Development of Influenza-Specific CD4 T Cell-Mediated Immunity in Children Following Inactivated Influenza Vaccination

Abstract

Background

While both cellular and humoral immunity are important in immunologic protection against influenza, how the influenza-specific CD4 T cell response is established in response to early vaccination remains inadequately understood. In this study, we sought to understand how the CD4 T cell response to inactivated influenza vaccine (IIV) is established and develops throughout early childhood.

Methods

Influenza-specific CD4 T cell responses were quantified following IIV over 2 influenza seasons in 47 vaccinated children between 6 months and 8 years of age who had no documented history of natural influenza infection during the study. Peripheral blood mononuclear cells were stimulated with peptide pools encompassing the translated regions of the pH1, H3, HAB, and NP proteins, and CD4 T cell responses were assessed via multiparameter flow cytometry.

Results

There was boosting of H3- and HAB-specific CD4 T cells but not cells specific for the pH1 HA protein post-vaccination. A positive correlation between age and the magnitude of the influenza-specific CD4 T cell response was seen, with an overall greater magnitude of IFNγ-producing cells in subjects ≥3 years of age. Changes in CD4 T cell functionality were also noted in older compared to younger children, with increases in CD4 T cells producing IFNγ and TNF or IL-2 as well as IFNγ alone.

Conclusions

Inactivated influenza vaccine elicits a CD4 T cell response to H3 and HAB, with increases in the magnitude of the CD4 T cell response and changes in cellular functionality throughout childhood. This suggests that repeated influenza vaccination contributes to the development of anti-influenza CD4 T cell memory in children.

Inactivated influenza vaccine stimulates an influenza-specific CD4 T cell response in children, with boosting of H3- and HAB-specific cells post-vaccination and changes in CD4 T cell functionality throughout childhood. This suggests that early childhood vaccination may contribute to imprinting of the influenza-specific immune response.

INTRODUCTION

Influenza remains an important health problem in young children, with high rates of morbidity in children younger than the age of 5 years as well as in those with preexisting conditions [1]. For this reason, yearly administration of inactivated influenza vaccine (IIV) is recommended beginning at the age of 6 months in the United States [2]. There is a growing understanding of influenza-specific immunity in adults; however, there are critical differences in innate and adaptive immune function in children, with evolution of the phenotype and function of T cells and the amount of preexisting immunologic memory over the course of childhood [1, 3]. While increases in serum HAI titers have been detected widely following 2 doses of influenza vaccine in children [4], the role of T cells in the development of anti-influenza immunity following childhood vaccination remains less well understood.

Influenza-specific CD4 T cells have been found to provide independent protection against infection [5–7], so understanding how cellular immunity develops against this virus is a high priority. CD4 T cell responses are commonly detected following IIV administration in adults, with a positive correlation between influenza-specific CD4 T cells and the antibody response [8–13]. Studies evaluating the anti-influenza CD4 T cell response in children have been conflicting. One previous study by Zeman et al. demonstrated a higher baseline frequency of CD4 T cells in children with more previous immunizations; however, an increase in the frequency of influenza-specific CD4 T cells post-vaccination was not seen [14]. Increases in influenza-specific CD4 T cells producing IFNγ were detectable post-vaccination in 2 subsequent studies that included children; however, neither of these studies addressed CD4 T cell specificity or functionality [15, 16].

In a related study, our group compared H3- and NP-specific CD4 T cell reactivity following acute H3N2 infection in 16 acutely infected children compared to 28 age-matched, IIV-immunized children. While we documented a greater magnitude H3-specific CD4 T cell response following acute infection, this advantage was not sustained post-vaccination in study year 2, with vaccinated subjects demonstrating a gradual increase in H3-specific CD4 T cell reactivity over time. However, this study only evaluated a subset of the present cohort and quantified reactivity to a limited number of proteins, with direct comparisons limited to between the infected and vaccinated groups [17].

In the present study, we sought to understand how CD4 T cell immunity is established and boosted by IIV in children, examining both response magnitude and functionality of the CD4 T cell response over time and in different age groups. We found that vaccination boosted CD4 T cells specific for H3 and HAB but not the pH1 protein. Further, when subjects were grouped by age, older children had a greater magnitude CD4 T cell response, with distinct shifts in CD4 T cell functionality. Overall, these results suggest that IIV shapes the HA-specific anti-influenza CD4 T cell response in children; however, there may be further editing of this response upon acute infection that broadens the response to more conserved CD4 T cell epitopes and shifts the functionality of the response toward a more Th1-like phenotype.

METHODS

Human Subjects

Following approval from the University of Rochester Research Subjects Review Board (protocol RSRB00058437), 47 evaluable pediatric subjects between the ages of 6 months and 8 years of age were enrolled into the vaccination cohort of this observational study, predominately in the fall of 2016 (Table 1). Inclusion criteria required subjects to be between 6 months and 8 years of age, born at ≥37 weeks gestational age, and have a history of previous influenza vaccination if >12 months of age and enrolled into the vaccination cohort. Subjects were excluded if they had a history of immunosuppression, active neoplastic disease, use of potentially immunosuppressive medications within the past year, a diagnosis of asthma requiring chronic controller medication, history of receipt of immunoglobulin or another blood product within the past year, a contraindication to influenza vaccination, previous administration of an influenza vaccine in the current influenza season (if in the vaccination cohort), or an acute illness within the past 3 days (if in the vaccination cohort). The primary objective of the study was to comprehensively evaluate the specificity and functional potential of influenza-specific CD4 T cell responses on a single cell level using intracellular cytokine staining (ICS) following restimulation with pools of HA and NP peptides. To accomplish this objective, venous blood was obtained from all enrolled children pre-vaccination and at 8–14 days and 20–28 days post-vaccination with seasonal IIV (Quadrivalent Fluzone). Enrolled subjects were then followed longitudinally and revaccinated with Quadrivalent Fluzone the next year, with blood again obtained pre- and post-vaccination. Due to greater intersubject variability at early timepoints post-vaccination, data from day 20–28 post-vaccination is presented in this manuscript. All subjects had previous vaccination history verified in the New York State Immunization Information System, and no subject in the vaccination cohort had an acute influenza infection documented during the study.

Table 1.

Subject Demographic Data

Subject Age	<3 yo	≥3 yo
Total (n)	20	27
Sex
Male (n)	10	15
Female (n)	10	12
Race
White (n)	9	18
Black (n)	6	1
Native American (n)	1	0
Mixed (n)	4	8
Ethnicity
Non-Hispanic (n)	18	27
Hispanic (n)	2	0
Vaccination Hx
Vaccinated ever (%)*	65	100
Vaccinated previous year (%)	65	92.6

^*In the <3-year-old population, 7 subjects were ineligible for vaccination in the previous influenza season due to age.

In completion of this study, experimentation guidelines of the United States Department of Health and Human Services and the University of Rochester were followed, and study procedures were performed in accordance with the ethical standards of the Helsinki Declaration. Parents of all study subjects provided written informed consent prior to study participation. The study was registered in the clinicaltrials.gov database (NCT02559505). A subset of data obtained from this cohort was previously published as part of a different analysis [17].

Isolation of PBMCs From Human Blood

Blood was centrifuged and plasma was removed within 6 hours of blood draw. Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll density gradient centrifugation (Ficoll-Paque Plus; GE Healthcare). The resulting PBMCs were frozen at 5–10 × 10⁶ cells/mL at a controlled rate in fetal bovine serum (Gibco) containing 10% dimethyl sulfoxide.

Synthetic Peptides and Libraries

Peptides encompassing the entire translated sequences of the examined viral proteins (17-mer and 15-mer overlapping by 11 amino acids) were obtained from the Biodefense and Emerging Infections Research Repository, National Institute of Allergy and Infectious Diseases, National Institutes of Health. The peptide arrays included A/California/04/2009(H1N1) pdm09 hemagglutinin (pH1; NR-15433), A/New York/384/2005(H3N2) hemagglutinin (H3; NR-2603), B/Florida/4/2006 (HAB) hemagglutinin (HAB; NR-18972), and A/New York/348/2003(H1N1) nucleoprotein (NP; NR-2611), with 111 peptides from Sin-Nombre Virus (NM H10) glycoprotein precursor protein (NR-4764) used as a negative control. All peptides from a given protein were pooled, with each peptide present at a final concentration of 1 μM in assays.

Intracellular Cytokine Staining

PBMCs were thawed and rested unstimulated at 37°C and 5% CO₂ for 16–18 hours overnight in RPMI-1640 (Gibco) containing 10% fetal calf serum. Cells were then stimulated with the pH1, H3, HAB, NP, and Sin-Nombre peptide pools for 16 hours, with monensin and brefeldin A (BD Biosciences) added after 8 hours of incubation. Post-stimulation, cells were stained for viability using Live/Dead Aqua (ThermoFisher Scientific). Surface and intracellular staining was performed as per previously published protocol [17]. Antibodies used included CD8 (RPA-T8, BD catalogue #560662), CD4 (RPA-T4, BD #555346), CD3 (SK7, BD #560176), IL-2 (MQ1-17H1, BD #564165), IFN-γ (B27, BD #562988), TNF (MAB11, BD #559321) and CD69 (FN50, BD #555532) as well as CD19 (HIB19, Biolegend #302236). Cell fixation and permeabilization were performed using the eBioscience FoxP3 Transcription Factor Fixation/Permeabilization Kit (ThermoFisher Scientific). Data were acquired using a BD LSR-II flow cytometer configured with 488, 633, 407, and 532-nm lasers. Sequential gating was performed using FlowJo v10.9 (TreeStar). Influenza-specific CD4 T cells were characterized as live, CD3+, CD4+ cells expressing cytokine (IFNγ, IL-2, or TNF) and CD69+ after subtraction of background (Supplementary Figure 1).

Statistical Analysis

Data are presented throughout the manuscript as a percentage of live CD4 T cells with background subtracted. Polyfunctionality graphs were generated using PESTLE and SPICE software (NIAID), while bar and scatter graphs were created using Prism 10 (GraphPad Prism). Data are presented as box plots, with the median shown as a horizontal line, the edges of the box demonstrating the 1st and 3rd quartiles, and the whiskers displaying the 10th and 90th percentiles. Each subject that is an outlier is identified by their own individual symbol. Statistical analyses were performed using SAS 9.4 (SAS Institute). The Wilcoxon rank-sum test was applied to compare cytokine responses across age groups, while the Wilcoxon Signed rank test was utilized for paired comparisons. Correlation analysis was completed using the Spearman correlation analysis. Multiple test adjustments were not applied due to the small sample size and exploratory nature of this analysis. Conclusions were confirmed by repeated ANOVA applied to simultaneously use all the data and to consider within-subject correlation. For all analyses, a P-value < .05 was considered significant.

RESULTS

This observational study was designed to evaluate the influenza-specific CD4 T cell response following IIV in children between 6 months and 8 years of age. A cohort of 47 children was evaluable, with gender evenly distributed between age groups (Table 1). Ninety-five percent of enrolled, vaccine-eligible children had been vaccinated the previous season (7 enrolled subjects were too young for IIV administration the previous year). Of the vaccine-eligible children, 17.5% had previously received at least 1 previous dose of live attenuated influenza vaccine. Upon enrollment, subjects were vaccinated with IIV after obtaining informed consent, with blood sampled pre- and post-vaccination. Children were then longitudinally followed through the year with passive reporting of influenza-like illness, with no active influenza infections documented in the vaccinated cohort. The following fall, all subjects were again evaluated before and after repeat vaccination with seasonal IIV (Figure 1A). Cryopreserved PBMCs were stained using ICS after stimulation with complete overlapping peptide pools representing the entire translated sequence of the pH1, H3, HAB, and NP proteins, with the resulting data gated on live, CD3+, CD4+ cells (Supplementary Figure 1). Cytokine production was quantified as the percentage of activated (CD69+) cells expressing IFNγ, IL-2, or TNF after subtracting background.

Figure 1.

There is evidence for boosting of activated CD4 T cells producing IFNγ following IIV administration. Pediatric subjects between 6 months and 8 years of age were enrolled and vaccinated with IIV in the fall of study years 1 and 2, as shown in (A). Isolated PBMCs obtained pre- and post-vaccination were stimulated with pools of overlapping peptides encompassing the entire coding regions of the pH1, H3, HAB, and NP proteins, after which they were stained by ICS. The resulting data were sequentially gated (as shown in Supplementary Figure 1), with (B) demonstrating the percentage of CD4 T cells expressing CD69 and IFNγ at pre- and post-vaccination timepoints in both study years. Data are depicted as box and whisker plots, with the box representing the 1st and 3rd data quartiles and the whiskers denoting the 10th to 90th percentiles. Outliers are shown using a separate symbol to identify each given subject. A Wilcoxon Signed rank test was applied to examine for differences between study timepoints, with P-values < .05 considered statistically significant.

In our initial analysis, we followed subjects longitudinally over time to evaluate for CD4 T cell boosting post-vaccination. When evaluating all subjects and comparing the pre-vaccination timepoint to day 20–28 post-vaccination, boosting of H3- and HAB-specific CD4 T cells was demonstrated when quantified by the percentage of IFNγ+CD69+ expressing cells. There was also an increase in the magnitude of the H3- and HAB-specific response over time when comparing between baseline and visit 6 (day 20–28 post-vaccination in study year 2; Figure 1B and Supplementary Figure 2). When younger and older children were examined independently, an increase in CD4 T cells specific for the H3 and HAB proteins was again demonstrated. Boosting of NP-specific CD4 T cells was also seen in the younger cohort between the baseline year 1 timepoint and both timepoints in year 2, possibly due to undiagnosed acute influenza infections or NP protein within the influenza vaccine [18, 19] stimulating an antigen-specific CD4 T cell response (Supplementary Figure 3). Interestingly, we were unable to detect a significant increase in CD4 T cells specific for pH1 HA post-vaccination in either the total cohort or when subjects were banded by age.

Next, we evaluated how the overall magnitude of the influenza-specific CD4 T cell response was influenced by subject age. We were able to document a positive correlation between age and CD4 T cell reactivity as measured by IFNγ production following stimulation with the H3, HAB, and NP but not the pH1 protein at baseline in study year 1 (Supplementary Figure 4). Next, the cohort was divided into younger (<3 years) and older (≥3 years) children and the magnitude of the CD4 T cell response to the pH1, H3, HAB, and NP proteins was compared between age groups at each timepoint (Figure 2). We found a greater percentage of CD69+IFNγ+CD4 T cells specific for H3 and HAB in older compared to younger subjects, with this advantage persisting through the pre-vaccination timepoint in study year 2. Interestingly, no statistically significant differences were identified at the post-vaccination timepoint in study year 2, suggesting a possible ceiling effect upon repeated influenza exposures. When NP-specific responses were examined, a significantly increased magnitude of NP-specific CD4 T cells was quantified at the baseline visit in year 1 but not at any subsequent study visits despite no documented influenza infections in enrolled subjects.

Figure 2.

Older subjects have a higher percentage of CD69+IFNγ+CD4 T cells specific for H3 and HAB through the pre-vaccination timepoint in study year 2. Subjects were banded into two age groups, consisting of those <3 years and ≥3 years of age. Following stimulation, reactivity to the pH1, H3, HAB, and NP proteins was quantified at pre- and post-vaccination timepoints following identification of CD69+IFNγ+CD4 T cells by sequential gating as shown in Supplementary Figure 1. Statistical analysis was performed using a Wilcoxon rank-sum test to compare cytokine responses across age groups. Data are shown as box and whisker plots, with the box depicting the 1st and 3rd quartiles and the whiskers showing the 10th to 90th percentiles.

Finally, we examined the functionality of the influenza-specific CD4 T cell response by utilizing combination gating to quantify co-expression of IFNγ, IL-2, and TNF in H3- and HAB-specific CD4 T cells over time, comparing between younger and older children (Figure 3). Most differences between younger and older children were seen in the initial year of the study, with a general trend toward increases in cell populations able to produce IFNγ alone and IFNγ in combination with either IL-2 or TNF in older children. Interestingly, the population of older children also had a significant expansion of HAB-specific CD4 T cells expressing IFNγ, IL-2, and TNF simultaneously, with that functionality making up only a very small fraction of H3-specific CD4 T cells at any study timepoint.

Figure 3.

Older subjects have a higher proportion of CD4 T cells producing IFNγ alone and IFNγ in combination with either IL-2 or TNF, especially in study year 1. PBMCs were stimulated and stained as previously described, with gating as shown in Supplementary Figure 1. Combination gates were created to evaluate expression of IFNγ, IL-2, and TNF. Differences in cytokine expression between older and younger subjects were compared following stimulation of PBMCs with the H3 (A) and HAB (B) proteins. Data are shown as pie graphs, with arcs highlighting specific cytokine production (IFNγ: red, IL-2: orange, TNF: blue) and pie slices depicting the percentage of cells with specific cytokine expression patterns as indicated in the figure. The Wilcoxon rank-sum test was used to compare each cytokine combination between age groups, with P-values denoted by the number of stars (* = P-value between .01 and .05, ** = P-value between .0001 and .01, *** = P-value ≤ .001).

DISCUSSION

In this study, we evaluated the development of CD4 T cell-mediated anti-influenza immunity to the pH1, H3, and HAB proteins as well as NP in children between 6 months and 8 years of age following vaccination with IIV, with a focus on the evolution of CD4 T cell specificity and function. We found evidence that IIV establishes and boosts influenza-specific CD4 T cell-mediated immunity to the H3 and HAB proteins in even the youngest children. Older children had an overall greater magnitude of CD4 T cells producing IFNγ through the pre-vaccination visit in study year 2, with a shift toward CD4 T cells producing IFNγ and TNF or IL-2 as well as IFNγ alone in older as compared to younger children.

One interesting finding was the more robust response seen in children in year 1 compared to year 2 of this study, with the younger age group more closely resembling the older cohort post-vaccination in study year 2. While a negative impact from repeated vaccinations has been documented in multiple studies, including a negative impact on cellular immunity [13], most eligible children were also vaccinated in the 2015–2016 influenza season in our cohort. This makes it less likely that blunting of the year 2 post-vaccination response was solely due to the repeat vaccination in study year 2. An alternative possibility is that, with repeated influenza exposures over the first several years of life, children reach a steady state in their anti-influenza CD4 T cell reactivity, resulting in more similar response magnitudes and less differences in CD4 T cell functionality between the younger and older children over time.

The blunted response to the pH1N1 HA protein seen within our cohort was unexpected; however, we do not believe this was due to an issue with the pH1 peptide pool based on previous experience using these same peptides in studies in both mice and humans [9, 11, 13, 20, 21]. While H3N2 viruses predominated in 2016–2017 and 2017–2018, H1N1 viruses were the predominant viruses circulating in 2013–2014 and 2015–2016, so it is likely that some of our cohort was previously exposed to an H1N1 virus via an acute infection and had preexisting immunologic memory. It is of note that several previous studies have documented a less robust CD4 T cell response to the H1 HA protein [11, 22], with fewer available H1 epitopes identified in mouse models [20, 23]. Potentially lower H1 immunogenicity could be due to lower epitope availability or differences in antigenic processing, with this possibly accounting for the low CD4 T cell reactivity to the pH1 protein in our cohort of children.

One limitation of this study was the challenge in documenting influenza infections. It is probable that most older children had exposure to influenza via infection prior to study enrollment, although the date of this infection was largely unknown. However, as this reflects the natural history of childhood influenza exposures [24], it does not invalidate the age-based comparisons made in this study. The increase in NP-specific CD4 T cell reactivity between study year 1 and study year 2 in younger children could support the occurrence of undocumented influenza infections in a subset of children, especially in the younger age cohort. However, Fluzone is a split vaccine known to contain variable amounts of NP [18, 19], so some children may also have had increases in NP-specific CD4 T cell reactivity in response to vaccine administration [9, 12]. Unfortunately, due to young, healthy children frequently not seeking medical care for an uncomplicated influenza infection, we were unable to capture influenza infections in our cohort through passive surveillance via medical history or chart review. Serologic studies on this cohort of children are currently ongoing.

In addition, this study was also limited as only cytokine-dependent methods were used to measure antigen-specific CD4 T cell responses, with not all CD4 T cell functionalities (Th2, Th17, Treg) able to be quantified due to limitations in the panel size and the small blood volumes able to be obtained from young children. In adults, the influenza-specific immune response is Th1-biased, and our study enabled us to track the development of Th1-mediated immunity in children. Future studies either utilizing cytokine-independent measures to quantify antigen-specific CD4 T cells or a broader panel to track additional CD4 T cell functionalities would allow for a more unbiased evaluation of the early immune response in the pediatric population.

Our finding that early life vaccination can establish and subsequently boost CD4 T cell responses to the H3 and HAB proteins supports a role for IIV in helping to shape early life anti-influenza CD4 T cell reactivity. Early influenza exposures have a critical role in establishing immunologic memory, with infections in childhood able to bias subsequent pandemic [25] and seasonal [26] influenza immune responses. While existing data support a critical role for memory B cells established by acute infection in immune imprinting [27], the role of CD4 T cells induced by vaccination in establishing early life immunologic memory is uncertain [28]. Vaccination prior to an initial infection may be beneficial as it could prime a CD4 T cell response enriched for reactivity against multiple HA proteins. However, the peripherally primed CD4 T cells elicited by early vaccination could also lack key specificities against more conserved internal virion proteins and have poor lung homing potential. Further research is needed to more precisely understand the role of influenza vaccination in shaping the development of influenza-specific CD4 T cell memory in the pediatric population.

Supplementary Data

Supplementary materials are available at the Journal of The Pediatric Infectious Diseases Society online (http://jpids.oxfordjournals.org).

Data availability

All primary data will be made available by the corresponding author upon request.

Notes

Financial support. This work was supported by grant number 2015098 from the Doris Duke Charitable Foundation to J. L. N. and has been funded in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services, under CEIRS Contract No. HHSN272201400005C.

Potential conflicts of interest . J. L. N. receives research grant support from Pfizer Inc., Moderna Inc., and Merck Inc. to conduct clinical trials unrelated to the present study. No other authors have any conflicts of interest relevant to the publication of this paper.