Seasonal Influenza Vaccination Does Not Effectively Expand H2 Cross-Reactive Antibodies in Humans

Abstract

Influenza viruses cause millions of infections and hundreds of thousands of deaths every year. Influenza virus vaccinations are produced every year and contain H1N1 and H3N2 influenza A viruses and either one or two influenza B viruses. In this study, we examined the effects of seasonal influenza vaccinations in people on both circulating serum antibody titers and memory B-cell activation to H2Nx influenza viruses. In addition to evaluating the human cohort as a whole, participants were also divided into three separate groups based upon their likelihood of being either exposed to or imprinted with H2N2 influenza viruses in the 1950s and 1960s. While participants born after H2N2 influenza viruses left the human population had lower HAI, ELISA and neutralizing antibody titers to these viruses, a select number of cross-reactive antibodies to some of the H2 HA proteins were boosted after seasonal influenza vaccination. However, these results varied and were not consistent by age group or specific H2 HA protein. Overall, seasonal influenza vaccination did not significantly expand cross-reactive antibodies to H2 HA antigens.

Introduction

Influenza is a seasonal respiratory virus that infects 25–50 million people in the United States every year (1). Yearly influenza vaccination elicits antibody responses to antigenically diverged seasonal influenza viruses defined as H1N1, H3N2 and type B. These yearly, seasonal vaccinations have a well-documented ability to elicit antibodies to each of the strains in the vaccine, as well as some antigenically distinct strains of the same subtype (2). Immune responses to the influenza virus subtypes within the vaccine has been examined previously (2–4). However, the ability of seasonal influenza virus vaccines to elicit antibodies to other influenza virus subtypes has not been well studied. This question is of particular importance for the H2N2 influenza viruses.

H2N2 influenza viruses caused the 1957 influenza pandemic. These novel H2N2 influenza viruses were the result of a reassortment event between a human H1N1 influenza virus and an avian H2N2 influenza virus (1). These novel H2N2 influenza viruses contained the HA, NA, and PB1 genome segments from the avian H2N2 influenza viruses and the other five genome segments from human H1N1 influenza viruses (5). This influenza pandemic caused an estimated one to two million deaths worldwide (5). Given that H2N2 influenza viruses have caused a previous pandemic in humans, it is likely that a future pandemic may be caused by an H2N2 influenza virus.

In 1960, Thomas Francis first described the term ‘original antigenic sin’ when comparing immune responses of different aged individuals to influenza viruses isolated in different decades (6). Francis theorized that an individual’s first infection with an influenza virus had an impact on their immune response to future influenza virus infections. Specifically, Francis hypothesized that first individuals would retain more B-memory cells from their first influenza virus infection than from subsequent influenza virus infections on the same subtype. Recently, the works of Gostic et al. have shown that ‘original antigenic sin’ or ‘immune imprinting’ of influenza viruses greatly affects a person’s immune response to heterologous influenza subtypes (7, 8). While most individuals alive today have been imprinted with either H1N1 or H3N2 influenza viruses, a significant subset of the human population was first infected with H2N2 influenza viruses in the 1950s and 1960s.

While previous studies have shown the lack of H2 influenza virus specific antibodies in individuals born after 1968 (9), no studies have examined the ability of seasonal influenza virus vaccinations to either generate de novo H2 influenza virus specific antibody responses or expand cross-reactive antibodies as a higher percentage of the active B-cell population. In this study, individuals likely imprinted with H1N1 influenza viruses, but still exposed to H2N2 influenza viruses (1934–1952) as well as individuals likely imprinted with H2N2 influenza viruses (1953–1964) and individuals unlikely to ever had been exposed to H2N2 influenza viruses (1965–1999) were compared. The dates were chosen based upon the probability of an individual born in a certain year having their first influenza virus infection being an H2N2 virus (7).

Using serum antibodies, the three groups within the human cohort were evaluated for both total antibody binding to HA and neutralizing antibodies against the virion. Following seasonal influenza virus vaccination, there was some increases in H2 influenza cross-reactive antibodies in both the hemagglutination inhibition (HAI) and ELISA assays in the different age groups within the human cohort. However, HAI and ELISA titers to other H2 HA proteins decreased after seasonal influenza vaccination. Additionally, peripheral blood mononuclear cells (PBMCs) were analyzed both before and after seasonal influenza vaccination for each group. Seasonal influenza vaccination did not stimulate memory B cells specific to H2 influenza in any of the groups. The data presented here suggests that seasonal influenza vaccination does not significantly increase cross-reactive antibodies to the majority of H2Nx influenza viruses regardless of the individual’s previous exposure to H2N2 influenza viruses.

Materials and Methods

Participants and vaccinations

Participants in this study were between the ages of 19 and 82 years old, consented to the study and enrolled in Athens, GA, USA. Eligibility was determined by a number of factors including those who had not yet received the seasonal influenza virus vaccine at the time of enrollment at the beginning of September in 2016. Influenza strains included in the vaccine were based upon the WHO recommendations for the Northern hemisphere: (A/California/7/2009–H1N1), (A/Hong Kong/4801/2014-H3N2), (B/Phuket/3073/2013-Yamagata-lineage), (B/Brisbane/60/2008-Victoria-lineage). Vaccinations of participants occurred between September-December 2016. Participants were vaccinated with the standard dose (15 μg/antigen) split-virion (IIV) version of licensed Fluzone (Sanofi Pasteur) influenza virus vaccine.

148 participants were enrolled in the study (Table 1). Approximately 80mL of blood was collected from each participant before vaccination (D0), 7 days post infection (D7), and 21 days post-infection (D21). Sera and peripheral blood mononuclear cells (PBMCs) were isolated from the blood samples. Sera was collected in Vacutainer serum separation tubes (SST) tubes (BD Biosciences) and processed within 48 hours, aliquoted and stored at −20°C. PBMCs were collected in Vacutainer cell preparation tubes (CPT) tubes (BD Biosciences) at D0, D7 and D21. PBMCs samples suspended in DMSO and FBS and were stored in liquid nitrogen.

Table 1:

Demographics of Human Participants A total of 138 individuals with 33 being male and 105 being female. There were 35 participants born from 1934–1952, 22 participants were born from 1953–1964 and 81 participants were born from 1965–1996. Caucasian participants numbered 110, Hispanics numbered 15, African Americans numbered 8 and the remaining 5 participants were either Asian, Pacific Islander or chose not to answer.

	Age Groups
Subject (#)	Total	1934–1952	1953–1964	1965–1996
Sex	Female	105	26	18	6110
	Male	33	9	4	20
Ethnicity	Caucasian	110	33	21	56
	Hispanic	15	1	0	14
	African American	8	0	1	7
	Asian	3	3	0	0
	Pacific Islander	1	1	0	0
	Did not repsond	1	0	0	1

Viruses, rHA antigens and VLPs

A/Chicken/Potsdam/4705/1984 (Chk/Pots/84), A/Chicken/PA/298101–4/2004 (Chk/PA/04), A/Duck/Hong Kong/273/1978 (Duk/HK/78), A/Mallard/Minnesota/AI08–3437/2008 (Mal/MN/08), A/Swine/Missouri/4296424/2006 (Sw/MO/06), A/Formosa/313/1957 (For/57), A/Japan/305/1957 (J/57), and A/Taiwan/1/1964 (Tw/64) were obtained from either the United States Department of Agriculture’s (USDA) Diagnostic Virology Laboratory (DVL) in Ames, Iowa, BEI resources, or provided by the laboratory of Dr. S. Mark Tompkins in Athens, GA. The H1N1 influenza viruses used in the study A/South Carolina/1/1918 (SC/18), A/Weiss/JY2/1943 (Weiss/43), A/Fort Monmouth/ 1/1947 (FM/47), A/Denver/1/1957 (Den/57), A/New Jersey/8/1976 (NJ/76), A/USSR/90/1977 (USSR/77), A/Brazil/11/1978 (BR/78), A/Chile/1/1983 (CL/83), A/Singapore/6/1986 (Sing/86), A/Texas/36/1991 (TX/91), A/Bejing/262/1995 (Bej/95), A/New Caledonia/20/1999 (NC/99), A/Solomon Islands/59/2006 (SI/06), A/Brisbane/59/2007 (Bris/07), A/California/07/2009 (CA/09; pandemic) and A/Michigan/45/20015 (MI/15; pandemic) were provided by either the Centers for Disease Control and Prevention (CDC) or Virapur LLC. Each virus was passaged using embryonated chicken eggs. Each virus was harvested from the eggs and aliquoted into tubes which were stored at −80°C. Each virus was tittered using a standard influenza plaque assay.

Recombinant HA (rHA) proteins were produced using the pcDNA 3.1+ plasmid. Each HA gene was truncated by removing the transmembrane (TM) domain and the cytoplasmic tail at the 3’ end of the gene. The TM domain was determined using the TMHMM Server v. 2.0 website: http://www.cbs.dtu.dk/services/TMHMM/. The HA gene was truncated at the first amino acid prior to the TM domain. A fold-on domain from T4 bacteriophage, an Avitag and a 6X histidine tag totaling 477 nucleotides were added to the 3’ end of the HA gene. The pcDNA 3.1+ vectors were then transfected individually into HEK293T suspension cells using ExpiFectamine 293 transfection reagent following manufacturer’s specifications (ThermoFisher Scientific). The supernatants were then harvested from the transfected HEK293T cells. Each rHA was then purified from the supernatant using a nickel-agarose column. The rHAs were then eluted from the column using imidazole. After elution, proteins were quantified using bicinchoninic assay (BCA) and stored at −80°C. Recombinant HA proteins produced for this study were A/Mallard/Netherlands/13/2001 (Mal/NL/01), J/57, A/Mallard/Wisconsin/08OS2844/2008 (Mal/WI/08).

For the virus-like particle (VLP) production, human endothelial kidney 293T (HEK-293T) cells (1 × 10⁶) were transiently transfected for the creation of mammalian virus-like particles (VLPs). DNA of each of the three pTR600 mammalian expression vectors (10) expressing the influenza neuraminidase (A/South Carolina/1/1918; H1N1), the HIV p55 Gag sequence, and one of the various H2 wild-type HAs were added in a 1:2:1 ratio with a final DNA concentration of 1ug. Following 72 h of incubation at 37°C, supernatants from transiently transfected cells were collected, centrifuged to remove cellular debris, and filtered through a 0.22 μm pore membrane. VLPs were purified and sedimented by ultracentrifugation on a 20% glycerol cushion at 23,500 × g for 4 h at 4°C. VLPs were resuspended in phosphate buffered saline (PBS), and total protein concentration was determined with the Micro BCA Protein Assay Reagent kit (Pierce Biotechnology, Rockford, IL, USA).

Hemagglutination activity of each preparation of VLP was determined by serially diluting volumes of VLPs and adding equal volume 0.8% turkey red blood cells (RBCs) (Lampire Biologicals, Pipersville, PA, USA) suspended in PBS to a V-bottom 96-well plate with a 30 min incubation at room temperature (RT). Prepared RBCs were stored at 4°C and used within 72 h. The highest dilution of VLP with full agglutination of RBCs was considered the endpoint HA titer. The HA sequences used for VLPs were Mal/NL/01, Chk/Pots/84, Muskrat/Russia/63/2014 (Musk/Rus/14), Duck/Cambodia/419W12M3/2013 (Duk/Cam/13), J/57, Moscow/1019/1965 (Mosc/65), T/64, Duk/HK/78, England/10/1967 (Eng/67), Mal/WI/08, Sw/MO/06, Quail/Rhode Island/16–018622-1/2016 (Qu/RI/16), Turkey/California/1797/2008 (Tur/CA/08), Mallard/Maryland/242/2001 (Mal/MD/01), Avian/Massachusetts/25756/1990 (Av/MA/90), Green-winged-teal/Ohio/175/1986 (GWT/OH/86).

Hemagglutination-inhibition assay

The hemagglutination inhibition (HAI) assay was used to quantify receptor-binding HA-specific antibodies by measuring the inhibition in the agglutination of turkey erythrocytes. The protocol was adapted from the WHO laboratory of influenza surveillance manual (11). To inactivate nonspecific inhibitors, the sera was treated with receptor-destroying enzyme (RDE) (Denka Seiken, Co., Japan) prior to being tested. Briefly, three parts RDE was added to one part sera and incubated overnight at 37°C. RDE was inactivated by incubating the serum-RDE mixture at 56°C for ~45 minutes. After the incubation period, six parts PBS was added to the RDE-treated sera. RDE-treated sera were two-fold serially diluted in V-bottom microtiter plates. An equal volume of each virus-like particle (VLP) was adjusted to approximately 8 hemagglutination units (HAU)/25 μL, and was added to each well of the V-bottom microtiter plates. The plates were covered and incubated at RT for 20 minutes before adding 50 μL of RBCs which were allowed to settle for 30 minutes at RT. The HAI titer was determined by the reciprocal dilution of the last well that contained non-agglutinated RBCs.

ELISAs

A high-affinity, 96-well flat-bottom enzyme-linked immunosorbent assay (ELISA) plate was coated with 50 uL of 2 μg/mL of recombinant hemagglutinin (rHA) in ELISA carbonate buffer (50mM carbonate buffer [pH 9.5] with 5 μg/mL BS) and incubated overnight at 4°C. The hemagglutinins included Mal/NL/01, J/57 and Mal/WI/08. Chimeric rHAs were designed using a wild-type H10 HA (A/Eurasian_Wigeon/Netherlands/3/2007) sequence and a wild-type H2 HA (A/Japan/305/1957) sequence. The cH10/H2 protein was designed using the HA1 region of the H10 HA sequence (amino acids 1–300) and the HA2 region of the H2 HA sequence. The HA2 sequence was defined as amino acids 301–562. Conversely, the cH2/H10 protein was designed using the HA1 region of the H2 HA sequence and the HA2 region of the H10 HA sequence.

After the overnight incubation, the plates were washed with PBS and non-specific epitopes were blocked with 1% bovine serum albumin (BSA) in PBS with 0.05% Tween 20 (PBST + BSA) solution for 1 hr at room temperature. The blocking buffer was then removed and three-fold serial dilutions of RDE treated sera were added to the plate with the highest initial dilution being 1:50. The plates were incubated at 37°C for 90 minutes. The plates were washed in PBS, and goat anti-mouse IgG-HRP was added to the plates at a 1:4000 dilution in PBST+BSA. Plates were then incubated at 37°C for 1 hour. Following another wash step, 2,2’-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) substrate in McIlvain’s Buffer (pH 5) was added to each well, and incubated at 37°C for 15 min. The colorimetric reaction was stopped with the addition of 1% SDS in ddH₂O, and the absorbance was measured at 414 nm.

Each sample was run in triplicate on a single plate. ‘Background’ wells without primary or secondary antibodies added were included on each plate (n=12). The average of each dilution of each of the triplicate samples was calculated as well as the average absorbance of the background wells. The area under the curve for each sample and the area under the curve of the background was calculated before the background area under the curve was subtracted from each samples’ area under the curve.

In vitro differentiation of B-cells

Approximately 2E+6 viable PBMCs were cultured in complete media containing RPMI 1640 media (MilliporeSigma) with 10% FBS (Atlanta Biologicals), 23.8 mM sodium bicarbonate (Thermo Fisher Scientific), 7.5 mM HEPES buffer (Amresco), 170μM Penicillin G (Tokyo Chemical Industry), 137μM streptomycin (Thermo Fisher Scientific), nonessential amino acid solution (Thermo Fisher Scientific), 500ng/mL R848 (Invivogen), and 5 ng/mL rIL-2 (R&D Systems) for 7–9 days in incubator at 37°C with 5% CO₂. Conditioned medium supernatants were harvested and evaluated for total and rHA-specific IgG abundance by ELISA starting at a 1:5 dilution.

Neutralization Assays

The neutralization assay was used to identify the presence of virus-specific neutralizing antibodies. The protocol was adapted from the WHO laboratory of influenza surveillance manual (11). Antibodies were diluted in ½-log increments with serum-free media and incubated with 100 times TCID50 for one hour. The antibody-virus mixture was then added to the incomplete (FBS-free) DMEM washed MDCK cells in the 96-well plate. After two hours, the MDCK cells were washed with incomplete DMEM. Approximately 200 uL of DMEM with P/S and TPCK was added to each of the 96 wells. The cell monolayers were checked daily for cytopathic effect (CPE). After three or four days, media in each well was removed, and the MDCK cells were fixed with 10% buffered formalin. MDCK cells were stained using 1% crystal violet. CPE was defined as >10% CPE of cells per well.

Study Approval

The study procedures, informed consent, and data collection documents were reviewed and approved by the IRB of the University of Georgia. Subjects were recruited at a medical facility in Athens, Georgia and enrolled with written, informed consent. Exclusion criteria included documented contraindications to Guillain-Barré syndrome, dementia or Alzheimer’s disease, allergies to eggs or egg products, estimated life expectancy <2 years, medical treatment causing or diagnosis of an immunocompromising condition, or concurrent participation in another influenza vaccine research study. Influenza virus did not circulate widely in the community during the time periods that the subjects participated, and as such, participants were not monitored for influenza virus infection during that time period. However, study participants were asked during each visit if they had experienced flu-like symptoms, and those who did were excluded from the study.

Statistics

Statistical significance for assays was defined as a p-value of less than 0.05. Limit of detection for HAIs is 1:10 and 1:5 was used for statistical analysis. The HAI titers were transformed by log2 for analysis and graphing for both paired t-test and one-way ANOVA analysis. Limit of detection of ELISAs is 0 area under the graphed curve. Background for the ELISAs was defined as the average absorbance of the control wells for each plate. The background for each plate was subtracted from the average of each samples’ replicates before paired t-test and one-way ANOVA analysis. Geometric mean titers were calculated for neutralization assays, but the Log 10 titers were used for both paired t-test and ANOVA analysis. All error bars on the graphs represent standard mean error.

Results

Demographics of Participants

Participant demographics are listed in table 1. Of the 138 participants, 105 are females. Caucasian was the most identified ethnicity with 110 out of the 138 individuals identifying as such. Of the remaining 28 participants, 15 identified themselves as Hispanic, 8 identified as African American and the remaining 5 participants identified as either Asian, Pacific Islander or chose not to answer. Of the 138 participants, 35 individuals were born before 1953 with the oldest being born in 1934. In the middle age group (1953–1964) consisted of 22 participants. The remaining 81 individuals were born after 1964 with the youngest born in 1996.

Study Outline

Serum from participants was evaluated for responses to both H1N1 and H2Nx influenza viruses. Individuals were divided into three age groups representing those born before the H2N2 pandemic and likely imprinted with H1N1 influenza viruses (1934–1952), those likely imprinted with H2N2 influenza viruses (1953–1964) and those likely not infected with H2N2 influenza viruses (1965–1996). Participants were vaccinated with the seasonal influenza virus split inactivated tetravalent vaccine. Serum from individuals in the cohort was evaluated using HAI for the presence of receptor binding site (RBS) specific antibodies to both H1N1 and H2Nx influenza virus HA proteins. Serum was also evaluated using ELISAs for total antibody binding to H2 HA protein on D0 and D21 post-vaccination. Additionally, individuals from each of the three age groups were randomly selected based upon availability of PBMCs. Sera from these selected individuals was tested for the presence of neutralizing antibodies to H2Nx influenza viruses. PBMCs from these select individuals were used to determine B-cell memory responses to both H1 and H2 rHAs.

H1N1 and H2Nx antibodies block attachment

Antibodies titers that block HA attachment by binding in and around the RBS of H1N1 influenza viruses were quantified using the HAI assay on samples from D0 and D21 post-vaccination. Before separating into age groups, the entire cohort had significant increases (p<0.05, paired t-test) from D0 to D21 to each of the H1N1 viruses except for the Den/57, Sing/86, Bej/91 and NC/99 viruses (Sup F1). On D0 and D21, HAI titers between each of the sixteen viruses in the panel were compared. On both D0 and D21, the FM/47, Den/57, NJ/76 and CL/83 HAI titers are significantly lower (p<0.05, ANOVA) than the HAI titer to nearly every other virus in the panel (Fig S1).

After separating into the three age groups, the youngest individuals (those born after 1964) showed significant (p<0.05, paired t-test) HAI titer increases to thirteen of the sixteen H1N1 viruses in the panel. The viruses that did not show significant increases were Weiss/43, FM/47 and Den/57 (Fig 1). Similar to the entire cohort, the youngest individuals had significantly lower (p<0.05, ANOVA) HAI titers for the FM/47, Den/57, NJ/76 and CL/83 viruses compared to nearly every other virus in the panel (Fig 1).

Figure 1: Attachment Blocking Antibodies to H1N1 Viruses.

Average HAI titers against 16 H1N1 influenza viruses for each of the three age groups both before and after vaccination. Serum from each participant was tested against and individual data points are shown. The mean and standard error are show for each group. Dotted lines indicate 1:40 and 1:80 HAI titer respectively. Error bars represent standard mean error.

The middle age group (1953–1964) only showed significant (p<0.05, paired t-test) HAI titer increases to SC/18, NJ/76, BR/78, CL/83, Bris/07, CA/09 and MI/15 (Fig 1). On D0, there was no significant difference in HAI titers between any of the H1N1 viruses while on D21, the CA/09 and MI/15 HAI titers became significantly higher (p<0.05, ANOVA) compared to nearly every other virus in the panel (Fig 1). The oldest individuals (1934–1952) showed significant (p<0.05, paired-test) HAI titer increases to SC/18, Weiss/43, FM/47, NJ/76, USSR/77, BR/78, CL/83, Bris/07, CA/09 and MI/15 from D0 to D21 (Fig 1). On D0, Weiss/43 had significantly higher HAI titers (p<0.05, ANOVA) to every other H1N1 influenza virus in the panel except for CA/09. On D21, Weiss/43, CA/09 and MI/15 had significantly higher (p<0.05, ANOVA) HAI titers to nearly every other H1N1 virus in the panel (Fig 1).

When combining all of the participants in the cohort, HAI titers to H2 HA expressing virus-like particles (VLPs) significantly increased (p<0.05, paired t-test) from D0 to D21 for the Chk/Pots/84, Duck/Cam/13, Sw/MO/06 and Tur/CA/08 HAs (Fig S2). Interestingly, the Mal/NL/01, J/57 and T/64 H2 HA expressing VLPs had a significant decrease (p<0.05, paired t-test) in HAI titers from D0 to D21. All of the other H2 HA expressing VLPs had no significant change in HAI titers from D0 to D21. On D0, Chk/Pots/84, T/64, Duck/HK/78, Mosc/65 and Av/MA/90 had significantly lower (p<0.05, ANOVA) HAI titers to nearly every other H2 influenza virus in the panel while J/57, GWT/OH/86, Sw/MO/06, Tur/CA/08 and Qu/RI/16 had significantly higher (p<0.05, ANOVA) HAI titers to nearly every other H2 influenza virus in the panel. On D21, the differences between the HAI titers are nearly identical to the differences on D0 (Fig S2).

In the oldest age group (1934–1952), HAI titers to the Chk/Pots/84, Duck/Cam/13, Sw/MO/06 and Tur/CA/08 H2 HA expressing VLPs significantly increased (p<0.05, paired t-test) from D0 to D21 while the Mal/NL/01 and J/57 both significantly decreased (p<0.05, paired t-test) (Fig 2). The remaining nine H2 HA expressing VLPs showed no significant increase or decrease from D0 to D21. On D0, the Chk/Pots/84, Mosc/65, T/64, Duk/HK/78 and Av/MA/90 H2 expressing VLPs had significantly lower (p<0.05, ANOVA) HAI titers compared to nearly every other H2 expressing VLP in the panel. Additionally, the J57, GWT/OH/86, Sw/MO/06, Tur/CA/08, Mal/WI/08 and Qu/RI/16 had significantly higher (p<0.05, ANOVA) HAI titers compared to nearly every other H2 expressing VLP in the panel. On D21, the GWT/OH/86, Sw/MO/06 and Tur/CA/08 H2 expressing VLPs had significantly higher (p<0.05, ANOVA) HAI titers compared to nearly every other H2 expressing VLP in the panel (Fig 2).

Figure 2: Attachment Blocking Antibodies to H2Nx Viruses.

Average HAI titers against VLPs expressing 16 H2 HA sequences for each of the three age groups both before and after vaccination. Serum from each participant was tested against and individual data points are shown. The mean and standard error are show for each group. Dotted lines indicate 1:40 and 1:80 HAI titer respectively. Error bars represent standard mean error.

In the middle age group (1953–1964), HAI titers to both the Chk/Pots/84 and Sw/MO/06 H2 HA expressing VLPs significantly increased (p<0.05, paired t-test) from D0 to D21 while the Mal/NL/01 and T/64 HAI titers significantly decreased (p<0.05, paired t-test). The twelve other H2 HA expressing VLPs had no significant change in HAI titers between the two time points. On D0, the Chk/Pots/84, Mosc/65 and Av/MA/90 H2 expressing VLPs had significantly lower (p<0.05, ANOVA) HAI titers to most of the other H2 expressing VLPs in the panel while the J/57, GWT/OH/86, Tur/CA/08 and Qu/RI/16 had significantly higher (p<0.05, ANOVA) HAI titers to most of the other H2 expressing VLPs in the panel (Fig 2). On D21, only the Sw/MO/06 and Tur/CA/08 H2 expressing VLP plasmids had significantly higher (p<0.05, ANOVA) HAI titers compared to most of the other H2 expressing VLPs in the panel (Fig 2).

In the youngest age group (1965–1996), the only H2 HA expressing VLP that significantly increased (p<0.05, paired t-test) in HAI titer from D0 to D21 was Mosc/65. The Chk/Pots/84 and Av/MA/90 H2 HA expressing VLPs both showed a significant decrease (p<0.05, paired t-test) in HAI titers from D0 to D21 while the other thirteen VLPs showed no significant change between the two time points. On D0, the Chk/Pots/84, Mosc/65 and Musk/Rus/14 H2 expressing VLPs had significantly higher (p<0.05, ANOVA) HAI titers compared to most of the other H2 expressing VLPs in the panel. On D21, only the Musk/Rus/14 retained its significantly higher (p<0.05, ANOVA) HAI titers over most of the other H2 expressing VLPs in the panel (Fig 2).

H2 HA Total Antibody Binding

Total antibody binding on D0 and D21 post-vaccination was quantified using ELISAs. Three H2 rHAs representing each of the three H2 HA phylogenetic clades were used for this study (Fig 3). These specific rHAs were selected due to their phylogenic diversity and based upon previously published articles (12, 13). When all of the participants were included, ELISA titers significantly decreased (p<0.05, paired t-test) to the Mal/NL/01 rHA and significantly increased to the J/57 rHA from D0 to D21. There was no significant change in ELISA titers to the Mal/WI/08 rHA. On D0, the ELISA titers to the J/57 rHA were significantly higher (p<0.05, ANOVA) than the ELISA titers to either the Mall/NL/01 or Mal/WI/08 rHAs. On D21, the ELISA titers to the J/57 rHA was again significantly higher (p<0.05, ANOVA) than the other two rHAs. Additionally on D21, the ELISA titers to the Mal/WI/08 rHA were significantly higher (p<0.05, ANOVA) than the ELISA titers to the Mal/NL/01 rHA (Fig 3A).

Figure 3: Serum rH2 HA ELISA Titers.

Participants were evaluated together and also separated into the three age groups. The average area under the curve for each participant was recorded and plotted above. Serum from participants were tested on Day 0 and Day 21 after vaccination. The mean and standard mean error are shown. Three H2 rHAs were used: Mal/NL/01, J/57 and Mal/WI/08.

When the three age groups are separated, ELISA titers to the Mal/NL/01 rHA significantly decreased (p<0.05, paired t-test) from D0 to D21 while the J/57 and Mal/WI/08 showed no significant difference across any of the three age groups (Fig 3B–D). On D0 in both the oldest (1934–1952) and youngest (1965–1996) age groups, the Mall/NL/01 rHA had significantly lower (p<0.05, ANOVA) ELISA titers than the J/57 rHA. In the middle age group (1953–1964), there was no significant difference between the ELISA titers of any of the rHAs (Fig 3B and D). On D21, the oldest (1934–1952) and middle (1953–1964) age groups both had significantly lower (p<0.05, ANOVA) ELISA titers for the Mal/NL/01 rHA compared to both the J/57 and Mal/WI/08 rHAs (Fig 3B and C). The youngest age group (1965–1996) also had significantly lower (p<0.05, ANOVA) ELISA titers for the Mal/NL/01 rHA compared to both the J/57 and Mal/WI/08 rHAs, but the Mal/WI/08 rHA also had significantly lower (p<0.05, ANOVA) ELISA titers compared to the J/57 rHA in this age group (Fig 3D).

H2Nx Neutralization Titers

Individuals born before 1965 had high neutralization titers (>1:150) to Sw/MO/06 both before and after vaccination (Fig 4A–C). These specific viruses were selected based on their phylogenic diversity and previously published articles (12, 13). Participants born after 1965 had low average neutralization titers (<1:15) to all viruses, except Sw/MO/06 both before and after vaccination (Fig 4D). One participant in the youngest age group had significantly higher (p < 0.05, t-test) neutralization titers to the Sw/MO/06 virus than the other participants in that age group.

Figure 4: Serum H2Nx Neutralization Titers.

Neutralization titers were obtained from select participants’ sera on Day 0 and Day 21. Titers were obtained by taking the geometric mean titer of the replicates for each of the vaccine groups. Serum was tested against seven H2Nx influenza viruses. The mean and standard mean error are shown. The lower limit of detection is 5 while the upper limit of detection is 640.

After vaccination, none of the age groups nor the combined participants showed any significant difference in neutralization titers from D0 to D21 for any of the H2 influenza viruses (Fig 4). On both D0 and D21 in the combined cohort, in the oldest age group (1934–1952) and the middle age group (1953–1964), the Sw/MO/06 virus had significantly higher (p<0.05, ANOVA) neutralization titers compared to the other H2 influenza viruses in the panel (Fig 4B–C). On both D0 and D21 in the youngest age group, there were no significant differences between any of the neutralization titers for any of the H2 influenza viruses (Fig 4D).

IgG secreting B memory cells

PBMC stimulation of select individuals was evaluated using ELISAs. When all individuals in the cohort were combined, memory IgG titers significantly increased (p<0.05, paired t-test) to the J/57 and T/64 rHAs from D0 to D21. The Mall/NL/01, Mal/WI/08, the cH2/H10 and the cH10/H2 showed no significant difference from D0 to D21. These results were the same for individuals in the oldest age group (1934–1952) (Fig 5A–B).

Figure 5: Memory B-cell Responses to H1 rHAs.

Individuals from each of the three age groups were randomly selected based upon availability of PBMCs. Titers of these select individual’s’ memory B cell–derived antibody (IgG) against four H1N1 rHAs prior to and 21 days following vaccination. This is measured by ELISA following in vitro differentiation of PBMC from select participants. Participants were evaluated together and also separated into the three age groups on D0 and D21 post-vaccination. The average area under the curve for each participant was recorded and plotted above. The mean and standard mean error are shown.

In the youngest individuals (1965–1996), there was no significant change in IgG titers in any of the six rHAs from D0 to D21 (Fig 5D). Individuals in the middle age group (1953–1964) only had significant increases (p<0.05, paired t-test) in IgG titers to Mal/WI/08. There was no significant change in ELISA titers to any of the other five rHAs (Fig 5C).

In the combined cohort, J/57 and T/64 had significantly lower (p<0.05, ANOVA) ELISA titers to each of the other rHAs on both D0 and D21. The Mal/NL/01 rHA also had significantly lower (p<0.05, ANOVA) ELISA titers to the cH10/H2 rHA on both D0 and D21. Mal/NL/01 also had significantly lower (p<0.05, ANOVA) ELISA titers to Mal/WI/08 on D0, but not on D21 post-vaccination (Fig 5A). On D0 in the oldest age group (1934–1952), ELISA titers to the J/57 and T/64 rHAs were significantly lower (p<0.05, ANOVA) than the ELISA titers to each of the other rHAs except for Mal/WI/08. The Mal/NL/01 ELISA titers were also significantly lower (p<0.05, ANOVA) to the ELISA titers for the cH10/H2 rHA. On D21 post-vaccination, J/57 and T/64 ELISA titers were significantly lower (p<0.05, ANOVA) to both the cH2/10 and cH10/H2 rHAs. The Mal/NL/01 and cH2/H10 rHA ELISA titers were also significantly lower (p<0.05, ANOVA) than the ELISA titers to the cH10/H2 rHA (Fig 5B).

In the middle age group (1953–1964) on D0, T/64 and J/57 had significantly lower (p<0.05, ANOVA) ELISA titers to the Mal/WI/08, cH2/H10 and cH10/H2 rHAs. The ELISA titers to the Mal/NL/01 rHA was significantly higher (p<0.05, ANOVA) than the T/64 ELISA titers and significantly lower (p<0.05, ANOVA) than the cH10/H2 ELISA titers. On D21, both the J/57 and T/64 ELISA titers were significantly lower (p<0.05, ANOVA) than the other rHA ELISA titers. The cH10/H2 rHA protein also had significantly higher (p<0.05, ANOVA) ELISA titers than the Mal/NL/01 rHA. In the youngest age group (1965–1996) on both D0 and D21, both the J/57 and T/64 ELISA titers were significantly lower (p<0.05, ANOVA) than the other rHA ELISA titers. The cH10/H2 rHA protein also had significantly higher (p<0.05, ANOVA) ELISA titers than the Mal/NL/01 rHA (Fig 5C).

Memory IgG ELISA titers to each of the four H1N1 rHAs did not change significantly from D0 to D21 in the combined group or any of the three separated age groups (Fig 6). In the combined age groups on both D0 and D21, the ELISA titers to Bris/07 and Cal/09 are both significantly higher (p<0.05, ANOVA) than the ELISA titers to the Sing/86 and NC/99 rHAs (Fig 6A). In the oldest age group (1934–1952) on D0, there was no significant difference between any of the groups. On D21, all of the ELISA titers were not significantly different form each other except for the Cal/09 titers being significantly higher than the NC/99 ELISA titers (Fig 6B). In the middle age group (1953–1964) on D0, the ELISA titers to Cal/09 were significantly higher (p<0.05, ANOVA) than the ELISA titers to both the Sing/86 and NC/99 rHAs. On D21, the ELISA titers to Cal/09 were also significantly higher (p<0.05, ANOVA) than the ELISA titers to Sing/86 (Fig 6C). In the youngest age group (1965–1996) on D0, the Cal/09 ELISA titers were significantly higher (p<0.05, ANOVA) than the ELISA titers to the NC/99 rHA. There were no significant differences between any of the rHAs on D21 (Fig 6D).

Figure 6: Memory B-cell Responses to H2 rHAs.

Titers of memory B cell–derived antibody (IgG) against three WT H2 rHAs prior to and 21 days following vaccination. Chimeric H2/H10 rHAs are also shown in the last two columns. The titers were measured by ELISA following in vitro differentiation of PBMC from select participants. Participants were evaluated together and also separated into the three age groups on D0 and D21 post-vaccination. The average area under the curve for each participant was recorded and plotted above. The mean and standard mean error are shown.

Discussion

H2N2 influenza viruses caused a pandemic in 1957 and continued to circulate in the human population until 1968. Given that most individuals are infected with an influenza virus in the first four years of their lives, our study divided individuals into three age groups. The first group was born before 1953 who were immunologically imprinted with H1N1 influenza viruses, but still exposed to H2N2 influenza viruses. The second group was likely imprinted with H2N2 influenza viruses and were born form 1953–1964. The third and final group was likely imprinted with either H1N1 or H3N2 influenza viruses and have likely never been exposed to H2N2 influenza viruses.

The effect of different primary infections of influenza viruses on immunological memory has been documented (6–8). B-cell expansion in response to strains within the seasonal influenza viruses has also been documented (14). While the difference in ages effects the immune response to H1N1, H2N2, H3N2, and influenza B viruses, there has been no study done on the effect of seasonal influenza virus vaccination on the expansion of B-cells to H2 HA proteins.

HAI assays quantify the titer of receptor binding site blocking antibodies. All participants had period specific H1N1 HAI titers based upon their age. Participants born before 1964 had antisera with HAI activity against the majority of the H2 viral particles in the panel, while those born after 1964 had antisera with low HAI activity to each of the H2 viral particles. While there were significant increases in HAI titers to H2 VLPs in each age group, the specific H2 VLPs varied between the groups. This indicates that there are some cross-reactive antibodies that block the RBS of the H2 HA proteins that are being elicited by the seasonal influenza vaccine. It is unclear if these cross-reactive antibodies are being elicited by the H1N1 or H3N2 influenza viruses or perhaps the presence of both of these components in the seasonal influenza vaccine. Interestingly, each age group also had significant decreases in HAI titers to several H2 VLPs. In order to avoid decreasing HAI titers to H2 HA proteins, an H2 influenza virus component would likely need to be included in the seasonal influenza vaccine or developed as a new stand-alone vaccine.

Total antibody binding to rH2 HA proteins was evaluated using ELISAs. In both the combined and all three of the separated age groups, the ELISA titers significantly decreased to the Mal/NL/01 rHA while the ELISA titers did not change to either of the other H2 rHA proteins. Since both the HAI and ELISA titers decreased to the Mal/NL/01 HA protein, it is likely that the antibodies that humans possess that cross-react with this HA protein are being drowned out by the expansion of B-cells that are specific to the H1 and H3 components of the seasonal influenza vaccine. It is unclear if this decrease in antibody titers to this specific H2 HA protein would persist long-term of if the significance would eventually be eliminated once the H1N1 and H3N2 specific B-cells returned to baseline. More research is also needed to determine if this effect would be replicated with the seasonal influenza vaccine from another season with different H1N1 and/or H3N2 components.

Neutralizing antibodies to H2Nx influenza viruses were evaluated in select participants representing each of the three age groups. There were no significant changes in neutralizing antibody titers following vaccination in any age group. All of the participants in the youngest age group had low neutralizing antibody titers to each of the seven H2Nx influenza viruses except for one individual who had significantly higher neutralization titers to the Sw/MO/06 virus than any of the other participants in that age group. Participants born after 1964 also had much lower neutralization titers to each of the seven H2Nx viruses than older individuals. These results indicate that if another pandemic were caused by H2 influenza viruses, the vast majority of younger individuals would not have any neutralizing antibodies to the virus.

The memory IgG titers to the H1 rHA proteins did not significantly increase as expected. This could be attributed to the H1 rHAs being evaluated were not the H1N1 influenza virus strain used in the seasonal influenza vaccine. However, the ELISA titers of the IgG secreting B memory cells did significantly change for the rH2 HA proteins. In both the combined and oldest age groups, the Mal/NL/01, Mal/WI/08, cH2/H10 and cH10/H2 rHAs did not significantly change after vaccination. However, the IgG antibody titers against both the J/57 and T/64 rHAs significantly increased after vaccination in both of these groups. The middle age group only showed significant ELISA titer increases to the Mal/WI/08 rHA and the youngest age group had no significant changes in ELISA titers to any of the H2 rHAs. These memory IgG results indicate that in the oldest age group, B-cells that are cross-reactive to the human H2 rHAs are being stimulated by the seasonal influenza virus vaccination. Interestingly, the middle age group who would’ve been imprinted with H2N2 influenza viruses only showed a significant increase in memory IgG titers to the Mal/WI/08 rHA. This difference between the oldest two age group could possibly be attributed to the oldest age group being imprinted with H1N1 influenza virus before subsequently being infected multiple times with H2N2 influenza viruses. This could possibly lead to these individuals developing and retaining more cross-reactive memory B-cells to the H2 HA proteins than the individuals who were imprinted with the H2N2 influenza viruses before being infected with H3N2 and finally H1N1 influenza viruses. More research is needed to determine if the exact order and number of influenza virus infections significantly changes the number of subtype cross-reactive memory B-cells in humans.

Taken together, the data in this paper demonstrates that seasonal influenza vaccination does not elicit cross-reactive antibodies to the majority of H2 HA proteins. While some HAI and ELISA titers for a select few H2 proteins did increase after vaccination, the increases were not consistent by number or specific H2 influenza virus across any of the different age groups. This suggests that individuals born during different time periods of influenza virus subtype circulation have different serum antibody and memory B-cell responses to H2 influenza viruses. While other factors including cellular responses play a role in influenza virus immunity, the lack of serum antibodies and memory B-cells, particularly in younger individuals suggests that the likelihood of H2N2 influenza viruses causing a future pandemic will likely increase every year as the population’s immunity continues to wane.

Supplementary Material

1. Supplemental Figure 1: Combined Cohort Attachment Blocking Antibodies to H1N1 Viruses.

Average HAI titers for combined cohort and age groups both before and after vaccination. Serum from each participant was tested against 16 H1N1 influenza viruses. The individual data points, mean and standard error are show for each group. Dotted lines indicate 1:40 and 1:80 HAI titer respectively. Error bars represent standard mean error.

2. Supplemental Figure 2: Combined Cohort Attachment Blocking Antibodies to H2Nx Viruses.

Average HAI titers for combined cohort and age groups both before and after vaccination. Serum from each participant was tested against 16 H2 HA sequences. The individual data points, mean and standard error are show for each group. Dotted lines indicate 1:40 and 1:80 HAI titer respectively. Error bars represent standard mean error.