Intermediate levels of preexisting protective antibody allow priming of protective T cell immunity against Influenza

Abstract

Overcoming interfering impacts of preexisting immunity to generate universally protective influenza A virus (IAV)-specific T cell immunity through vaccination is a high priority. Here, we passively transfer varied amounts of H1N1-IAV-specific immune serum prior to H1N1-IAV infection to determine how different levels of preexisting Ab influence the generation and protective potential of heterosubtypic T cell responses in a murine model. Surprisingly, IAV nucleoprotein (NP)-specific CD4 and CD8 T cell responses are readily detected in infected recipients of IAV-specific immune serum regardless of the amount transferred. When compared to responses in control groups and recipients of low and intermediate levels of convalescent serum, NP-specific T cell responses in recipients of high levels of IAV-specific serum, which prevent overt weight loss and reduce peak viral titers in the lungs, are, however, markedly reduced. While detectable at priming, this response recalls poorly and is unable to mediate protection against a lethal heterotypic (H3N2) virus challenge at later memory timepoints. A similar failure to generate protective heterosubtypic T cell immunity during IAV priming is seen in offspring of IAV-primed mothers that naturally receive high titers of IAV-specific Ab through maternal transfer. Our findings support that priming of protective heterosubtypic T cell responses can occur in the presence of intermediate levels of preexisting Ab. These results have high relevance to vaccine approaches aiming to incorporate and evaluate cellular and humoral immunity against IAV and other viral pathogens against which T cells can protect against variants escaping Ab-mediated protection.

Introduction

Influenza virus (IAV) remains a significant health concern causing millions of cases of severe illness globally despite widespread annual vaccination (1, 2). Current IAV vaccine formulations aim largely to stimulate the generation of neutralizing antibodies against the surface hemagglutinin (HA) and neuraminidase (NA) proteins of IAV strains predicted to circulate during the ‘flu’ season. Most neutralizing antibodies, however, offer limited protection against antigenically drifted strains or reassorted strains expressing different HA and NA proteins that have pandemic potential (3). A powerful way to improve protection against IAV is for vaccines to target conserved epitopes between disparate viral strains. Along this line, the generation of Ab recognizing conserved regions of the HA stalk (4) as well as the generation of cross-protective memory T cells (5–8) both have the potential improve vaccine efficacy.

IAV-specific memory T cells, in addition to recognizing the external HA and NA viral coat proteins, recognize the internal viral nucleoprotein (NP), acidic polymerase (PA), and basic polymerase (PB-1) proteins (5, 9). These proteins are highly conserved amongst IAV isolates as they are essential for the viral life cycle (10). NP, PA, and PB-1 specific T cell responses have been extensively characterized in murine models (9, 11, 12). At memory timepoints following primary infection or vaccination with live attenuated cold-adapted virus (LAIV) vaccines, IAV-specific T cell responses can protect against severe disease caused by IAV strains expressing HA and NA combinations not neutralized by infection- or vaccine-induced antibodies (13–15). This mode of T cell-dependent protection is known as ‘heterosubtypic’ immunity (16). Heterosubtypic immunity, mediated by IAV-specific CD4 or CD8 T cells alone, can protect against IAV infection independently of B cells (16–18). However, a concerted anti-viral response involving IAV-specific T as well as B cell responses (19–21), provides the most optimal and robust protection against infection. In line with murine findings, positive correlations between the magnitude of internal IAV protein specific T cell responses and the degree of protection against severe symptoms following controlled and seasonal IAV infections have been reported in human studies (22–24). Collectively, these findings support that targeting the stimulation of T cell responses through vaccination to generate heterosubtypic immunity could contribute to strong universal protection against IAV.

A potential hurdle to optimally stimulating the generation of universally protective T cell immunity through vaccination is the presence of preexisting IAV-specific humoral immunity (25–28), the levels of which vary greatly in the general population (29–31). By restricting infectivity and thereby reducing antigen loads in addition to altering inflammatory responses, preexisting neutralizing Ab can hinder effective generation of protective pathogen-specific T cell and B cell Ab responses (32). Exactly how different levels of preexisting IAV-specific Ab impact the ability to prime new protective T cell responses has not previously been explored. This is an important question to address given varied levels of IAV-specific Ab in the human population and the pressing need to increase IAV vaccine efficacy. Here, to gain quantitative insights into how preexisting humoral immunity impacts the depth and breadth of the T cell responses against subsequent homotypic IAV, we passively transferred high, intermediate, and low amounts of well-characterized convalescent immune serum to otherwise naive mice and assessed the priming of T cell responses induced by IAV infection. The relationship between preexisting humoral immunity impacts on T cell priming and the subsequent protection afforded during a memory recall response following heterosubtypic IAV challenge, which is a model of pandemic infection, is also examined.

We find that IAV nucleoprotein (NP)-specific CD4 and CD8 T cell responses are readily primed in all infected recipients of IAV-specific immune serum regardless of the amount transferred. However, priming of heterosubtypic T cell responses that can persist and mediate robust protection against heterologous viral challenge at memory timepoints only occurs in the presence of intermediate and low levels of preexisting IAV-specific Ab. These findings support that evaluation of preexisting humoral immunity prior to universal vaccine approaches that incorporate both cellular and humoral immunity against IAV could markedly improve vaccine efficacy. These findings also have relevance to vaccine approaches for other viral infections such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (33–37), against which T cells can protect against variants escaping Ab-mediated protection (38–41), and for individuals which have received passive immunotherapy or preexposure prophylaxis infusions of pathogen-specific Ab (42).

Materials and Methods

Mice

Adult C57BL/6 mice were used in experiments when 6 to 12 weeks old. C57BL/6 mice were obtained from Jackson Laboratories. In some experiments mice were bred in the vivarium of the University of Central Florida. All experimental animal procedures were conducted in accordance with guidelines outlined by the Office of Laboratory Animal Welfare (OLAW), National Institute of Health, USA. Protocols were approved by the Institutional Animal Care and Use Committee of the University of Central Florida (Orlando, FL).

Virus stocks and infections

A/Puerto Rico/8/34 (PR8), (H1N1) originating from stocks at St. Jude Children’s Hospital, and A/Philippines/2/82/x-79 (A/Phil), (H3N2), (kindly provided by S. Epstein, NIH) that contains the internal proteins of PR8 were grown in the allantoic cavity of embryonated hen eggs at the Trudeau Institute and the egg infective dose (EID₅₀), tissue culture infective dose (TCID₅₀), and lethal dose (LD₅₀) characterized. Mice were infected intranasally under light isoflurane anesthesia (Webster Veterinary Supply) with the stated doses of virus in 50 μl of PBS. Timed-pregnant female C57BL/6 mice generated within the University of Central Florida vivarium were infected with PR8 when in the first or second trimester. Pups from uninfected control and PR8-infected dams were weaned 21 days later and used in homotypic and heterotypic virus challenge experiments as indicated. Juvenile mice were infected with the stated doses of virus in 25 μL of PBS. All infected mice were monitored daily for percentage weight loss, hunched posture, ruffled fur, and lack of movement and humanely euthanized as described (43).

Generation, characterization, and administration of IAV-specific immune serum

Convalescent anti-PR8 serum was collected from sub-lethally infected animals at days greater than 30 days post infection. The concentration of PR8 specific IgG in convalescent serum, 2.5 mg/mL, was used to as a marker to quantify the level of pathogen specific Ab and was determined by ELISA as described below using purified anti-H1N1 specific IgG (Chemicon International) as a standard. PR8-specific IgG titers for HA, NA, Matrix protein external peptide (M2e), and NP in the convalescent serum were determined by ELISA. Plates were coated with 1 μg/mL of recombinant IAV proteins (Sino Biological) or 15 μg/mL of M2e peptide (Vivitide) as previously described (44, 45). The Hemagglutination Inhibition (HAI) titer of the convalescent serum, 160 HAI, was determined using previously described standard methods (46). Convalescent serum diluted to the indicated concentrations in PBS or undiluted naïve control serum in 200 μL was injected intraperitoneally (i.p.) 24 hours prior to PR8 infection to ensure optimal levels of passively transferred Ab were present in respiratory tissues (47).

The “take” or systemic level of the passively transferred PR8-specific IgG Ab detectable on day 4 post infection in recipient mice was determined by ELISA. Briefly, Nunc ELISA plates coated with UV-inactivated influenza PR8 were washed and blocked with PBS containing 2% BSA. Serum samples serially diluted in PBS-Tween 20 with 1% BSA were incubated 2 hr. at room temperature. After washing, HRP-conjugated Abs specific for mouse total IgG (Southern Biotech) were added at 0.2 μg/ml in PBS-Tween 20 with 1% BSA, and plates were incubated 1hr. at room temperature. After washing, the HRP substrate o-phenylenediamine dihydrochloride was added and the OD of the color reaction was measured at 492 nm. Endpoint serum titers were defined by the last serum dilution that gave an OD reading above 2 SD of the mean of conjugate blank readings.

Detection of IAV titer

At day 4 of primary or day 5 to 6 of secondary viral infection, mice were euthanized by cervical dislocation followed by exsanguination by perforation of the abdominal aorta. Lungs were harvested and flash frozen in liquid nitrogen and pulmonary viral titer determined by quantitation of viral RNA. RNA was prepared from whole lung homogenates using TRIzol (Sigma-Aldrich), and 2.5 μg of RNA was reverse transcribed into cDNA using random hexamer primers and Superscript II Reverse Transcriptase (Invitrogen). Quantitative PCR was performed to amplify the acidic polymerase (PA) gene of PR8 using the QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems) with 50 ng of cDNA per reaction and the following primers and probe: forward primer, 5’-CGGTCCAAATTCCTGCTGA-3’; reverse primer, 5’CATTGGGTTCCTTCCATCCA-3’; probe, 5’−6-FAM-CCAAGTCATGAAGGAGAGGGAATACCGCT-3’ (LGC BioSearch Technologies). Data were analyzed with QuantStudio 7 software. The copy number of the PA gene per 50 ng of cDNA was calculated using a PA-containing plasmid of known concentration as a standard. The number of copies of the PA gene is presented as the number of IAV copies per lung.

Detection of inflammatory cytokines and chemokines

Lungs were collected on 7 dpi and homogenized in RPMI-1640 supplemented with 2mM L-glutamine (Invitrogen), 100 international Units penicillin, 100 μg per mL streptomycin (Gibco) 10 mM HEPES (Gibco), 50μm 2-mercaptoethanol, and 7.5% FBS (Hyclone). Concentrations of cytokines and chemokines were determined using a Milliplex mouse cytokine/chemokine magnetic bead premixed kit (Millipore) read on a Bio-Plex 200 System (Bio-Rad) as previously described (48).

Tissue preparation and flow cytometry

At the indicated time points following virus infection, mice were euthanized by cervical dislocation, exsanguinated by perforation of the abdominal aorta, and lungs perfused by injecting 10 mL of PBS in the left ventricle of the heart. Lungs, spleen, and draining mediastinal lymph node (dLN) were prepared into single cell suspensions by mechanical disruption of organs and passage through a nylon membrane. Cell suspensions were washed, resuspended in FACS buffer (PBS plus 0.5% BSA and 0.02% sodium azide (NaN₃); Sigma-Aldrich) and incubated on ice with 1 μg anti-FcR (2.4G2) followed by saturating concentrations of the following fluorochrome-labeled Abs for surface staining: anti-Thy1.1 (OX-7), anti-Thy1.2 (53–2.1), anti-CD4 (RM4.5), anti-CD8 (H35–17.2), anti-CD44 (IM7), anti-CD45.2 (104), anti-CD11c (N418), anti-CD11b (M1/70), anti-NK1.1 (PK136), anti-CD3 (17A2), anti-MHC-II (M5/114.15.2), anti-Ly6G/C (RB6–8C5), and anti-Siglec F (E50–2440) (Pharmingen, eBioscience/ThermoFisher, or BioLegend). To enumerate IAV-specific CD8 T cells in IAV-infected mice, cells were stained for 1 hr. at room temperature with H2D^b NP_366–374 fluorochrome-labelled tetramer prior to staining for cell surface markers (49).

To visualize virally infected lung cells, single cell suspensions of were washed, resuspended in FACS buffer, and incubated on ice with 1 μg anti-FcR (2.4G2) followed by saturating concentrations of fluorochrome-labeled Abs for surface staining of CD45.2 (104), CD11c (N418), and MHC-II (M5/114.15.2) (Pharmingen, eBioscience/ThermoFisher, or BioLegend). Intracellular staining for IAV-NP (D67J, ThermoFisher) and prosurfactant C (EPR19839, Abcam), visualized with donkey anti-rabbit fluorochrome-labeled Ab (Poly4064, BioLegend) was performed as previously described (50) following fixation and permeabilization with the FOXP3 Transcription Factor/Fixation Permeabilization Concentrate and Diluent (Life Technologies, ThermoFisher).

ELISPOT Assay

ELISPOT analysis for detection of IFN-γ and IL-2 secreting T cells was performed as previously described (14). Briefly, 96-well plates (Millipore) were coated overnight with purified anti-mouse IFN-γ or IL-2 Ab. Plates were washed and blocked with complete T cell media prior to the addition of 10⁵ lung cells or 10⁶ spleen or lymph node cells and 10⁶ syngeneic APC from spleens of naïve mice. Wells received 10 μg of I-A^b NP_261–275 and NP_311–325 peptides to elicit CD4 T cell responses or 10 μg of H2D^b NP_366–374 and PA_224–233, or H2K^b PB_703–711 peptides to elicit CD8 T cell responses. Plates were incubated overnight with biotinylated detection Ab against IFN-γ or IL-2 followed by streptavidin alkaline phosphate substrate. Plates were developed with 5-bromo-4-chloro-indolyl phosphate substrate and resulting spots were counted on an ImmunoSpot reader (CTL).

Histology

For assessment of immunopathology following IAV challenge infection, lungs lobes were isolated and immediately fixed in 10% neutral buffered formalin on 5 dpi. Lung samples were subsequently processed, embedded in paraffin, sectioned at 5μm thickness, placed on L-lysine-coated slides, and stained with Hematoxylin and Eosin (H&E) using standard histological techniques. Stitched Z-stack images of whole lung sections at 20X magnification were acquired with the ImageXpress Pico Automated Cell Imaging System (Molecular Devices) and CellReporterXpress software (V 2.7).

Statistical analysis

Group sizes of n = 3 to 9 per experiment were employed, and experiments were repeated two to three times. For Unpaired, Students t-tests, ∝ = 0.05, were used to assess whether the means of two normally distributed groups differed significantly. One-way ANOVA analysis with the appropriate multiple comparison post-test, Bonferroni’s or Tukey’s, was employed to compare multiple means. Pearson’s correlation test was used measure associations between variables of interest. All error bars represent the standard deviation. Significance is indicated as * P < 0.05, ** P < 0.005, *** P < 0.001, **** P < 0.0001.

Results

Passive Ab transfer controls viral burden and prevents morbidity following IAV infection

To model preexisting anti-IAV humoral immunity in a manner that permits precise control of the amount of virus-specific Ab present in vivo, we passively transferred varied amounts of convalescent IAV immune serum to otherwise naive mice. The convalescent IAV serum used was collected from mice at 30 dpi of a sublethal PR8 infection. Such serum contains high titers of virus-specific IgM, IgG, and IgA antibody isotypes (51). To characterize the protective potential of the sera, we determined the level of PR8-specific IgG present as prior studies have shown that depletion of IgG from convalescent IAV serum abrogates its ability to passively transfer protection against homologous viral infection (51). A range of 0.25 to 250 μg of PR8-specific IgG Ab, which contains Ab reactive against PR8 external as well as internal proteins (Table 1 and Supplemental Fig 1), was passively transferred to otherwise naïve animals that were subsequently infected with a lethal (2 LD₅₀) dose of homologous PR8 virus.

Table 1.

Convalescent PR8-specific serum reactivity against IAV proteins

IAV Protein	Naïve serum AUC (SD)†	PR8-specific serum AUC (SD)
HA	4.80 × 102 (0.70)	1.73 × 105 (0.01)
NA	2.10 × 101 (0.03)	3.19 × 104 (0.02)
M2e	7.70 × 101 (1.40)	2.20 × 103 (0.24)
NP	6.70 × 103 (3.20)	1.83 × 105 (0.01)

^†Area under curve (AUC) of optical density readings at 492 nm and standard deviation (SD) of the mean.

At 4 dpi, a timepoint preceding the generation of endogenous PR8-specific Ab, we determined the “take” or the in vivo titer of the transferred PR8-specific Ab in the serum. No increase in PR8-specific Ab over baseline was detected in recipients of 0.25 or 2.5 μg of PR8-specific Ab, or in control recipients of serum from naive mice. However, passive transfer of serum containing 25 μg of PR8-specific IgG resulted in reciprocal endpoint Ab titers that, while low, were significantly higher than titers detected in recipients of control serum (Fig 1a). Markedly higher titers of transferred PR8-specific Ab were present in recipients of 250 μg of PR8-specific IgG (Fig 1a). We also determined the HAI units of the transferred serum to get an indication of the in vivo “take” of HA-reactive Ab. All recipients of PR8-specific immune serum had in vivo HAI titers clinically considered to be protective (greater than or equal to 40) (52, 53). Regardless of the amount transferred, the HAI units in recipients of convalescent serum was 80 whereas control recipients of naïve serum had unprotective titers of 20 HAI (data not shown).

Figure 1. Passive Ab transfer controls virus burden and morbidity following IAV infection.

C57BL/6 control and recipients of the indicated amounts of PR8-specific convalescent serum were infected with a lethal 2 LD₅₀ dose of PR8. On the fourth day post infection (dpi), serum and lungs were collected for determination of (a) in vivo PR8-specific Ab titers, (b) viral burdens in the lung, and (c) the relationship between the two parameters as determined by regression analysis (n = 5–7 mice per group per replicate experiment). Convalescent serum (anti-PR8) is shown in (a) as a positive control. (d) PR8-specific Ab recipients and control animals were monitored for morbidity by weight loss following infection up to 20 dpi and percent initial weight determined (n = 5 mice per group per replicate experiment). On 2 and 4 dpi, virally infected lung cells were visualized by flow cytometry (n = 4 mice per group per replicate experiment). (e) Representative gating of CD45.2⁻ cells and staining for Pro-SpC in the presence of both primary and secondary-fluorescently labeled Ab (black-lined histogram) or the secondary-fluorescently labeled Ab alone (grey-lined histogram). (f) Representative IAV-NP⁺ staining in Pro-SpC⁺ and Pro-SpC⁻ CD45.2⁻ lung cells and (g) frequency of IAV-NP⁺ cells at the indicated days post challenge.

To determine if passively transferred PR8-specific immune serum neutralizes virus in the lung, peak viral burdens were evaluated on 4 dpi. Mice receiving 25 μg of PR8-specific IgG had ~1 log reductions in viral titer compared to control mice or mice receiving lower amounts of immune serum. Viral copies detected in recipients of 250 μg of PR8-specific IgG were reduced by ~2–3 logs compared to controls (Fig 1b). Regression and correlation analysis revealed a significant negative, non-linear relationship (R² = 0.8212, P = 0.0052) between the in vivo reciprocal endpoint titer of the passively transferred serum and the viral burden detected at 4 dpi (Fig 1c). Interpolation of the curve revealed that the relative concentration of transferred PR8-specific IgG required to reduce the PR8 burden in the lung by 50% is 5 μg (IC₅₀).

In agreement with the protective in vivo HAI titers observed, all recipients of PR8-specific Ab were protected from a lethal 2 LD₅₀ homologous PR8 virus challenge although a range of morbidity, measured by weight loss, was observed (Fig 1d). Animals that received 250 μg did not lose any weight and were indistinguishable from uninfected controls despite the detection of roughly a million copies of virus in the lung, whereas recipients of 2.5 μg of PR8-specific IgG presented with significant weight loss prior to complete recovery.

We next evaluated whether recipients of 250 μg of PR8-specific IgG harbored any lung cells actively infected with virus to rule out that the viral copies detected above could be largely restricted to particles that had been neutralized by the transferred serum. On d2 post infection, recipients of naive and PR8-specific serum both had IAV-nucleoprotein (NP)⁺ cells within non-hematopoietic, CD45.2⁻, prosurfactant C⁺ type II alveolar epithelial cells (54, 55) as well as within CD45.2⁻, prosurfactant C⁻ lung cells, which includes type I alveolar epithelial cells (54) (Fig 1e–g). As anticipated, the frequency of non-hematopoietic lung cells positive for IAV-NP was significantly lower in the recipients of PR8-specific IgG versus in infected recipients of naïve serum (Fig 1f and g). The frequency of lung cells positive for IAV-NP was significantly higher in infected recipients of control naïve serum on 4 dpi versus 2 dpi but not in recipients of 250 μg of PR8-specific IgG (Fig 1g). These findings suggest that a subclinical and contained active viral infection occurs in permissive lung cells in the presence of high levels of preexisting, PR8-specific Ab. Collectively, these results demonstrate the well-documented protective impact of passively transferred immune serum during PR8 infection.

We next evaluated how the presence of preexisting IAV-specific Ab impacts the inflammatory milieu present in the lung at 7 dpi to determine whether and at which doses passively transferred anti-IAV IgG can reduce the ‘cytokine storm’ associated with severe infection (56). To do so, we assessed a broad sampling of inflammatory cytokines and chemokines in lung homogenates as in our previous studies (48). Unexpectedly, when compared to recipients of naïve serum, the levels of three analytes, IFN-γ, IL-1β, and KC, were significantly increased in the lungs of recipients of 0.25 μg of PR8-specific IgG (Fig 2), suggesting that low concentrations of Ab unable to completely neutralize virus can contribute to Ab-dependent enhancement of disease (57, 58, 59, 60, 61). Levels of inflammatory cytokines and chemokines in the lungs of recipients of 2.5 or 25 μg of PR8-specific IgG were either similar, increased without reaching significance, or gradually decreased versus levels in controls receiving naive serum. However, recipients of 250 μg of PR8-specific IgG displayed markedly reduced levels of all analytes detected versus controls and were comparable to uninfected animals (Fig 2). Together, these results demonstrate that IAV-induced inflammation is reduced only by the passive transfer of relatively high levels of immune serum.

Figure 2. Passive Ab transfer impacts lung inflammation following IAV infection.

C57BL/6 control and recipients of the indicated amounts of PR8-specific convalescent serum were infected with a lethal 2 LD₅₀ dose of PR8. Lung homogenates from all groups of mice were collected on 7 dpi and assessed for inflammatory cytokines and chemokines (n = 3 mice per group per replicate experiment). Dashed lines in graph panels represent basal levels of analytes present in naïve mice.

The generation of anti-viral T cell responses in the presence of preexisting IAV-specific Ab

To evaluate how preexisting IAV-specific Ab impacts the generation of anti-viral T cell responses, total and activated T cells (Fig 3), as well as internal IAV protein-specific T cells (Fig 4) were characterized in PR8-infected recipients of 2.5 μg of PR8-specific IgG (half of the IC₅₀ of immune serum required to reduce viral burdens by 50%) or of 250 μg of PR8-specific IgG (which prevents weight loss entirely). We assessed T cell responses at 10 dpi, their peak, versus responses in control mice receiving naive serum.

Figure 3. High levels of preexisting IAV-specific Ab reduce anti-viral T cell responses in the lung.

C57BL/6 control and recipients of the indicated amounts of PR8-specific convalescent serum were infected with a lethal 2 LD₅₀ dose of PR8. On 10 dpi, (a) the total number of lymphocytes, as well as the (b) frequency of CD44^hi T cells present in the spleen, dLN, and lung were determined. (c) Representative CD44 surface marker expression and gating on CD4 and CD8 T cells in the different organs of the infected animals and (d and e) enumeration of the total number of CD44^hi CD4 and CD8 T cells (n = 5 mice per group per replicate experiment). Significant differences between control and PR8-specific Ab recipients in the spleen (top bar) and lung (lower bar) are indicated.

Figure 4. High levels of preexisting IAV-specific Ab reduce NP-specific CD4 and CD8 T cell responses.

C57BL/6 control and recipients of the indicated amounts of PR8-specific convalescent serum were infected with a lethal 2 LD₅₀ dose of A/PR8. On 10 dpi, (a) IFN-γ (b) and IL-2 CD4 T cell responses were enumerated in the stated organs by ELISPOT using of I-A^b NP_261–275 and NP_311–325 peptides to stimulate the cells (n = 5–7 mice per group per replicate experiment). On 10 dpi, (c) H-2D^b NP_366–374-specific CD8 T cells were enumerated using tetramer staining and flow cytometry (n = 3 mice per group per replicate experiment). (d) Representative tetramer staining of the different groups of mice as well as naïve controls (nC). (e) Regression analysis of the observed weight loss following infection in the different groups and the number of NP_366–374-specific CD8 T cells present in the different organs.

We first enumerated total lymphocytes in the spleen, dLN, and lungs. Total numbers of lymphocytes in the spleens and dLN were similar across all groups (Fig 3a). In contrast, recipients of 250 μg of PR8-specific IgG had lower numbers of lymphocytes in the lung compared to controls receiving naïve serum (Fig 3a). Total T cells expressing high levels of the activation marker CD44 were also similar across groups in the spleen and dLN but were only reduced two-fold in the lungs of recipients of 250 μg PR8-specific IgG compared to recipients of 2.5 μg of PR8-specific IgG or of naive serum (Fig 3b). When T cells were separated into CD44^hi CD4 and CD8 T cells (Fig 3c) and cells enumerated, both CD4 and CD8 T cells were reduced in the lung and dLN of the recipients of 250 μg PR8-specific IgG versus the other groups (Fig 3d and e).

We next assessed NP-specific T cell responses on 10 dpi by ELISPOT to enumerate CD4 T cells, and tetramer staining and flow cytometry to enumerate CD8 T cell responses. IFN-γ production elicited by the H-2^b NP_261–275 and NP_311–325 IAV peptides was readily detected in the spleen, dLN, and lung of all infected animals, but the responses were reduced in recipients of 250 μg of PR8-specific IgG compared to the other groups (Fig 4a). Recipients of 2.5 μg of PR8-specific IgG had similar IFN-γ responses to control mice in the spleen and draining lymph node but slightly reduced NP_261–275-specific responses in the lung. To detect a potentially broader array of NP-specific CD4 T cells, we also assessed peptide-dependent IL-2 responses. IL-2 production in response to both NP peptides was comparable across all groups in the dLN (Fig 4b). Interestingly, splenic NP_261–275 but not NP_311–325-dependent IL-2 spots were reduced in recipients of 250 μg of PR8-specific IgG, while in the lung NP_311–325-dependent but not NP_261–275-dependent spots were reduced in the presence of 250 μg of PR8-specific IgG. These results indicate reductions in the magnitude of anti-viral CD4 T cell responses by high levels of preexisting antibodies, but not by lower levels that are still able to confer protection.

H-2D^b NP_366–374-specific CD8 T cell responses in the spleens, dLN, and lungs of infected recipients of 250 μg of PR8-specific IgG were also reduced (Fig 4c and d). NP_366–374- as well as PA_224–233-, and PB_703–711-specific CD8 T cells are widely accepted to be robust IFN- γ producing cells (12, 62), which was verified by ELISPOT (data not shown). Interestingly, while low in number, NP_366–374-tetramer⁺ CD8 T cells were still readily detected in recipients of 250 μg of PR8-specific IgG and were increased roughly 40-fold versus in comparison to uninfected animals (Fig 4d). Regression and correlation analysis of the number of H-2D^b NP_366–374-specific CD8 T cells in the spleens, dLN, and lungs versus observed weight loss on 10 days post infection revealed significant non-linear relationships between the two parameters in the spleen (R² = 0.9128, P = 0.0002) and lung (R² = 0.7766, P = 0.0015), but not dLNs (R² = 0.4719, P = 0.4180) (Fig 4e). The latter observation indicates that the magnitude of H-2D^b NP_366–374-specific CD8 T cell response generated is directly associated with weight loss and the antigen burden.

Protective capacity of heterosubtypic T cell memory generated in the presence of preexisting Ab

Whether anti-viral T cell immunity generated during PR8 infection in the presence of preexisting PR8-specific antibodies optimally functions during recall is not clear. To evaluate this, we treated unprimed mice with varied concentrations of PR8-specific IgG and then primed the mice with a sublethal 0.5 LD₅₀ dose of PR8. At 30 dpi all mice were challenged with a lethal (150 LD₅₀) challenge of the heterotypic A/Philippines/2/82/x-79 (A/Phil, H3N2) virus. Robust heterosubtypic immunity, as indicated by minimal weight loss and survival, was seen in mice that received 0.25 or 2.50 μg of PR8-specific IgG Ab during priming. Recipients of 25.0 μg were also fully protected but did lose more weight during the first 4 dpi than mice that received 0.25 or 2.50 μg of Ab during priming. In contrast, the recipients of 250 μg of PR8-specific IgG prior to priming, which did not display signs of morbidity during the homologous infection, were not protected from heterotypic virus challenge and mirrored unprimed mice challenged with A/Phil (Fig 5a and b) despite having detectable lung IAV-specific T cell responses at 10 dpi (Fig 5c and d). While NP-specific CD8 T cells were detectable, albeit reduced in recipients of 250 μg of PR8-specific IgG post priming, this pattern of morbidity and mortality indicates that high levels of preexisting Ab during primary IAV infection compromises the generation of T cell immunity that can protect against future heterotypic IAV threats.

Figure 5. Protective heterosubtypic IAV-specific T cell memory is generated in the presence of intermediate levels of preexisting Ab.

C57BL/6 control and recipients of the indicated amounts of PR8-specific convalescent serum were infected with a sublethal 0.5 LD₅₀ dose of PR8 (H1N1). At a memory timepoint, 30 dpi, animals were challenged with a lethal 150 LD₅₀ dose of A/Phil (H3N2) and (a) morbidity determined by percent initial weight, and (b) mortality monitored (n = 5–7 mice per group per replicate experiment). (c and d) Enumeration of H-2D^b NP_366–374-specific CD8 T cells 10d post PR8 priming by tetramer staining and flow cytometry (n = 3 mice per group per replicate experiment). On 5dp A/Phil heterosubtypic viral challenge, groups of mice were assessed (e) for control of virus in the lung, (f and g) H-2D^b NP_366–374-specific CD8 T cell responses, and (h) IFN-γ and IL-2 recall responses in the indicated organs enumerated via ELISPOT following stimulation with NP CD4 and CD8 peptides (n = 5–7 mice per group in replicate experiments). Significant differences between control and PR8-specific Ab recipients in (g) the spleen (top bar) dLN (middle bar) and lung (lower bar) and in (h), the spleen (top bar) and lung (lower bar) for IFN-γ and the spleen for IL-2 are indicated.

We next assessed viral titers at 5 days post A/Phil challenge. Mice receiving 2.5 μg of PR8-specific IgG during PR8 priming effectively controlled virus by several logs compared to unprimed mice challenged with A/Phil (Fig 5e). In contrast, mice receiving 250 μg of PR8-specific IgG during PR8 priming, which were not protected against morbidity and mortality, only marginally controlled virus compared to unprimed controls (Fig 5e).

To ascertain whether compromised protection against A/Phil in mice receiving 250 μg of PR8-specific IgG during PR8 priming correlates with reductions in heterosubtypic T cell responses, NP-specific CD8 T cell responses as well as the combined magnitude of IAV-specific memory CD4 and CD8 T cell IFN-γ recall response was determined. Indeed, relatively weak recall responses were detected in the spleen and lungs of recipients of 250 μg of PR8-specific IgG on day 5 post A/Phil challenge (Fig 5f–h). IAV-specific memory CD4 T cell IL-2 responses were similarly reduced in the spleens of recipients of 250 μg of PR8-specific IgG (Fig 5h) but were similar between groups in the lung and dLN.

To gain additional insights into the compromised recall responses described above, we subsequently evaluated innate immune cell influx into the lung in unprimed and PR8-primed recipients of PR8-specific or naïve serum following A/Phil challenge. On 5d post challenge with A/Phil, both unprimed and primed recipients of 250 μg of PR8-specific IgG had elevated numbers of lung NK cells, neutrophils, and monocytes (Fig 6a and b), all of which have been associated with lung injury and immunopathology following IAV infection (18, 63–65). In contrast, significantly lower numbers of NK cells, neutrophils, and monocytes were detected in the lungs of PR8-primed recipients of control naïve serum at this timepoint. Lung eosinophils, alveolar macrophages, and dendritic cells were similar between all experiential groups (Fig 6a and b). Further evaluation of lung histopathology by H&E staining 5d post challenge showed that PR8-primed recipients of PR8-specific serum had diffuse lymphocytic infiltrates that mirror the infiltrates seen in unprimed animals more than the focal perivascular and peribronchial lymphocytic infiltrates seen in PR8-primed recipients of naïve serum (Fig 6c). Together, these observations support that optimal priming of protective heterosubtypic T cell immunity against IAV infection is compromised in the presence of high levels of pre-existing Ab.

Figure 6. Impaired generation of heterosubtypic immunity in the presence of preexisting Ab is associated with immunopathogenic innate immune responses following challenge infection.

C57BL/6 control and recipients of the indicated amounts of PR8-specific convalescent serum were infected with a sublethal 0.5 LD₅₀ dose of PR8 (H1N1). At a memory timepoint, 30 dpi, unprimed and primed animals were challenged with a lethal 150 LD₅₀ dose of A/Phil (H3N2) and (a) lung NK cells, neutrophils, eosinophils, alveolar macrophages (MQ), conventional dendritic cells (cDC), and monocytes enumerated on 5dp challenge (n = 3–4 mice per group per replicate experiment). The dashed line in graph panels represents enumeration of innate immune cells in uninfected, naïve animals. Representative staining and gating is shown in (b). In separate experiments, lungs were harvested on 5d post challenge and H&E-stained sections imaged to visualize histopathological changes (n = 3–4 mice per group per replicate experiment). Representative images in (c) are stitched 20X magnification Z-stacks of whole lung sections and the corresponding 7X zoom of indicated regions.

Maternally transferred Ab compromises generation of protective heterosubtypic T cell immunity

Newborn mice receiving maternal Ab transfer from IAV-primed mothers are a translationally relevant model where neutralizing Ab against IAV is present prior to other components of IAV-primed immunity. We thus tested whether the levels of PR8-specific Ab transferred naturally through maternal transfer from PR8-primed mothers similarly hinders the generation of protective heterosubtypic T cell immunity following PR8 infection. To do so, gravid dams in their first trimester were either mock infected or infected with 0.5 LD₅₀ PR8. The PR8-infected dams steadily gained weight, at a rate only slightly delayed compared to uninfected dams until parturition, while non-gravid infected females showed a typical weight loss pattern following infection (Fig 7a). Post-wean, juvenile (21-day old) offspring from uninfected and PR8-infected dams were challenged with 0.25 LD₅₀ PR8 and were monitored for weight loss and recovery. Juvenile mice weaned from PR8-infected dams that received PR8-specific Ab naturally through maternal transfer, which have high PR8-specific IgG titers that are comparable to maternal titers (Fig 7b), continued to gain weight until an adult weight was reached, with male mice gaining more weight than female mice (Fig 7c). Juvenile mice weaned from uninfected dams showed a typical weight loss pattern following infection with recovery of weight commencing on 10 dpi. These results indicate the strong protective impact of maternally transferred Ab.

Figure 7. Impaired generation of heterosubtypic IAV-specific T cell immunity in the presence of maternally transferred Ab.

C57BL/6 unmated females (non-gravid) and timed-pregnant dams (gravid) were mock infected or infected with a sublethal 0.5 LD₅₀ dose of PR8 (H1N1). Time-pregnant dams were in their first trimester at the time of infection. (a) Morbidity evaluated by percent initial weight was monitored to 20 dpi and arrows indicate parturition for gravid dams (n = 5 per group per replicate experiment). Offspring from unprimed or PR8-primed dams were weaned at 21 post-natal days and separate groups evaluated for (b) PR8-specific Ab titers (n = 3 per group per replicate experiment), and (c) changes from initial weight following infection with homologous PR8 virus (n = 9 per group per replicate experiment, and maternal Ab (MatAb)). At a memory timepoint, 30 dpi, PR8-primed offspring and unprimed controls were challenged with a 150 LD₅₀ dose of A/Phil (H3N2), and (d) morbidity and (e) control of virus in the lung determined (n = 3 to 5 per group per replicate experiment). The PR8-specific Ab titer of control naïve serum is represented as a dashed line in panel (b).

Thirty days post PR8 challenge, juvenile mice weaned from uninfected and PR8-infected dams, as well as naïve unmanipulated controls, were challenged with a lethal dose of A/Phil and morbidity and survival monitored. In agreement with our observations from mice receiving high levels of passively transferred PR8-specific IgG prior to priming, offspring harboring maternally acquired PR8-specific Ab succumbed to infection in a manner like unprimed controls. In contrast, mice weaned from uninfected dams that as juveniles lost weight during PR8 infection were completely protected from A/Phil challenge (Fig 7d). Finally, we asked if the compromised heterosubtypic immunity in mice weaned from PR8-primed mothers correlates with impaired A/Phil clearance. Indeed, while PR8-primed animals weaned from uninfected mothers cleared virus by 6 dpi, the offspring of PR8-primed dams did not control A/Phil titers (Fig 7e). Together, these results indicate that maternally transferred IAV-specific Ab in mice interferes with the priming of heterosubtypic IAV immunity in juvenile offspring. Furthermore, we find that such maternal Ab interference remains evident at 8 and 12 wks., timepoints when titers are expected to be at least half of their original level due to decay (data not shown) (66).

Discussion

The priming of universally protective immunity against IAV through vaccination is a high priority. An attractive means to generate such immunity is through the induction of IAV-specific memory T cells that have been shown in animal and in clinical studies to mediate robust protection against IAV-induced disease (8, 13–16, 67). A major hurdle for stimulating robust T cell memory through vaccination is the presence of preexisting pathogen-specific immunity (68, 69), which can likewise hinder the capacity to stimulate the generation of Ab responses from B cells recognizing novel viral epitopes (25, 70).

Following passive transfer of differing amounts of well-characterized anti-A/Puerto Rico/8/1934 (PR8, H1N1)-specific immune serum to otherwise naïve mice, we find the virus-specific Ab titer detected in vivo positively correlates with protection against subsequent homotypic PR8 infection. While reduced compared to mice not receiving antibodies, PR8-specific T cell responses are primed in recipients of transferred IAV-specific immune serum sufficient to prevent overt weight loss and reduce viral burdens in the lungs by two to three logs. However, PR8-primed T cell responses generated in the presence of high levels of anti-PR8 antibodies fail to protect the mice from a lethal heterotypic (H3N2) IAV challenge that is readily cleared in controls and recipients of intermediate and low levels of Ab during PR8-priming. We observed a similar loss in the ability to generate protective heterosubtypic T cell immunity during PR8 priming in the presence of high levels of preexisting humoral immunity received naturally through maternal transfer.

Collectively, these findings support the concept that the generation of protective T cell immunity against IAV, as previously reported in children and in various animal models following vaccination or infection (71–75), is impaired by high levels of preexisting Ab. Nevertheless, and importantly as we report here, protective responses can be primed in the presence of intermediate and low levels of Ab still sufficient to provide a level of homotypic protection. Our data supports that vaccine regimes targeting both cellular and humoral immunity against viral pathogens like IAV and SARS-CoV-2 (33–35), against which pathogen-specific T cells can protect against variants escaping Ab-mediated protection (38–41, 76), can provide balanced protection.

Surprisingly, NP-specific T cell responses were readily detectable at the peak of the primary immune response against IAV in all recipients of transferred PR8-immune serum. For example, NP_366–374 specific CD8 T cells were detected at frequencies roughly 40-fold higher than that of the naïve state in recipients of high levels of PR8-specific IgG suggesting that sufficient viral antigen is still present and able to prime anti-viral T cells despite virtually no outward signs of disease in the infected mice. These findings support the concept that even low exposures to infectious IAV can stimulate the generation of anti-viral T, as well as B cell humoral responses (77–79). Whether these responses optimally survive to memory and mediate protection against heterologous virus challenge is not clear from observations reported in prior literature. We show here during heterotypic virus challenge, mice primed in the presence of the highest amount of IAV-specific IgG, which had 2–3 log reductions in peak viral burdens in the lung, have relatively weak CD4 and CD8 T cell recall responses compared to mice receiving intermediate IAV-specific IgG Ab through passive transfer. Furthermore, this compromised memory response is unable to control a heterotypic virus challenge that is controlled by mice primed in the presence of lower amounts of protective, preexisting Ab. These results highlight the importance of challenge experiments in the evaluation of vaccine-induced T cell immunity to accurately determine the magnitude and protective potential of primed T cells.

We previously reported the existence of a ‘memory checkpoint’ operating from days 5 through 7 after IAV infection when receipt of TcR and ultimately of IL-2 signals are required for activated CD4 T cells to survive contraction and form stable memory populations (80, 81). We speculate that when high levels of preexisting Ab are present, the amount of viral antigen available during the memory checkpoint, while able to support relatively robust expansion of cells initially, is not sufficient to support optimal memory T cell formation. While the timing of the receipt of signals has yet to be defined, integration of inflammatory and TcR signals can also impact the ability of CD8 T effector cells to transition to memory cells (82–85). Mechanistically, high levels of antigen are required to overcome Bim-induced pro-apoptotic signals in antigen-specific effector CD8 T cells in the absence of inflammatory signals (83). It is thus reasonable to speculate that the presence of high levels of preexisting Ab, which prevent logarithmic growth of the virus, as well as inflammatory responses in the lung, precludes the ability of memory precursor IAV-specific CD8 T cells to survive and persist. Further studies are needed to probe whether these and other potential pathways regulating memory generation contribute to weak heterosubtypic immunity and whether increasing antigen availability during the memory checkpoint can increase the strength of T cell memory primed in the presence of preexisting humoral immunity.

While further exploration is required as to how preexisting humoral immunity impacts the ability to stimulate the generation of T cell responses, how it interferes with the stimulation of de novo Ab responses has been widely studied (25, 86, 87). The magnitude, the neutralizing capacity, as well as the depth of the reactivity of new IAV-specific Ab responses can all be impacted by prior pathogen exposure (29, 32, 87, 88), with preexisting Ab levels differentially affecting outcomes for B cell responses in a manner similar to that reported here for T cell responses. Most striking in adults is the profound effect of preexisting Ab levels on the reactivity or epitope dominance of the Ab repertoire generated following vaccination with universal Ab vaccines (87). In healthy individuals with high preexisting IAV-specific Ab titers, newly generated Ab responses tend to favor the generation of repertoire with reactivity dominance against epitopes in the HA-globular head, which can vary significantly among seasonal isolates. Whereas healthy individuals with low preexisting IAV-specific Ab titers tend to favor the generation of an Ab repertoire with dominant specificity for the more conserved epitopes in the HA stalk that can confer broad reactivity following vaccination (87). These findings in conjunction with ours support that the generation of IAV-specific immunity with universal protective capacity mediated either T or B cells can be achieved in the presence of low or intermediate levels of preexisting IAV humoral immunity.

For neonates and infants that are too young to be vaccinated due to the immature status of their immune system, the transfer of maternal Ab offers important immunological protection against pathogens (89, 90). Indeed, vaccination of expectant mothers during the third trimester is being explored as a potential means to protect highly susceptible infants from serious respiratory virus infection with IAV, RSV, as well as SARS-CoV-2 (91–93). Maternal Ab interference with subsequent infant vaccination is anticipated, yet there are conflicting observations in support of as well as in opposition of this concept (94–102). Our observations suggest that the level of preexisting maternal Ab may be a key determinant in the subsequent vaccination outcomes in infants. When the level of IAV-specific maternal Ab transferred is high, as is generated by natural infection of pregnant dams, robust and long-lived homologous immunity is evident in the offspring. However, much like the observations obtained with the passive transfer of IAV-specific Ab to adult mice, this high level of maternal IAV-specific Ab hinders the generation of protective heterosubtypic T cell responses in young mice, which we show can indeed be primed in this cohort as others have also shown (103). It is thus plausible that maternal Ab reduction of antigen loads and the concomitant alteration of inflammatory responses at the memory checkpoint similarly reduces the ability to optimally prime protective memory T cells in juvenile offspring.

In summary, our findings thus suggest that improving the robustness and longevity of vaccine-induced T cell immunity in all age groups requires a thorough quantitative understanding of how preexisting immunity, induced by infection, vaccination, maternal transfer, or introduced clinically via therapeutic intervention, impacts the antigen and inflammatory signals received by responding T cells.

Key Points.

1. A range of transferred IAV-specific Ab protects mice from lethal homotypic IAV.

2. Detectable IAV-specific T cell priming occurs in the presence of preexisting Ab.

3. Protective T cell immunity decreases with increasing preexisting IAV-specific Ab.