Vaccination with 2014–15 Seasonal Inactivated Influenza Vaccine Elicits Cross-Reactive Anti-HA Antibodies with Strong ADCC Against Antigenically Drifted Circulating H3N2 Virus in Humans

Abstract

It is well established that virus neutralizing (VN) antibodies to hemagglutinin (HA) antigens of influenza A viruses provide optimal protection against antigenically matched strains of influenza A viruses. In contrast, little is known about the potential role of HA-specific, non-neutralizing antibodies in protection against human influenza illness at present. In this study, we show that individuals vaccinated with the 2014–15 seasonal inactivated influenza vaccine displayed strong A/H3N2 HA-specific antibody-dependent cell-mediated cytotoxicity (ADCC) activities against an antigenically drifted H3N2 virus, despite poor induction of cross-reactive neutralizing antibodies against the antigenic variant. Given that passive transfer of influenza HA-monospecific immune sera with negligible levels of HA-specific VN antibodies can often confer considerable cross protection against lethal challenge with heterologous influenza viruses in animal models, it is conceivable that HA-specific, non-neutralizing antibodies may provide certain degree of cross protection against antigenically drifted influenza A viruses through ADCC in case of influenza vaccine mismatches. This may have important implications for public health.

Introduction

Vaccination is an effective public health tool to prevent influenza illness. Currently licensed influenza vaccines induce both virus neutralizing (VN) and non-neutralizing antibodies directed toward the hemagglutinin (HA) of influenza A and B viruses. VN antibodies provide optimal protection against antigenically matched influenza virus strains and are an immune correlate of protection against influenza infection (6). In contrast to VN antibodies, assessment of the possible role in protection for HA-specific, non-neutralizing antibodies is largely neglected due to a lack of understanding of the biological functions of this subset of anti-HA antibodies. The available evidence shows that current influenza vaccines may confer some level of cross protection against antigenically drifted viruses despite poor induction of cross-reactive neutralizing antibodies (7,8,21). Therefore, it is important to identify new correlates of protection in case of vaccine mismatches.

The emergence of antigenically drifted H3N2 viruses in the 2014–15 Northern Hemisphere influenza season offered a unique opportunity to address this issue. The majority of A/H3N2 viruses of the season circulated globally were antigenically distinct from the A/H3N2 vaccine and belonged to genetic group 3C.2a or 3C.3a (5). In this study we show that strong HA-specific, cross-reactive antibody-dependent cell-mediated cytotoxicity (ADCC) activity against the representative antigenically drifted A/Switzerland/9715293/2013 (3C.3a) virus was readily detectable in sera from individuals who received the 2014–15 seasonal inactivated quadrivalent influenza vaccine (QIV) containing an A/Texas/50/2012 (3C.1) H3N2 component.

Materials and Methods

Human serum samples

Human serum samples were obtained from 20 healthy adults (mean age: 31.5 years; range: 21–46 years; male: 5; female: 15) who received one dose of 2014–15 seasonal QIV containing Tex/50 virus as the H3N2 vaccine component (Fluzone Quadrivalent; Sanofi Pasteur). The sera were collected pre- (day 0) and post- (day 21–28) vaccination and were selected based on achievement of a fourfold or more rise in titers against the H3N2 A/Texas/50/2012. The sera were acquired through a contract and received as anonymous samples and thus were exempt from CDC Institutional Review Board review.

Influenza A viruses

All of the influenza virus stocks used in the present study, including A/Texas/50/2012 (Tex/50, H3N2, Genetic group: 3C.1), A/Nebraska/04/2014 (NE/4, H3N2, Genetic group: 3C.2a), and A/Switzerland/9715293/2013 (SZ/9715293, H3N2, Genetic group: 3C.3a), were propagated in MDCK-SIAT 1 cells (19).

Recombinant HA antigens

Full-length, recombinant HA (rHA) antigens of influenza A viruses with his-tag were obtained from Influenza Reagent Resource (www.influenzareagentresource.org). The rHA antigens were produced by baculovirus infection of Trichoplusia ni insect larvae. Trimeric, soluble HA0 form of the rHA proteins, was purified using standard chromatographic methods and verified by SDS-PAGE.

Microneutralization assay

Due to low or no hemagglutination activity of many drifted A/H3N2 viruses, titers of ferret and human sera were determined by a standard microneutralization (MN) assay as described previously (22). Briefly, sera were first heat-inactivated for 30 min at 56°C. The sera were then twofold serially diluted in 96-well ELISA plates from a starting dilution of 1:10. Hundred TCID₅₀ of Tex/50 2014–15 H3N2 vaccine virus and NE/4 or SZ/9715293 virus were then added into each well of the plates, gently mixed and incubated for 1 h at 37°C. 1.5 × 10⁴ of MDCK-SIAT 1 cells were added into each serum–virus mixture. After overnight incubation at 37°C, viral nucleoprotein in virally infected MDCK-SIAT 1 cells was detected by ELISA. Each serum sample was tested in duplicate. The final titer was the GMT of the duplicate titers.

ADCC NK cell activation assay

Influenza HA-specific ADCC activities of human serum samples were quantified using a modified ADCC natural killer (NK) cell activation assay described in detail elsewhere (Zhong, et al., article in preparation). Briefly, nickel-coated plates were first coated with rHA antigens described above. Serially diluted human serum samples were then added into each well of the HA antigen-coated plates. The start dilution was 1:40. The plates were then washed five times with sterile phosphate-buffered saline. Human NK cell line expressing high-affinity (158 V/V) FcγRIIIa receptor (1) was mixed with appropriately diluted PE-conjugated mouse anti-human CD107a (BD Pharmingen) and added into each well. Following incubation and washing, the percentages of CD107a expression on NK cells were determined on a Becton Dickenson LSR II flow cytometer using high throughput sampler. The results were expressed as endpoint titers, for example, the highest serum dilution that achieved the 3% of the arbitrary threshold. Each serum sample was tested in duplicate. The final titer was the geometric mean titer (GMT) of the duplicate titers.

Results

Ferret antisera raised against the Tex/50 exhibited an eightfold reduction in titer against antigenically drifted reference virus SZ/9715293 (3C.3a) and (NE/4. 3C.2a) compared with the titer to Tex/50 virus itself (Table 1), demonstrating that these viruses as antigenically drifted from the Tex/50 vaccine virus. Ferret antisera raised to SZ/9715293 and NE/4 reacted with the reciprocal virus to titers that were within fourfold of the titer to homologous virus, indicating antigenic similarity between these viruses representing 3C.3a and 3C.2a genetic groups, respectively. For this reason, SZ/9715293 virus was chosen to represent the antigenically drifted viruses for further analyses in this study.

Table 1.

Antigenic Characterization of H3N2 Viruses by Microneutralization Assay Using Ferret Antisera

		Reference ferret antisera to
Influenza antigen	Genetic group	Tex/50	NE/4	SZ/9715293
A/Texas/50/2012	3C.1	320	160	80
A/Nebraska/04/2014	3C.2a	40	160	56
A/Switzerland/9715293/2013	3C.3a	40	40	160

Ferret antisera were raised against each indicated H3N2 viruses propagated in MDCK-SIAT1 cells. Virus neutralization titers were determined by the microneutralization assays using MDCK-SIAT1 cells. The values represent GMT of duplicate tests. Numbers in bold indicate the homologous titers.

GMT, geometric mean titer.

Ferret antisera do not always recapitulate complex exposure history of humans to influenza viruses (12). We thus used a panel of sera from persons that received the 2014–15 seasonal QIV to further characterize the anti-HA neutralizing antibody response to the H3N2 vaccine component and the antigenically drifted virus, SZ/9715293. As shown in Table 2, pre- and postvaccination GMT of sera against the Tex/50 virus were 19 and 251, respectively, an average 12-fold increase in VN titers. Fourfold or greater increases in MN titers post vaccination were detected in 90% of the serum pairs. In contrast, a fourfold or greater increase in postvaccination titer to SZ/9715293 was detected in only 45% of the 20 serum pairs. The postvaccination GMT of VN antibodies against the drifted virus was only 28.

Table 2.

2014–15 Seasonal QIV Vaccine Elicits Robust, Non-Neutralizing Anti-HA Antibody Response with ADCC Activity Against Antigenically Distinct Circulating H3N2 Virus in Healthy Adults

		A/Texas/50/2012 virus		A/Switzerland/9715293/2013 virus
In vitro functional activity		Prevaccination	Postvaccination	Prevaccination	Postvaccination
VN	GMT (95% CI)	19 (11–34)	251 (137–461)	6 (5–8)	28 (17–48)
	No. of seroconversion^a (%)	—	18 (90.00)	—	9 (45.00)
ADCC	GMT (95% CI)	70 (46–107)	343 (215–547)	63 (40–99)	315 (203–488)
	No. of seroconversion^a (%)	—	18 (90.00)	—	18 (90.00)

Twenty paired serum samples were collected from healthy individuals on day 0 (prevaccination) and day 21–28 following vaccination with 2014/15 seasonal QIV (postvaccination) for MN and ADCC testing against the respective H3N2 vaccine strain and 2014/15 predominant circulating H3N2 virus as indicated, respectively. All MN assays were performed using MDCK-SIAT1 cells.

Fourfold or more rise in titers and titers of post-vaccination serum ≥40.

ADCC, antibody-dependent cell-mediated cytotoxicity; HA, hemagglutinin; MN, microneutralization; QIV, quadrivalent influenza vaccine; VN, virus neutralization.

Despite poor induction of cross-reactive, neutralizing antibodies to the antigenically drifted H3N2 virus in adults that received the 2014–15 influenza vaccine as judged by conventional serological analysis, we found that the same individuals exhibited strong ADCC activity against both the H3N2 vaccine strain and the antigenically drifted virus (Table 2). Postvaccination sera had similar GMT ADCC titers to Tex/50 (GMT 343) and to SZ/9715293 (GMT 315), an average of fourfold rise in titers compared to prevaccination sera. Ninety percent of the serum pairs had a fourfold or greater increase in ADCC titers against both the H3N2 vaccine virus and the circulating H3N2 virus.

Discussion

Influenza A virus-associated ADCC activities have been reported in a number of recent studies. It has been shown that cross-reactive influenza ADCC antibodies were detectable in the sera of healthy individuals as well as from those infected with 2009 pandemic H1N1 virus (13,14). Moreover, HA-specific, ADCC-mediating antibodies have been shown to contribute to the control of 2009 pandemic H1N1 virus infection in a rhesus macaque model (15). In the present study, we show that robust HA-specific ADCC activities against an antigenically drifted circulating H3N2 virus were readily detectable in a group of healthy adults, despite substantially lower titers of VN antibodies induced following vaccination with 2014/15 inactivated seasonal influenza vaccine.

There is ample evidence that influenza vaccines can often provide a degree of cross-protection against challenges with antigenically drifted viruses in the absence of detectable cross-reactive neutralizing antibodies (7,9,20). Furthermore, it has been shown that passive transfer of human or ferret immune sera with low or no VN antibodies conferred considerable protection against subsequent challenge with influenza A viruses in the recipient mice (17,23). Although signs of clinical influenza illness were generally observed, most of the experimental animals survived the lethal challenges with either homologous or the antigenically drifted viruses, along with reduced viral shedding and/or reduced systemic viral spread. The precise immune mechanism(s) that correlate with this protective effect are not fully understood at present. Several humoral immune mediators may contribute to the cross-protection under such circumstances, including cross-reactive anti-HA stalk antibodies, neuraminidase-inhibition antibodies and anti-M2 antibodies. The observation that the degree of cross-protection was generally associated with the antigenic relatedness of HA antigens between the vaccine strain and the challenge viruses (18) suggests that HA-specific, non-neutralizing antibodies may also contribute. In this regard, the HA of the Tex/50 virus shares over 97% of the amino acid sequence identity with those of the NE/4 and SZ/9715293 viruses (data not shown). Mutations in antigenic site B of the HA of the SZ/9715293 were likely contributed to the reduced binding of VN antibodies raised to Tex/50 virus (3). Our finding that anti-HA antibodies induced following 2014–15 influenza vaccination possess nearly equally strong capability to trigger ADCC activities against both the vaccine stain as well as the antigenically drifted SZ/9715293 virus suggest that the cross-reactive, HA-associated ADCC antibodies induced under such circumstance may be directed toward both neutralizing and non-neutralizing antigenic sites of HA antigens shared by the vaccine strain and the antigenically mismatched H3N2 viruses.

Early estimates of overall vaccine effectiveness (VE) of the 2014–15 vaccine against laboratory-confirmed A/H3N2 influenza associated with medically attended acute respiratory illness among U.S. children and adults was only 22% (11). Our view that HA-associated ADCC antibodies may contribute to cross-protection against antigenically drifted influenza viruses does not correlate with this low VE estimate. However, as the authors acknowledged, the VE estimates was limited to the prevention of real-time RT-PCR-confirmed influenza A virus infection among patients with symptoms of an acute respiratory illness, and not severe influenza illness. Current seasonal influenza vaccines, even those containing components with reduced antigenic similarity to circulating viruses may offer benefits of reduction in disease severity (2,4,10,16). In this regard, it is tempting to speculate that although not capable of preventing influenza virus infection, cross-reactive HA-specific ADCC antibodies may contribute to mitigation of influenza clinical illness, a benefit that was not measured by the current VE estimates.

In sum, our study reveals that vaccination with the 2014–15 seasonal influenza vaccine induces HA-specific, non-neutralizing antibodies with strong ADCC activities against the antigenically drifted circulating H3N2 viruses. It remains to be demonstrated whether in vitro measurement of anti-HA non-neutralizing antibodies with ADCC activities represent a new correlate of protection against antigenically drifted viruses.

Footnotes

Acknowledgments

The authors thank Ms. Bonnie Dighero-Kemp, Heather Tatum, Leilani Thomas, and Mr. Eric Gillis, CDC Influenza Division, for their excellent assistance in specimen and data management for this study; and NantKwest, Inc. and Dr. Kerry Campbell, Fox Chase Cancer Center, for kindly providing human NK cell lines for the ADCC NK cell activation assay. This work was supported by Centers for Disease Control and Prevention. The views expressed in this study solely represent those of the authors and do not reflect the official policy of CDC.

Author Disclosure Statement

J.M.K. has received research grants from Juvaris Biotherapeutics and Glaxo SmithKline. For all other authors, no competing financial interests exist.

References

Binyamin

, Alpaugh

, Hughes

, et al. Blocking NK cell inhibitory self-recognition promotes antibody-dependent cellular cytotoxicity in a model of anti-lymphoma therapy. J Immunol, 2008; 180:6392–6401.

Castilla

, Godoy

, Dominguez

, et al. Influenza vaccine effectiveness in preventing outpatient, inpatient, and severe cases of laboratory-confirmed influenza. Clin Infect Dis, 2013; 57:167–175.

Chambers

, Parkhouse

, Ross

, et al. Identification of hemagglutinin residues responsible for H3N2 antigenic drift during the 2014–2015 influenza season. Cell, 2015; 12:1–6.

Deiss

, Arnold

, Chen

, et al. Vaccine-associated reduction in symptom severity among patients with influenza A/H3N2 disease. Vaccine, 2015; 33:7160–7167.

D'Mello

, Brammer

, Blanton

, et al. Update: influenza activity—United States, September 28, 2014–February 21, 2015. MMWR Morb Mortal Wkly Rep, 2015; 64:206–212.

Dormitzer

, Galli

, Castellino

, et al. Influenza vaccine immunology. Immunol Rev, 2011; 239:167–177.

Ducatez

, Webb

, Crumpton

, Webby

. Long-term vaccine-induced heterologous protection against H5N1 influenza viruses in the ferret model. Influenza Other Respir Viruses, 2013; 7:506–512.

Ellebedy

, Ducatez

, Duan

, et al. Impact of prior seasonal influenza vaccination and infection on pandemic A (H1N1) influenza virus replication in ferrets. Vaccine, 2011; 29:3335–3339.

Endo

, Itamura

, Iinuma

, et al. Homotypic and heterotypic protection against influenza virus infection in mice by recombinant vaccinia virus expressing the haemagglutinin or nucleoprotein of influenza virus. J Gen Virol, 1991; 72 (Pt 3):699–703.

10.

Ferdinands

, Olsho

, Agan

, et al. Effectiveness of influenza vaccine against life-threatening RT-PCR-confirmed influenza illness in US children, 2010–2012. J Infect Dis, 2014; 210:674–683.

11.

Flannery

, Clippard

, Zimmerman

, et al. Early estimates of seasonal influenza vaccine effectiveness—United States, January 2015. MMWR Morb Mortal Wkly Rep, 2015; 64:10–15.

12.

Hensley

. Challenges of selecting seasonal influenza vaccine strains for humans with diverse pre-exposure histories. Curr Opin Virol, 2014; 8:85–89.

13.

Jegaskanda

, Laurie

, Amarasena

, et al. Age-associated cross-reactive antibody-dependent cellular cytotoxicity toward 2009 pandemic influenza A virus subtype H1N1. J Infect Dis, 2013; 208:1051–1061.

14.

Jegaskanda

, Vandenberg

, Laurie

, et al. Cross-reactive influenza-specific antibody-dependent cellular cytotoxicity in intravenous immunoglobulin as a potential therapeutic against emerging influenza viruses. J Infect Dis, 2014; 210:1811–1822.

15.

Jegaskanda

, Weinfurter

, Friedrich

, Kent

. Antibody-dependent cellular cytotoxicity is associated with control of pandemic H1N1 influenza virus infection of macaques. J Virol, 2013; 87:5512–5522.

16.

Kostova

, Reed

, Finelli

, et al. Influenza illness and hospitalizations averted by influenza vaccination in the United States, 2005–2011. PLoS One, 2013; 8:e66312.

17.

Krammer

, Jul-Larsen

, Margine

, et al. An H7N1 influenza virus vaccine induces broadly reactive antibody responses against H7N9 in humans. Clin Vaccine Immunol, 2014; 21:1153–1163.

18.

Larson

, Tyrrell

, Bowker

, et al. Immunity to challenge in volunteers vaccinated with an inactivated current or earlier strain of influenza A(H3N2). J Hyg (London), 1978. 80:243–248.

19.

Lin

, Gregory

, Collins

, et al. Neuraminidase receptor binding variants of human influenza A(H3N2) viruses resulting from substitution of aspartic acid 151 in the catalytic site: a role in virus attachment?. J Virol, 2010; 84:6769–6781.

20.

Lipatov

, Hoffmann

, Salomon

, et al. Cross-protectiveness and immunogenicity of influenza A/Duck/Singapore/3/97(H5) vaccines against infection with A/Vietnam/1203/04(H5N1) virus in ferrets. J Infect Dis, 2006; 194:1040–1043.

21.

O'Donnell

, Wright

, Vogel

, et al. Effect of priming with H1N1 influenza viruses of variable antigenic distances on challenge with 2009 pandemic H1N1 virus. J Virol, 2012; 86:8625–8633.

22.

WHO Global Influenza Surveillance Network. Serological diagnosis of influenza by microneutralization assay. In: Manual for the Laboratory Diagnosis and Virological Surveillance of Influenza. Geneva, Switzerland: WHO. 2011:63–77.

23.

Yassine

, Boyington

, McTamney

, et al. Hemagglutinin-stem nanoparticles generate heterosubtypic influenza protection. Nat Med, 2015; 21:1065–1070.