culture supernatant of stimulated cells was performed using 162

LEGENDplex™ Human CD8/NK Panel (13-plex, BioLegend) suitable for detection of IL-2, 163 IL-4, IL-10, IL-6, IL-17A, TNF-α, sFas, sFasL, IFN-γ, Granzyme A, Granzyme B, Perforin, 164

Granulysin, according to the manufacturer's instructions. Briefly, freshly prepared provided 165 cytokine standard or thawed cell culture supernatant was mixed with cytokine-specific beads, 166 incubated for 2 hours and washed. After sequential incubation of bead-bound cytokines with 167 biotin-labeled cytokine detection antibodies and streptavidin PE antibodies, non-binding anti-168 bodies were washed off and PE-labeled bead-bound cytokines were subsequently analyzed by 169 flow cytometry. Quantification of cytokines was carried out using the standard. The data was 170 analyzed using the online LEGENDplex™ Data Analysis Software of the manufacturer. The 171 assay was performed in duplicates and mean values of each sample were used for further anal-172 ysis. 173

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint Data analysis, graphs and statistics were prepared with FlowJo version 10, and R 4.0.5 (pack-175 ages: tidyverse, rstatix, splines, emmeans, kableExtra, magrittr, heatmaply). Paired sample 176 analyses were performed by paired t-test. The stimulation index and fold increases showed 177 highly skewed distributions and were therefore log-transformed for further analysis. An un-178 paired t-test was used for comparing two groups. Categorical variables were compared using 179

Fisher's exact test. Comparisons between three groups were done with one-way ANOVA and 180 post hoc pairwise t-tests, if the ANOVA-F test was significant, thus following the closed test 181 principle, no adjustment for multiple testing was necessary for pairwise comparisons. The as-182 sociation between age and the stimulation index was explored using a non-parametric spline 183 regression with age and serology group and sex as independent variables. An interaction be-184 tween spline age and serology group was initially included in the model, but it was removed 185 from the final model, if it did not significantly increase the model fit. A p-value < 0.05 was 186 considered as statistically significant in all analyses. As this study had an exploratory nature, 187 we refrained from adjusting for multiple testing. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint

The study cohort consisted of 126 participants: 51 seropositive children as determined by pos-200 itivity in all of three separate serological tests covering the viral Spike and Nucleocapsid, 24 201 seronegative siblings living in a shared household with an infected individual and 51 age-and 202 gender-matched unexposed controls. There were no significant differences in the age-and gen-203 der distribution of these groups (Table 1) . For children whose families were able to give a de- perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in

The copyright holder for this this version posted January 28, 2022.

PBMC from all donors were stimulated with peptide MegaPools derived from the first de-223 scribed original SARS-CoV-2 strain (Wuhan-Hu-1) spanning the entire Spike glycoprotein 224 (Spike-OS_MP) and predicted epitopes from the remaining proteome (R_MP). CD4+ T cell 225 response was measured based on the expression of AIM markers (OX40 and CD69) and com-226 pared to a negative control consisting of the peptide mix solvent DMSO at the same concentra-227 tion as in the peptide mix ( perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint Additionally, reactive T cells were detected in 8 % (2 of 24 Spike-OS_MP) and 13 % (3 of 23 246 R-MP) of seronegative siblings, as well as in 14 % (7 of 51 Spike-OS_MP) and 12 % (6 of 51 247 R_MP) of unexposed controls. 248 Without detectable differences in T cell response between seronegative siblings and unexposed 249 controls, our data indicate that the intensity of exposure to SARS-CoV-2 in household mem-250 bers, which did not lead to a humoral immune response (seronegative siblings) was generally 251 also not sufficient to induce a systemic CD4+ T cell response. 252

Naive versus memory phenotype of AIM+ CD4+ T cells was determined using CD27 and 253 We evaluated the T cell response towards the beta variant, a WHO variant of concern due to its 266 capability to escape humoral immunity 24 perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint SI > 3 in any of the peptide pools had a positive HCoV serology. This indicates that the identi-295 fied SARS-CoV-2 reactive T cells in seronegative siblings and unexposed controls cannot be 296 explained only by cross reactive T cell memory from prior HCoV infections. 297

Cytokine profile of SARS-CoV-2-reactive T cells 298

To further characterize the T cell response, thirteen different cytokines were measured in cul-299 ture supernatants after exposure of PBMC to SARS-CoV-2 peptide MegaPools. We could de-300 tect increased cytokine levels for nearly all of the quantified cytokines after peptide stimulation 301 when compared to DMSO negative control (data not shown). To control for unspecific or by-302 stander immune activation and to be able to attribute cytokine secretion to a specific T cell 303 response, we calculated fold increases in cytokine concentration after peptide stimulation over 304 cytokine concentration in corresponding DMSO negative controls analogous to the stimulation 305 index for T cell response. These were compared between seropositive children with a clear T 306 cell response (SI > 3 as quantified using activation induced markers) and unexposed controls 307 and seropositive children both without a clear T cell response (SI < 3). 308

The analysis revealed levels of IFN-γ and Granzyme B, both associated with viral defense, were 309 consistently higher in seropositive T cell responders after Spike-OS_MP, R_MP and Spike-310 BV_MP stimulation compared seropositives and unexposed controls both without a T cell re-311 sponse (Figure 3 A -F). Since PBMC were stimulated in bulk, the elevated IFN-γ and espe-312 cially Granzyme B levels could indicate concomitant CD8+ activation. 313

In seropositive T cell responders, R_MP stimulation led to an increased secretion of IL-2 (Fig-314 ure 3 G) and Granzyme A (Figure 3 H) seronegative siblings as well as unexposed controls. We provided further evidence that a high 335 proportion of children who seroconverted to SARS-CoV-2 were able to mount specific CD4+ 336 T cell responses, that were still detectable up to over 100 days after infection. In contrast to 337 adult cohorts 31 we could identify specific T cell responses in only a minority of SARS-CoV-2 338 exposed but not seroconverted children (seronegative siblings). In our pediatric cohort we 339 showed a T cell response attributable to a prior SARS-CoV-2 infection in 61% of seropositive 340 children, which is less than the reported response rates of 77 - perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in Until now data on pediatric T cell response after SARS-CoV-2 infection is scarce with cohort 345 sizes not allowing for analysis over different age groups 15,16 . Importantly, we showed an in-346 crease in SARS-CoV-2-specific CD4+ T cell responses by age within this pediatric cohort. 347

Reduced capacity to mount specific T cell responses was particularly seen in preschool children. and SARS-CoV-2 was demonstrated 37 . We analyzed the influence of prior HCoV infections 369 on pediatric SARS-CoV-2 T cell response. In our cohort, we could not detect a difference in 370 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. CoV-2 Wuhan-Hu-1 strain. Based on our data generated by stimulation of archived samples 396 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. Here we showed that pediatric CD4+ T cell responses after infection with an ancestral SARS-417

CoV-2 variant are age dependent, with reduced capability of the youngest to mount specific 418 responses. Antigen specific T cells persist over three months after infection and are cross reac-419 tive with the SARS-CoV-2 variant of concern B.1.351-beta variant. We detected a strong anti-420 viral cytokine response in association with SARS-CoV-2-specific T cell activation. Our 421 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint The respective peptide pool is indicated on top of each graph. The effect of age on the magnitude of T cell response, was further analyzed with a non-parametric multivariate regression analysis, by using a spline model (blue lines with light blue areas indicating 95% confidence intervals).

F: T cell responses towards Spike -OS_MP stimulation were compared between participants with different serostatus for "common cold" coronaviruses (HCoV, strains 229E, NL63, OC43 and HKU1). Analyses were conducted within study groups, which were defined by SARS-CoV-2 exposure and serostatus. Unpaired t -test was used to quantify P values. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns -not significant All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in Figure 3 The concentration of 13 cytokines was determined in cell culture supernatants after stimulation with SARS-CoV-2-derived peptide MegaPools. The respective peptide pool, used for stimulation, is indicated on top of each graph. Comparison was conducted between seropositive children with and without a clear T cell response and unexposed controls without a T cell response. T cell response was defined as a stimulation index (SI) > 3 in the AIM assay upon peptide stimulation A -I: Comparison of IFN-γ, Granzyme B, IL-2, Granzyme A and IL-10 secretion between groups. To adjust for unspecific cytokine secretion, comparison was performed by using the fold increase in cytokine concentration after peptide stimulation over cytokine concentration in DMSO treated samples. One way ANOVA and post hoc pairwise t -tests were used to quantify P values. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, not significant -not displayed perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint Supplementary Figure 3 . T cell response to peptide stimulation compared between different HCoV serostatus within study groups Comparison of T cell response towards R_MP and Spike -BV_MP stimulation between their different serostatus for "common cold" coronaviruses (HCoV). Analyses were conducted within study groups, which were defined by SARS-CoV-2 exposure and serostatus. Unpaired t -test was used to quantify P values. ns -not significant.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted January 28, 2022. ; https://doi.org/10.1101/2022.01.27.22269976 doi: medRxiv preprint

Approach Can Predict Candidate Targets for Immune 496 Responses to SARS-CoV-2

Phenotypic analysis of 498 the pediatric immune response to SARS-CoV-2 by flow cytometry

SARS-CoV-2-specific T cells in infection and 501 vaccination

Safety and immunogenicity of 503 an inactivated SARS-CoV-2 vaccine, BBIBP-CorV: a randomised, double-blind, placebo-504 controlled, phase 1/2 trial

Inactivated Vaccine Candidate, BBIBP-CorV, with Potent Protection against SARS-CoV-507 2

Phase I/II 509 study of COVID-19 RNA vaccine BNT162b1 in adults

SARS-CoV-2 mRNA vaccine design enabled by prototype pathogen preparedness. 512