The effect of age on the magnitude and longevity of Th1‐directed CD4 T‐cell responses to SARS‐CoV‐2

Abstract Age is associated with changes in the immune system which increase the risk for severe COVID‐19. Here, we investigate SARS‐CoV‐2‐reactive CD4 T cells from individuals recovered from SARS‐CoV‐2 infection with mild COVID‐19 symptoms after 3, 6 and 9 months using incubation with SARS‐CoV‐2 S1, S2 and N‐peptide pools, followed by flow cytometry for a Th1‐activation profile or proliferation analyses. We found that SARS‐CoV‐2‐reactive CD4 T cells are decreasing on average after 9 months but highly polyfunctional CD4 T cells can peak after 6‐month recovery. We show that individuals older than 60 years of age have significantly more SARS‐CoV‐2‐reactive T cells in their blood after 3 months of recovery compared to younger individuals and that the percentage of SARS‐CoV‐2‐reactive Th1‐directed CD4 T cells in the blood of mild‐COVID‐19‐recovered individuals correlates with age. Finally, we show that individuals over the age of 40 have significantly increased the amounts of highly polyfunctional SARS‐CoV‐2‐S‐peptide‐reactive CD4 T cells, compared to SARS‐CoV‐2 naïve individuals, than those under the age of 40. These findings suggest that in individuals recovered from mild COVID‐19, increased age is associated with significantly more highly polyfunctional SARS‐CoV‐2‐reactive CD4 T cells with a Th1‐profile and that these responses persist over time.


INTRODUCTION
The novel human pathogen severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [1] which causes coronavirus disease  has led to significant worldwide mortality since its discovery in late 2019.
Evidence, thus, far has indicated that the severity of COVID-19 after SARS-CoV-2 infection is linked to inefficient virus-specific immune responses. More severe COVID-19 cases have been linked with ineffective innate immunity [2,3] which can lead to a delayed adaptive immune response and a higher viral burden, a situation which has been associated with fatal cases of COVID-19 [4]. Furthermore, adaptive immune response involving the presence of detectable SARS-CoV-2-specific T cells in convalescent individuals has been significantly associated with milder disease [5][6][7]. Evidently, clearance of SARS-CoV-2 is dependent upon adaptive immune responses to SARS-CoV-2 antigens. Effective T-cell immunity has also been associated with the absence of disease progression for multiple viral infections and, specifically, the ability of polyfunctional T cells to generate multiple cytokines and activation markers upon stimulation with viral antigen [8,9]. Virus-specific T cells commonly develop a Th1 profile, which have welldocumented anti-viral properties through the secretion of cytokines such as IFNγ, TNFα and IL-2 and which have shown to be protective against lethal SARS-CoV infection in mice [10].
Age has been identified as one of the largest risk factors for COVID-19 mortality with significantly increased rates of fatalities in individuals aged 65 years or older [11,12] and mortality from SARS-CoV-2-infection being approximately 10 times higher for a 40-year old compared to a 20-year old [13]. Ageing is associated with increased T-cell senescence, which may contribute to inefficient responses to viral infections [14]; however, conversely, senescence is also associated with increased T-cell inflammatory responses [15], which may add to the immunopathology of COVID- 19. Several studies have demonstrated that age is associated with a perturbed and reduced functional immune response to SARS-CoV-2 infection involving both the innate and adaptive immune system, which leads to increased inflammation and adverse COVID-19 outcome [16,17]. Furthermore, increased age is also associated with a decrease in naïve T cells [18] and COVID-19 severity has been inversely correlated with naïve T-cell frequency [5]. The ability of an immune system responding to a novel virus relies upon the repertoire of naïve T cells available and so an older individual is less likely to have an immune system as capable to respond as a younger one [19]. This may lead to a slower T-cell response and more severe course of disease.
Whilst increased age is associated with an increased likelihood of severe COVID-19, the effect of age on the adaptive immune response to SARS-CoV-2 in those individuals who do not develop severe symptoms is less well studied. We, therefore, investigate whether Th1-directed CD4 T-cell responses to SARS-CoV-2 peptides are different in individuals recovered from mild COVID-19 at 3, 6 and 9 months after recovery within different age groups.

Study participants
A cohort of volunteers in a German SARS-CoV-2 high incidence area were recruited for sampling to determine the seropositivity prevalence in the population; the results of which were described elsewhere [20] (German Clinical Trials Register https://www.drks.de ID: DRKS00021306). A follow-up study to re-visit these individuals was approved and individuals were sampled 3 months and 6 months after the initial visit (German Clinical Trials Register, https://www.drks.de, ID: DRKS00023113). The study was approved by the Ethics Committee of the Medical Faculty of University Bonn (approval number 085/20). In all visits, plasma was collected and PBMCs isolated through density centrifugation and cryopreserved for downstream analyses. Study T A B L E 1 Study participants. V1 < 3-month recovery due to the proximity to the arrival of the SARS-CoV-2 virus in Europe in January 2020 -WHO situation reports [42], V2 (6-9-month recovery), V3 (9-12 months recovery participants across all three visits are detailed in Table 1 with reported symptom severity within age groups outlined in Table 2. Participants who were vaccinated or PCR positive for SARS-CoV-2 (methods described elsewhere, [20]), at any time-point, were not included within this study. SARS-CoV-2 naïve individuals were sampled from individuals within the first visit group.

Proliferation assay
Cryopreserved PBMCs were thawed, washed and left overnight in RPMI 1640 (Gibco) media containing 5% human AB serum (PAN-Biotech), 2 mM L-Glutamine, 1% penicillin/streptomycin and 5 U/ml Benzonase Nuclease HC, Purity >99% (Millipore). The next day, viable cells were counted via trypan blue exclusion and plated in media containing 5% human AB serum (PAN-Biotech), 2 mM L-Glutamine and 1% penicillin/ streptomycin at 375000 cells per well into 96-well flatbottomed plate (Corning) in triplicate and stimulated at 0.25 μg/ml with Prot_N (N), Prot_S1 (S1), Prot_S (S2) (Miltenyi Biotec), as well as positive control (PHA 0.5 μg/ml) and negative control (DMSO). Plates were incubated for 48 h. Proliferation was detected using a BrdU Cell Proliferation Assay kit (Sigma) as per manufacturer instructions. In short, BrdU labelling solution was added to wells for the last 20 h of incubation, cells were then dried, fixed and incubated for 90 min with anti-BrdU-POD (peroxidase) monoclonal antibody.
Next, substrate solution (Tetramethylbenzidine) was added and colour change halted after 5 min with 25 μl 1 M H 2 SO 4 . Plates were read in an ELISA reader at 450 nm. Data are shown as stimulation index (SI) as a fold change compared to DMSO control wells.

Statistical analyses
Data collection and analysis were performed using FlowJo 10 and GraphPad Prism 9.1. Data are shown as mean + 95% confidence intervals. Activation marker positive T cells were calculated by subtracting the unstimulated DMSO control for each donor with negative values being interpreted as zero. In all figures, * represents a p value of <0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

Persistence of SARS-CoV-2-reactive adaptive immune responses after recovery from COVID-19
Previous studies have demonstrated that the presence of SARS-CoV-2-reactive T cells within an individual is protective against severe COVID-19 outcome [5] [21]. To quantify SARS-CoV-2 reactive CD4 T cells in individuals who recovered from COVID-19 at least 9 months post recovery, PBMCs were cultured with peptide pools from the Nucleocapsid (N) or Spike (S1) (S2) regions of SARS-CoV-2 and CD4 T cells were analysed via flow cytometry for upregulation of Th1-related cytokines IFNγ, IL-2 and TNFα as well as expression of CD154 (representative gating shown in Figure S1). Analysis of SARS-CoV-2-reactive CD4 T cells with two upregulated activation markers showed the strongest response at the earliest timepoint, less than 3 months after initial infection which then decreased in magnitude over the next 9 months (Figure 1a). S1 and N, but not S2-reactive CD4 T cells with two upregulated functional activation markers were still detectable in individuals who recovered from COVID-19 at least 9 months after recovery when compared to SARS-CoV-2 naïve individuals ( Figure 1, A, 9-12 months recovery). Analyses of SARS-CoV-2-peptide-induced increase in proliferation revealed significant proliferative responses at less than 3-month recovery with N and S2-peptide pool stimulation (Figure 1b, <3-month recovery, N p = 0.007, S2 p = 0.001), at 6-month recovery with S2 but not N or S1 stimulation (Figure 1b, 6-9-month recovery, S2 p = 0.045) and at 9-month recovery with all three peptide pools used (Figure 1b, 9-12 months recovery, N p = 0.002, S1 p = 0.43, S2 p = 0.001).
Taken together, these data indicate individuals who recovered from COVID-19 retain CD4 T cells capable of responding to SARS-CoV-2 peptides at least 9 months after recovery towards both the spike and nucleocapsid of SARS-CoV-2.
Collectively, these data indicate that after 3 months of recovery, less polyfunctional Th1-directed responses to SARS-CoV-2 peptides decrease over time but more polyfunctional CD4 T-cell responses to SARS-CoV-2 involving IFNγ, IL-2 and CD154 and with or without TNFα continue to increase after more than 3 months of recovery from COVID-19.

Association of increased age with SARS-CoV-2-reactive CD4 T cells in individuals recovered from COVID-19
There is evidence to suggest that age is associated with an elevated CD4 T-cell immune response to SARS-CoV-2 in adults compared to children [24]; however, the CD4 T-cell immune response over time to SARS-CoV-2 throughout the ageing population who do not develop severe COVID-19 is less well studied. To investigate the effect of age on F I G U R E 1 N-, S1-and S2-induced immune activation over 9 months in individuals recovered from mild COVID-19. (a) Comparison of dual-functional activation marker-positive CD4 T-cell responses from three time-points over 9 months from SARS-CoV-2-recovered or naïve individuals is shown after 6-h incubation with N, S1 or S2 peptide pools. (b) Comparison of SARS-CoV-2 N-, S-and S2-related peptideinduced proliferative responses from PBMCs taken from SARS-CoV-2-recovered or naïve individuals. Data are shown as fold change compared to DMSO control proliferation (Stimulation Index, SI). Group <3-month recovery n = 20, 6-9-month recovery n = 29, 9-12 months recovery n = 30, SARS-COV-2 naïve n = 27. Data are represented as mean + 95% confidence intervals with differences between groups assessed using Mann-Whitney corrected for multiple comparisons using the Holm-Šíd ak method *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (c) Proportional analyses of three-to-four-functional marker-positive CD4 T-cell responses over 9 months from SARS-CoV-2 recovered or naïve individuals after incubation with N, S1 or S2 peptide pools. Data show mean values as a percentage of all three-to-fourfunctional CD4 T cells measured which is stated below each pie chart. The frequency of these poly-functional responses can be seen in D. Group <3-month recovery n = 154, 6-9-month recovery n = 62, 9-12-month recovery n = 43, SARS-COV-2 naïve n = 327. Data are represented as mean + 95% confidence intervals. Differences between groups were assessed using two-way ANOVA with Tukey's multiple comparisons test *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 the frequency of SARS-CoV-2-reactive CD4 T cells after incubation with SARS-CoV-2 peptides, individuals were grouped by age and CD4 T-cell responses analysed. The most robust SARS-CoV-2-induced CD4 T-cell responses were observed in recovered individuals older than 60 years with significantly higher amounts of SARS-CoV-2-reactive CD4 T cells at 3-, 6-and 9-month recovery (Figure 2a, 3a).
Notably, age was positively associated with significant increases in IFNγ+ IL-2+ CD4 T cells at under 3 and 6 months post recovery upon challenge with S1 peptide against representative responses from all younger age groups (Figure 2a). More poly-functional age-specific increases in SARS-CoV-2-reactive CD4 T cells are seen in Figure 3a with individuals over the age of 60 having  (Figure 3a, <3-month recovery, N and S1), 6-month recovery (Figure 3a, 6-9-month recovery, S1 and S2) and at 9 months post recovery with individuals over the age of 60 having elevated S2 responses over the younger age groups under 40 years of age (Figure 3a, 9-12 months recovery, S2).
Whilst increasing age groups are positively associated with the magnitude of particular polyfunctional CD4 Tcell responses to SARS-CoV-2 peptides, analyses of polyfunctional CD4 T-cell response profiles can indicate the quality of CD4 T-cell response, as well as the evolution over time. Analyses of the mean three-to-four-functional S1-response-profiles of all age groups are seen over time in Figure 3b. Poly-functional CD4 T-cell response profiles to S1-peptide pools did remain visually comparable between all age groups at 3 and 6 months of recovery and are seen in Figure 3b (<3-month recovery, 6-9-month recovery); however, older groups progressively recruited increased percentages of total three-to-four-functional CD4 T cells at every time-point (Figure 3b <3-month recovery, 6-9-month recovery and 9-12-month recovery-% CD4 T cells). The evolution of S1-reactive polyfunctional CD4 T-cell profile at 3-, 6-and 9-month recovery was comparable for age groups under 60 years of age with a general increase over time in the proportion of responding three-functional CD154 + IFNγ + IL-2 + CD4 T cells (Figure 3b 0-19 YO, 20-39 YO, 40-59 YOpurple segment); whereas, in individuals over the age of 60, more poly-functional CD154 + IFNγ + IL-2 + TNFα + CD4 T cells can be seen increasing over time (Figure 3b 60+ YOorange segment) indicating that individuals of increased age can generate a more polyfunctional SARS-CoV-2 response profile.
Taken together, the detectable increase in adaptive immune responses against SARS-CoV-2 peptides within older individuals compared to younger age groups, particularly those under the age of 40, indicates that age may be associated with the development of increased adaptive Th1-directed immune responses against SARS-CoV-2-infection even when COVID-19 symptoms are only mild or asymptomatic.

DISCUSSION
Our data indicate that increased age is associated with a more robust CD4 T-cell response to SARS-CoV-2-peptides with a Th1 phenotype in individuals with comparably mild COVID-19 symptoms and recovery time. Individuals over the age of 60 had significantly more SARS-CoV-2-reactive polyfunctional CD4 T cells with a Th1 phenotype than those of younger age groups which persisted for at least 9 months after recovery when compared to those under the age of 40. Age has been associated as one of the largest risk factors for adverse COVID-19 outcomes [11,12], whilst adaptive immune responses involving SARS-CoV-2-specific T cells in convalescent individuals have been associated with milder COVID-19-symptoms [5][6][7]. Within this study, we show that in individuals who do not develop severe COVID-19, age is associated with a more robust polyfunctional Th1-CD4 T-cell response to SARS-CoV-2 peptides, indicating a protective role for SARS-CoV-2-directed Th1-CD4 T-cell responses with increasing age. A Th1-mediated response, as investigated here, has been regarded as a more protective phenotype when related to SARS-CoV-2 immunity [10]; however, it is also related to the general immunopathogenesis of more severe COVID-19 [25]. The involvement of Th1 immune responses, therefore, seems to be beneficial, however only when SARS-CoV-2-specific. It has also been reported that SARS-CoV-2 challenged T-cell responses from individuals recovered from mild COVID-19 lean towards a Th2-phenotype [26], which is also the case with children who similarly tend towards a Th2 phenotype, which may explain the generally milder COVID-19 symptoms seen within the young [27]. Investigations into a Th2 phenotype were beyond the scope of this project and future investigations into age-specific differences in those recovered from mild COVID-19 should include Th2 phenotypic analyses.
Lower symptomatic infection and more beneficial outcome of SARS-CoV-2 infection have been associated with a robust adaptive immune response to viral antigen and a growing body of evidence indicates that a successful recovery from SARS-CoV-2-infection relies upon significant T-cell involvement [5,7] involving high avidity SARS-CoV-2-specific T cells [28]. In this study, we show that in individuals that recovered from mild COVID-19, SARS-CoV-2-reactive CD4 T cells have persistent and strong immune responses. The presence of SARS-CoV-2-reactive CD4 T cells is an indication of protective immunity from re-infection or severe course of infection [29,30]. Indeed, it has been demonstrated for other related respiratory viruses such as SARS-CoV-1 [31] that induced memory T cells which were detectable 17 years after infection [32]. As with data shown here, others have also shown SARS-CoV-2 antigen-induced T-cell responses in convalescent individuals which persist for at least 6 months and then begin to wane [33][34][35][36].
Polyfunctional T cells have been strongly linked with an increased ability to combat viral infections [22,23] and early protection, even before antibody production [21]. The highly polyfunctional SARS-CoV-2-reactive CD4 T cells upregulating functional activation markers IFNγ, IL-2 and CD154 (which were also either TNFα positive or negative) peaked after around 6-month recovery. This increase was also accompanied with a subsequent steady decrease in polyfunctional CD4 T cells which were not upregulating IFNγ. This shift towards a larger percentage of polyfunctional CD4 T cells upregulating IFNγ and TNFα could be attributed to differentiation towards a Th-1 central memory phenotype, which has been shown to be increasing until at least 6 months post symptom onset [33] and is consistent with polyfunctional CD4 T-cell responses in individuals infected and challenged with SARS-CoV-1-peptides [37].
The effect of ageing on the immune system can be characterized by a higher initial pro-inflammatory status alongside a progressive inability of the immune system to mount efficient responses [38]. A potential explanation for this is an association with age and a progressive decrease in naïve T-cell populations [18], which has also been associated with more severe COVID-19 symptoms [5]. It is possible that a reduction in naïve T cells with age leads to a delayed adaptive immune response to the novel SARS-CoV-2 antigens, which, even in those individuals who develop mild symptoms, have a dampened response to the virus, leading to an increase in viral load, a more drawn-out recovery, and a more robust CD4 T-cell response to SARS-CoV-2 peptides upon challenge.
The increase in SARS-CoV-2-reactive polyfunctional CD4 T cells at 6 months post recovery was not accompanied with an increase in mono-functional (data not shown) or dual-functional activation marker-positive SARS-CoV-2-reactive CD4 T cells which were equal to or significantly decreased at 6-month recovery compared to at under 3-month recovery, yet still significantly increased over SARS-CoV-2-naïve responses. Corroborating publications towards a robust CD4 T-cell response at around 6-month recovery from COVID-19 also indicate the involvement of polyfunctional T cells involving IL-2, IFNγ and TNFα [29,30]. Additionally, decreases in monofunctional T-cell responses to S-peptides have been reported previously at 5 months after diagnosis of SARS-CoV-2-infection in health-care workers; however, polyfunctional T-cell responses were not analysed [39]. Limitations within this study include the use of PBMCs to investigate SARS-CoV-2-reactive CD4 T cells, excluding resident memory T cells, which form an integral part of the immune homeostasis and cannot be sampled from venous blood [40]. Further studies should include longevity and age-associated SARS-CoV-2-reactive tissuespecific T-cell analyses.
Immune responses in SARS-CoV-2 naïve individuals in particular towards S but not towards N, have been reported, with around 40-60% of unexposed donors responding to SARS-CoV-2 peptides [41]. As it was not possible to exclude which of the SARS-CoV-2-naïve donors in this study had cross-reactive T cells, it is likely that analyses comparing S1 and S2-reactive CD4 T-cell responses in SARS-CoV-2-exposed, against SARS-CoV-2-naïve, are less significant than if those cross-reactive naïve individuals were excluded, as would be ideal.
Taken together, our findings indicate that in individuals who recovered with comparably mild COVID-19 symptoms, SARS-CoV-2-reactive CD4 T cells are still detectable at least 9 months after SARS-CoV-2 recovery. We show that age is associated with significantly more robust Th1-directed CD4 T-cell responses, with higher amounts of SARS-CoV-2-reactive polyfunctional CD4 T cells in individuals over the age of 60 compared to those under the age or 40. However, we show that some SARS-CoV-2-reactive CD4 T cells with three to four Th1-directed positive functional activation markers remain comparable or increased at over 9-month recovery when compared to at under 3-month recovery, including in individuals under 40 years of age. The reduced CD4 T-cell Th1 responses to SARS-CoV-2 peptide-challenge in individuals recovered from mild COVID-19 under the age of 40 raises questions regarding the long-term protective immunity that natural infection imbues in the young.