Distinct T cell polyfunctional profile in SARS-CoV-2 seronegative children associated with endemic human coronavirus cross-reactivity

Summary SARS-CoV-2 infection in children typically results in asymptomatic or mild disease. There is a paucity of studies on SARS-CoV-2 antiviral immunity in African children. We investigated SARS-CoV-2-specific T cell responses in 71 unvaccinated asymptomatic South African children who were seropositive or seronegative for SARS-CoV-2. SARS-CoV-2-specific CD4+ T cell responses were detectable in 83% of seropositive and 60% of seronegative children. Although the magnitude of the CD4+ T cell response did not differ significantly between the two groups, their functional profiles were distinct, with SARS-CoV-2 seropositive children exhibiting a higher proportion of polyfunctional T cells compared to their seronegative counterparts. The frequency of SARS-CoV-2-specific CD4+ T cells in seronegative children was associated with the endemic human coronavirus (HCoV) HKU1 IgG response. Overall, the presence of SARS-CoV-2-responding T cells in seronegative children may result from cross-reactivity to endemic coronaviruses and could contribute to the relative protection from disease observed in SARS-CoV-2-infected children.


Seronegative children exhibit a lower proportion of polyfunctional T cells
CoV-2-specific T cell and endemic HKU1 IgG responses correlate in seronegative children ll OPEN

ACCESS
In contrast to the low COVID-19 severity in the majority of younger individuals, it is known that children are more susceptible than adults to other acute viral respiratory tract infections including respiratory syncytial virus (RSV), rhinovirus (RV), influenza virus and common circulating endemic human coronaviruses (HCoV). 73][24][25][26] Endemic HCoVs account for 15 to 30% of respiratory infections reported annually in children. 27,28These HCoVs belong to the alpha-coronavirus subfamily (HCoV-229E and HCoV-NL63) and the beta-coronavirus subfamily (HCoV-OC43 and HCoV-HKU1), are generally considered to have seasonal infection peaks during the winter season and are responsible for high rates of infection among children. 29,30In contrast, the influenza surveillance data from India showed that HCoV-HKU1 peaked during spring to summer over the 2015-2016 period. 31t is well established that protective immune responses to SARS-CoV-2 encompass both an antibody and T cell component. 324][35][36] However, data on T cell responses in children compared to adults are conflicting, with studies reporting lower SARS-CoV-2-specific T cell responses in children, 23,37,38 no differences 39 or greater T cell responses in children. 22Many of these studies make use of ELISPOT assays which preclude analysis of T cell phenotypes or functional cytokine profiles.
In this study, we prospectively characterized specific T cell responses in SARS-CoV-2 seropositive and seronegative children during the COVID-19 pandemic and determined the functional profiles of SARS-CoV-2-specific T cells, in the context of their pre-existing immunity against endemic beta-HCoVs.This study provides insights into cross-reactive immunity to SARS-CoV-2 in children.

Study cohort
To investigate SARS-CoV-2-specific immune responses, we measured immunoglobulin (Ig) G and T cell responses in 71 healthy children recruited in Cape Town, Western Cape, South Africa.The participants are described in Figure 1A.The median age of the children was 7 years (interquartile range (IQR) 2.8-9 years) and 34% (24/71) were female.The children included in this study had not received any SARS-CoV-2 vaccine prior to recruitment and no PCR-confirmed infection data were available, although 58% showed SARS-CoV-2 seropositivity (defined as being positive for either anti-spike or anti-nucleocapsid IgG).Samples were collected between 1 February 2021 and 20 May 2021, after two infection waves in South Africa, that were dominated by the ancestral D614G strain followed by the Beta variant of concern. 40

SARS-CoV-2-specific T cell responses in children
The magnitude of SARS-CoV-2-specific T cell responses in children was quantified using a whole blood assay and intracellular cytokine staining followed by flow cytometry.SARS-CoV-2-specific CD4 + and CD8 + T cell responses were measured as the total cytokine production of interferon-g (IFN-g), tumor necrosis factor-a (TNF-a) or interleukin-2 (IL-2) in response to a combined peptide pool covering SARS-CoV-2 spike (S), nucleocapsid (N) and membrane (M) proteins (Figure 2A).
We assessed the proportion of responders and magnitude of SARS-CoV-2-specific T cell responses according to the serostatus of the children (Figures 2B and 2C).As expected, 83% (34/41) of seropositive children had detectable SARS-CoV-2-specific CD4 + T cells.Interestingly, 60% (18/30) of seronegative children also had detectable SARS-CoV-2-specific CD4 + T cells, although this was a significantly lower proportion than observed in the seropositive group (p = 0.031).The magnitude of the SARS-CoV-2-specific CD4 + T cell response in seronegative responders was comparable to that observed in seropositive children (median of responders: 0.04% and 0.057% respectively; p = 0.437, Figure 2B).Conversely, we did not observe significant differences in the proportion of SARS-CoV-2 CD8 + T cell responders ($50%) or the frequency of SARS-CoV-2-specific CD8 + T cells between seropositive and seronegative children (median of responders: 0.019% and 0.025% respectively; p = 0.214, Figure 2C).
Next, we evaluated the polyfunctional profile of SARS-CoV-2-responding CD4 + and CD8 + T cells based on the co-production of IFN-g, IL-2 or TNF-a (Figures 2D and 2E).The overall functional profile of SARS-CoV-2-specific CD4 + T cells in seropositive children was distinct from that of seronegative children (p = 0.007, Figure 2D).SARS-CoV-2-specific CD4 + T cells in seropositive children were more polyfunctional, exhibiting a higher proportion of triple functional IFN-g+IL-2+TNF-a+-producing CD4 + T cells (p = 0.0044), and dual expressing cells (IL-2+TNF-a+, p = 0.0048; IFN-g+IL-2+, p % 0.0001) compared to the seronegative children.In contrast, seronegative children were characterized by an increased proportion of IFN-g monofunctional CD4 + T cells (p = 0.0048).Considering the CD8 compartment, we detected no significant differences in the overall functional profile of SARS-CoV-2-responding CD8 + T cells between the two groups (p = 0.098), although seropositive children did have a larger proportion of CD8 + T cells producing both IFN-g and IL-2 (p = 0.027) and single IFN-g (p = 0.011) (Figure 2E).
Taken together, these data show that a significant proportion of seronegative children have detectable SARS-CoV-2-specific CD4 + and CD8 + T cells.Notably, these CD4 + T cells exhibit a distinct polyfunctional cytokine profile compared to those found in SARS-CoV-2-exposed children.

Pre-existing immunity to endemic common circulating human coronaviruses
The presence of SARS-CoV-2-responding CD4 + T cells primarily exhibiting a monofunctional profile observed in seronegative children led us to hypothesize that the T cell response might be due to cross-reactivity resulting from prior infection with common circulating endemic HCoVs.Therefore, we measured endemic beta HCoV-HKU1 and HCoV-OC43 spike IgG in SARS-CoV-2 seropositive and seronegative children.The magnitude of both HCoV-HKU1 (Figure 3A) and HCoV-OC43 (Figure S2A) spike IgG was comparable in the seropositive and seronegative groups (median OD: 0.746 vs. 0.420; p = 0.153 for HCoV-HKU1 and 1.769 vs. 1.281; p = 0.07 for HCoV-OC43, respectively).

SARS-CoV-2-specific T cell responses in SARS-CoV-2 seropositive children compared to COVID-19 convalescent adults
Blood and plasma samples were collected from 30 COVID-19 convalescent healthcare workers (HCWs) participating in a longitudinal study at Groote Schuur Hospital (Cape Town, Western Cape, South Africa) (Figure 4A).All convalescent adults had a prior mild/asymptomatic PCRconfirmed SARS-CoV-2 infection (median 224 days prior to sampling) and had not received a COVID-19 vaccine at the time of sampling.Samples were collected between 22 January 2021 and 23 February 2021.

DISCUSSION
T cells have been associated with protection from severe COVID-19. 43 CD4 + and CD8 + T cells, but their CD4 + T cells exhibited a distinct functional cytokine profile compared to seropositive children.Importantly, in seronegative children, the frequency of SARS-CoV-2-reactive CD4 + T cells positively associated with HCoV-HKU1 spike-specific IgG antibodies.Additionally, we showed that convalescent adults had a higher magnitude of CD4 + T cell responses against SARS-CoV-2 compared to seropositive children, which associated positively with age.
Our data are in accordance with several studies showing that children develop robust humoral and cellular immunity to SARS-CoV-2 infection 20,22,39,44,45 and BNT162b2 vaccine. 46In addition, SARS-CoV-2-specific T cell responses have been detected in 40-60% of SARS-CoV-2 unexposed individuals including children and adults 22,23,[47][48][49][50][51][52] suggesting possible cross reactivity to HCoVs.One study demonstrated that the proportion of T cell responders was higher in children than adults. 22In contrast, a study by Tsang et al. found that SARS-CoV-2 uninfected children failed to mount detectable T cell responses. 44A further study showed that only a small proportion of seronegative children mounted a SARS-CoV-2-specific CD4 + T cell response (13%) compared to 60% in seropositive siblings, despite similar exposure in shared households. 45These two studies differ from our findings, where we showed that 60% of seronegative children had detectable SARS-CoV-2-specific T cell responses, consistent with Dowell et al. 22 The discrepancies between the studies could be due to a difference in the seasonal prevalence of circulating HCoVs in each study setting, 29,30,53 different T cell assays used to analyze T cell responses, and/or the cohort demographics.
Although the source of SARS-CoV-2 reactive T cells in seronegative children remains unclear, mounting evidence argues for cross-reactivity to endemic HCoVs.HCoVs (including beta HCoVs-HKU1 and OC43, and alpha HCoVs-NL63 and 229E) have partial sequence homology with SARS-CoV-2. 52,54While cross reactive humoral responses have been shown not to underly this protection, 55 studies have reported preexisting T cell responses to endemic HCoVs with cross-reactivity to SARS-CoV-2 in unexposed adults 50,[56][57][58][59][60][61] and children. 22,24,62,63These cellular responses were targeted mainly toward the S2 subunit which is highly conserved among coronaviruses. 22,25In our study, we showed a positive correlation between the frequency of SARS-CoV-2 CD4 + T cells and the magnitude of HKU1 spike-specific IgG in SARS-CoV-2 seronegative children.Our findings are supported by similar findings showing that 58% (7/12) seronegative children had binding antibodies against alpha and/or beta HCoV spike proteins, which they concluded were likely due to recent HCoV infection. 2261]64 Sagar et al. showed that individuals with a documented recent or ongoing HCoV infection and SARS-CoV-2 infection had fewer COVID-19-associated complications when hospitalized, compared to individuals without HCoV infection. 64Previously, the clinical relevance of pre-existing immunity and cross-reactive cellular responses from endemic HCoVs in providing protection from SARS-CoV-2 infection in children was unclear.Recently however, Dowell et al. demonstrated that SARS-CoV-2-reactive T cell responses in SARS-CoV-2 seronegative children were associated with relative protection against Omicron infection, suggesting a protective role in SARS-CoV-2 unexposed children. 657][68][69] These include protection against cytomegalovirus (CMV) infection after lung transplantation, 67 improved viral control of hepatitis C virus (HCV) 68 and detection during human immunodeficiency virus (HIV) infection in long term non-progressors. 69We found that seropositive children had a more polyfunctional profile of SARS-CoV-2-specific CD4 + T cells while CD4 responses in seronegative children exhibited a predominantly monofunctional profile.The SARS-CoV-2-specific polyfunctional Th1 CD4 T cell response (characterized by the co-expression of IFN-g, TNF-a and/or IL-2), as seen in seropositive children, may be necessary for effective viral control and has been documented in COVID-19 convalescent adults. 70,71he distinct CD4 and CD8 polyfunctional profiles reported in our findings could be a result of differing degrees of memory T cell differentiation and/or lower SARS-CoV-2 peptide binding affinity with HCoV-specific T cells resulting in low T cell activation.The affinity between MHC molecules presenting peptides (pMHC) and the T cell receptor (TCR) plays a role in antigen recognition.3][74] TCR cross-recognition of pMHCs that are not structurally identical results in lower binding affinity which decreases T cell polyfunctionality. 73,75It is therefore plausible that the distinct polyfunctional profile of SARS-CoV-2 CD4 + T cell responses in seronegative children observed in our study may be mediated by the cross-recognition of pre-existing T cell immunity to HCoV-HKU1, where partial sequence homology between HCoV-HKU1 and SARS-CoV-2 results in low peptide binding affinity and a different polyfunctional T cell profile. 52,58,72,75,76It is noteworthy to report that we did not find any association with the other beta HCoV-OC43 spike-specific IgG and SARS-CoV-2-specific T cells in seronegative children, which could be related to differences in seasonal prevalence of types of HCoV infections.8][79][80] In particular, a study by Friedman et al. showed a high prevalence of HCoV-OC43 was the most prevalent HCoV in winter but HCoV-HKU1 was the most prevalent in spring to summer during the 2015-2016 time period. 3139]44 Initially it was thought that children may have a higher magnitude of T cell responses given their relative resistance to severe disease and the link between T cell responses and protection from severe disease. 8,81However, children were shown to have lower T cell responses than adults in most studies, including the current study. 23,37,38Several hypotheses could possibly explain the differences in the magnitude of SARS-CoV-2-specific CD4 + T cell responses between children and adults in our study.These include 1) differences in disease severity between the groups, 2) the immaturity of children's immune system and 3) the timing of infection in relation to sampling.Adults have a mature immune system with more differentiated memory T cell subsets endowed with increased cytokine capacity, whereas children have an immature immune system, with many more naive T cells which have reduced cytokine producing capacity, are enriched for monofunctional responses, and have increased antigen dependence. 82However, as the exact time of SARS-CoV-2 infection in the children group could not be defined, we cannot fully exclude that the higher magnitude of SARS-CoV-2-specific CD4 + T cells observed in the adult group could be related to a more recent infection compared to the children group.Despite lower circulating SARS-CoV-2-specific T cells, children and adults may also have different responses in the respiratory tract.There is evidence in adults demonstrating the presence of SARS-CoV-2-specific T cell responses in the nasal mucosa after infection and vaccination. 83,84However, comparative studies have yet to be performed in children.17]85 Recent studies have shown that, compared to adults, SARS-CoV-2 seronegative children exhibited an increased number of innate immune cells with pre-activated signatures leading to the early production of interferon-mediated antiviral effects in the upper respiratory tract. 18,19n conclusion, our study shows that a robust SARS-CoV-2-specific T cell response is observed in children, including those with no evidence of prior SARS-CoV-2 infection.We demonstrate that the magnitude of SARS-CoV-2-reactive CD4 + T cells in seronegative children correlates with HCoV-HKU1 exposure.This, together with a distinct functional profile of SARS-CoV-2-specific responding CD4 + T cells observed between seropositive and seronegative children may provide further evidence for pre-existing T cell responses cross-reactive to SARS-CoV-2.

Limitations of the study
This study relied on serology to determine SARS-CoV-2 exposure in the pediatric cohort.With no PCR confirmation of SARS-CoV-2 infection, the exact time of infection of the children could not be determined.This caveat in our study may have an impact in the differences of SARS-CoV-2-specific T cell responses seen in adults compared to seropositive children.Additionally, our study investigated T cell responses against SARS-CoV-2 spike, nucleocapsid and membrane proteins, and not the non-structural viral proteins that can also serve as targets for the T cell response, thus not capturing the full extent of T reactivity in infected participants.While our study focused solely on Th1-associated cytokines, it remains to define whether other Th CD4 subtypes such Th2 or Th17 could as also play a role in COVID-19 protection.Furthermore, we did not address durability of the T cell responses in children due to the cross-sectional design of the study.Moreover, it would be of interest to assess the HCoV-HKU1-specific memory T cell responses and define the relationship with SARS-CoV-2 specific immune responses and their potential protective effect against SARS-CoV-2.Further experiments assessing HCoV-IgG responses as well as the neutralizing antibody activity against HCoVs, SARS-CoV-2 and its variants of concern are warranted to gain a better understanding of the relative protection from disease observed in SARS-CoV-2-infected children compared to adults.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

Figure 1 .
Figure 1.SARS-CoV-2-specific antibody responses in children (A) The demographic characteristics of 71 unvaccinated children included in this study.Age, gender, SARS-CoV-2 vaccination status, SARS-CoV-2 serology and collection date are shown.(B) Proportion of children exhibiting antibody responses to SARS-CoV-2 spike (S) and nucleocapsid (N) proteins.(C) The magnitude of SARS-CoV-2 S and N IgG antibodies (OD 490nm ) measured by ELISA in seropositive children (n = 41).The dotted lines indicate the cut-off for positivity which was calculated as the mean optical density of COVID-19 prepandemic control samples.Statistical analysis was performed using the Wilcoxon signed rank test; p values < 0.05 were considered statistically significant and are bolded.See also Figure S1.

Figure 2 .
Figure 2. SARS-CoV-2-specific T cell responses in children (A) Representative flow cytometry plots of SARS-CoV-2-specific interferon-g (IFN-g), tumor necrosis factor-a (TNF-a) and interleukin-2 (IL-2) cytokine production from CD4 + (left) and CD8 + (right) T cells in response to SARS-CoV-2 peptide stimulation.NS: no stimulation, SNM: combined peptide pool of SARS-CoV-2-spike, nucleocapsid and membrane proteins.(B) Frequency of SARS-CoV-2-specific CD4 + T cells producing any of the measured cytokines (IFN-g, IL-2 or TNF-a).Children were grouped according to SARS-CoV-2 serostatus (light blue: n = 41 SARS-CoV-2 seropositive; dark blue: n = 30 SARS-CoV-2 seronegative).Bars represent the median of the responders, and median values are indicated.The pie charts represent the proportion of responders with detectable T cell response to SARS-CoV-2 SNM peptides.(C) Frequency of SARS-CoV-2-specific CD8 + T cells producing any of the measured cytokines (IFN-g, IL-2 or TNF-a).Children were grouped according to SARS-CoV-2 serostatus (light red: n = 41 SARS-CoV-2 seropositive children; dark red: n = 30 SARS-CoV-2 seronegative children).Statistical comparisons in (B) and (C) were performed using the Mann-Whitney test between seropositive and seronegative children and the Chi-square test to compare the percentage of responders; p values < 0.05 were considered statistically significant and are bolded.(D)Polyfunctional profile of SARS-CoV-2-specific CD4 + and (E) CD8 + T cells in seropositive and seronegative unvaccinated children.The x axis illustrates each combination which is indicated with a black circle for the presence of IFN-g, IL-2 and TNF-a.The medians and interquartile range are shown.Each response pattern (any possible combination of IFN-g, IL-2 and TNF-a production) is color coded and summarized in the pie charts, with each pie slice representing the median contribution of each combination to the total SARS-CoV-2 responses.The permutation test was used to compare the statistical differences between the pie charts and the Mann Whitney Sum Test to compare response patterns between seropositive and seronegative children; p values < 0.05 were considered statistically significant and are bolded.

Figure 3 .
Figure 3. SARS-CoV-2 cross-reactivity to endemic beta-HCoV in children (A) The magnitude of HCoV-HKU-1 spike IgG levels were measured by ELISA in SARS-CoV-2 seropositive (light blue; n = 41) and seronegative (dark blue; n = 29) children.Plasma sample was insufficient for one seronegative child therefore OD for HCoV-HKU1 was not measured for this participant.The bars represent the median values.A statistical comparison was performed using the Mann-Whitney test between seropositive and seronegative children; a p value < 0.05 was considered statistically significant.(B) Correlation between the frequency of SARS-CoV-2-specific CD4 + T cells and HCoV-HKU-1-spike IgG levels in SARS-CoV-2 seronegative children (n = 29).One participant had insufficient sample available to be included in this assay.(C) Correlation between the frequency of SARS-CoV-2-specific CD4 + T cells and HCoV-HKU-1-spike IgG levels in SARS-CoV-2 seropositive children (n = 41).Statistical comparisons for (B) and (C) were performed using a two-tailed non-parametric Spearman rank tests; p values < 0.05 were considered statistically significant and are bolded and correlation coefficients values are shown.See also Figure S2.

Figure 4 .
Figure 4. SARS-CoV-2-specific T cell responses in children compared to convalescent adults (A) The demographic characteristics of 30 unvaccinated convalescent adults included in this study.Age, sex, SARS-CoV-2 PCR-positivity, days since PCR test and collection date are shown.(B) Frequency of SARS-CoV-2-specific CD4 + T cells producing any of the measured cytokines (IFN-g, TNF-a, or IL-2) in SARS-CoV-2 convalescent HCW (purple; n = 30) and seropositive children (light blue; n = 41).The pie charts represent the proportion of responders with a detectable T cell response to SARS-CoV-2 SNM combined peptide pools.Bars represent median of the responders.(C) Frequency of SARS-CoV-2-specific CD8 + T cells producing any of the measured cytokines (IFN-g, TNF-a, or IL-2) in SARS-CoV-2 convalescent HCW (dark red; n = 30) and seropositive children (light red; n = 41).The bars represent the median of the responders and median values are indicated.Statistical comparisons were performed using the Mann-Whitney test between seropositive children and adults and the Chi-square test was used to compare the percentage of responders for (B) and (C); p values < 0.05 were considered statistically significant and are bolded.(D) Correlations between SARS-CoV-2-specific CD4 + or CD8 + T cells and age in convalescent HCW (purple and dark red; n = 30) and seropositive (light blue and light red; n = 41) children.Statistical comparisons were performed using a two-tailed non-parametric Spearman rank tests; p values < 0.05 were considered statistically significant and are bolded and correlation coefficients are shown.

TABLE
d RESOURCE AVAILABILITY B Lead contact B Materials availability B Data and code availability d EXPERIMENTAL MODEL AND SUBJECT DETAILS B Study participants d METHODS DETAILS B SARS-CoV-2 and HCoV antigens B Enzyme-linked immunosorbent assay (ELISA) B Whole blood-based T cell assay B Cell staining and flow cytometry d QUANTIFICATION AND STATISTICAL ANALYSIS