A Computationally Optimized Broadly Reactive Antigen Subtype–Specific Influenza Vaccine Strategy Elicits Unique Potent Broadly Neutralizing Antibodies against Hemagglutinin

This article is featured in In This Issue, p.241

Vaccination is the most cost-effective tool to prevent infectious diseases and their associated complications. Hypervariable pathogens, such as influenza viruses, represent the greatest challenge for development of efficacious vaccines (1). The influenza hemagglutinin (HA) and neuraminidase surface glycoproteins are the major targets of neutralizing Abs (1). However, amino acid substitutions in these two proteins enable viral escape from host immunity. Therefore, because of mismatches between circulating strains and those selected for formulation of seasonal standard of care influenza vaccines, the effectiveness of seasonal influenza vaccination is variable from season to season. Overall, the average coverage provided by the current standard of care influenza vaccine ranges between 40 and 60% in the overall population, but can be as low as 10% in some seasons (2). Therefore, developing a more effective next-generation, universal, or broadly protective influenza vaccine is a high priority. A next-generation influenza vaccine needs to confer long-lasting immunity against a broad spectrum of influenza viruses, both seasonal and pandemic, in all populations. Currently, several different vaccine candidates are being tested in preclinical and clinical trials (3). Our research group has recently described the use of computationally optimized broadly reactive Ags (COBRA) for development of HA-based broadly protective influenza vaccine (4–7). In brief, the COBRA methodology consists of multiple rounds of consensus sequence generation. This approach yields unique HA proteins that incorporate HA epitopes based upon not only the phylogenetic sequence of each isolate but also the time frame and outbreaks in which each isolate was collected (6). This process reduces potential biases in the input sequences used to generate the COBRA sequence by eliminating sequences derived from a single outbreak from a human or zoonotic source in a single location at one specific period of time. Moreover, the weighting of each of the included isolates for sequence, outbreak group, and time of isolation allows COBRA designs to retain highly immunogenic and cross-reactive epitopes.

COBRA HA-based immunogens representing H1N1, H3N2, and H5N1 subtypes have been shown to elicit Abs that neutralize multiple influenza strains and protect mice, ferrets, and nonhuman primates against viral challenge (4–6, 8, 9). Importantly, COBRA HA sequences are effective immunogens when expressed on the surface of virus-like particles (VLP) (4, 6), as HA-ferritin nanoparticles (10), and recombinant HA (rHA) proteins (manuscript in preparation). COBRA HA Ags also mediate infection when expressed on live influenza virions, which can then be used for production of inactivated or live-attenuated influenza vaccines (11).

To effectively develop these COBRA HA-based vaccines, it is fundamental to understand the mechanism(s) by which they elicit broad protective immunity against a diverse panel of influenza viruses. In this study, we used a candidate H1N1 COBRA (named P1) known to elicit a broadly reactive Ab response (4, 11). mAbs were isolated from murine B cells following immunization with the P1 H1N1 COBRA HA Ag. Each mAb was then characterized for breadth of binding to an array of HA proteins and functional activity (hemagglutination inhibition [HAI] and neutralization) against a panel of H1N1 influenza viruses to shed light on the mechanism(s) of breadth conferred by a COBRA HA-based vaccine.

SP2/0 mouse myeloma cells (kindly provided by Dr. L. Wysocki, University of Colorado at Denver) and generated hybridoma cell lines were maintained in B cell medium (BCM) consisting of RPMI 1640 medium (Sigma-Aldrich, Saint Louis, MO) containing 10% FBS (Atlanta Biologicals, Flowery Branch, GA), 23.8 mM sodium bicarbonate (Thermo Fisher Scientific, Waltham, MA), 7.5 mM HEPES (Amresco, Solon, OH), 170 μM penicillin G (Tokyo Chemical Industry, Tokyo, Japan), 137 μM streptomycin (Sigma-Aldrich), 50 μM 2-ME (Sigma-Aldrich), nonessential amino acids (Thermo Fisher Scientific), and 1 mM sodium pyruvate (Thermo Fisher Scientific). For mAb production, hybridoma cell lines were grown in BCM containing 2% Super Low IgG FBS (HyClone, Logan, UT).

Madin–Darby canine kidney (MDCK) cells were maintained in DMEM (Corning, Corning, NY) supplemented with penicillin–streptomycin, BSA fraction V 7.5% solution (Thermo Fisher Scientific), 25 mM HEPES buffer, and 10% heat-inactivated FBS. Stable transfected MDCK cells with the cDNA of human 2,6-sialtransferase (SIAT1) MDCK-SIAT1 (provided by Center for Disease Control and Prevention, Atlanta, GA) were maintained similarly to MDCK cells and supplemented with 1 mg/ml of geneticin (G418 sulfate; Thermo Fisher Scientific). Expi293F cells were grown in Expi293 Expression Medium (Thermo Fisher Scientific) and maintained according to the manufacturer’s instructions.

All the 7:1 recombinant A/Puerto Rico/8/1934 (PR/34) reassortant COBRA P1 and PR/34, A/Chile/1/1983 (Chile/83), A/New Caledonia/20/1999 (NC/99), A/Solomon Island/3/2006 (SI/06), A/Brisbane/59/2007 (Brisb/07) viruses and wild-type A/Singapore/6/1986 (Sing/86), A/Texas/36/1991 (TX/91), A/California/07/2009 (CA/09), and A/Michigan/45/2015 (Mich/15) viruses were propagated in embryonated chicken eggs as previously described (9). The A/Philadelphia/1/2013 (Phil/13) was kindly provided by Scott Hensley (Department of Microbiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA). These viruses were titrated and used for immunization, HAI, and focus reduction assay (FRA) experiments described below. The pandemic A/South Carolina/1/1918 (SC/18) HA-expressing VLP were generated as previously described (4) and used for HAI assays. Many of the reassortant viruses were provided by Virapur (San Diego, CA).

Plaque assays were performed similarly to previously described protocols (11). In brief, low-passage MDCK cells were plated at a density of 3.8 × 10⁵ cells per well in six-well plates (Greiner Bio-One, Monroe, NC) 2 d before the assay. Virus samples were diluted and overlaid onto the cells in 100 μl of DMEM supplemented with penicillin–streptomycin and incubated for 1 h.

Samples were removed, cells were washed, and medium was replaced with 2 ml of Modified Eagle Medium plus 0.8% agarose (Cambrex, East Rutherford, NJ) and incubated for 72 h at 37°C with 5% CO₂. Agarose was removed and discarded. The cells were then fixed with 10% buffered formalin and then stained with 1% crystal violet (Fisher Science Education) for 15 min. Following thorough washing in distilled water to remove excess crystal violet, the plates were dried, the numbers of plaques were counted, and the plaque forming units per milliliter was calculated.

Truncated rHA encoding HA from COBRA P1 and wild-type Chile/83, Sing/1986, NC/99, Brisb/07, CA/09, Mich/15, and the chimeric cH6/1 protein, composed of the head region from H6 and the stem region from H1, were cloned, expressed, and purified as previously described (11) and used for all the binding experiments. In brief, the different HA proteins were expressed through a transient transfection of the Expi293F cells with the different COBRA and H1N1 HA pcDNA3.1/Zeo(+) encoding vectors following the instruction provided by the manufacturer. Alternatively, HA proteins were expressed through the generation of stable transfected cells, supplemented with 100 μg/ml of zeocin (Invivogen, San Diego, CA). rHA proteins were then purified through the ÄKTA Pure system using HisTrap columns (GE Healthcare, Chicago, IL) according to the manufacturer’s instructions. For all in-house purified rHA proteins, purity was assessed through SDS-PAGE using Bolt 10% Bis-Tris Plus gels (Thermo Fisher Scientific), and concentration was determined through a Micro BCA Assay kit (Thermo Fisher Scientific). The TX/91 rHA protein was kindly provided by Florian Krammer (Icahn School of Medicine at Mount Sinai, New York, NY) and the PR/34 and SI/06 rHA proteins were provided by BEI Resources. The monomer and truncated HA1 rHA versions of the PR/34, Brisb/07, and CA/09 proteins were purchased from Immune Technology (New York, NY). The HA protein with C-terminal histidine tag from influenza virus, A/Puerto Rico/8/1934 (H1N1), recombinant from baculovirus, NR-19240 and H1 HA protein from influenza virus, A/Solomon Islands/3/2006 (H1N1), recombinant from baculovirus, NR-15170, monoclonal anti-influenza virus H1 HA, A/South Carolina/1/1918 (H1N1), clone 5D3 (produced in vitro), NR-13451, clone 6B9 (produced in vitro), NR-13452, clone 39E4 (produced in vitro), NR-13453, monoclonal anti-influenza A virus HA: clone IC5-4F8 (produced in vitro), NR-48783, monoclonal anti-influenza A virus HA domain 2 (HA2), clone RA5-22 (produced in vitro), NR-44222, monoclonal anti-influenza virus H1 HA, A/California/04/2009 (H1N1)pdm09, clone 1C5 (produced in vitro), NR-42015, clone 5C12 (produced in vitro), NR-42019, clone CA09-02 (ascites, mouse), NR-28665, clone CA09-09 (ascites, mouse), NR-28666, clone CA09-11 (ascites, mouse), and NR-28667, clone CA09-15 (ascites, mouse), NR-28668 were all obtained from BEI Resources, National Institute of Allergy and Infectious Diseases/National Institutes of Health (Manassas, VA).

The postfusion conformation of P1, SI/06, Brisb/07, and CA/09 rHA proteins were obtained according to previously described protocols (12). HA cleavage was determined by SDS-PAGE and Western blot under reducing conditions.

BALB/c mice (female, 8–10 wk of age), Ab negative for circulating influenza A (H1N1 and H3N2) and influenza B viruses, were purchased from Envigo (Indianapolis, IN) and housed in microisolator units and fed ad libitum. Mice were handled in accordance with protocols approved by the University of Georgia Institutional Animal Care and Use Committee and were cared under U.S. Department of Agriculture guidelines for laboratory animals. Mice that showed signs of severe morbidity or lost >20% of their original weight were humanly euthanized.

Mice (n = 8 per group) were primed intranasally with Chile/83, Sing/86, NC/99, Brisb/07, COBRA P1, or CA/09 influenza viruses and then boosted i.p. 21 d later with the same virus mixed with aluminum hydroxide adjuvant, as previously described (11). For the generation of COBRA P1 and CA/09 B cell hybridomas, a subset of mice (n = 4 per group) received an additional i.p. boost of the same virus on day 42 containing 5 μg of HA quantified as previously described and resuspended in PBS (Corning) (Fig. 1A). On day 45, spleens were harvested and immediately processed for B cell hybridoma generation. Alternatively, splenocytes from immunized mice (n = 4 per group) were harvested and stored frozen in liquid nitrogen until usage.

ELISpot were performed as previously described (13). In brief, MultiScreen_HTS HA filter plates (EMD Millipore, Billerica, MA) were coated overnight at 4°C with 2 μg/ml of the different rHA (COBRA P1, Chile/83, Sing/86, NC/99, Brisb/07, CA/09, and cH6/1) or 2.5 μg/ml of goat anti-mouse IgG, (Sigma-Aldrich) in carbonate buffer (pH 9.4) containing 5 μg/ml BSA (50 μl per well) in a humidified chamber. Additional plates were coated with carbonate buffer containing 5 μg/ml BSA alone. Plates were washed three times with PBS before blocking with BCM for at least 1 h at room temperature. Serially diluted mouse splenocytes were then incubated in ELISpot plates for ∼20 h at 37°C, 5% CO₂. ELISpot plates were then washed three times with PBS + 0.1% Triton X-100 (Sigma-Aldrich), and after an additional five washes with PBS, alkaline phosphatase-conjugated goat anti-mouse IgG1 (Southern Biotech, Birmingham, AL) diluted 1:4000 in blocking buffer (PBS containing 2% BSA [Sigma-Aldrich], 1% bovine gelatin [Sigma-Aldrich], and 0.05% Tween 20 [Thermo Fisher Scientific]) was added and plates incubated for 2 h at 37°C. Plates were washed five times with PBS and then twice with distilled water (dH₂O) before addition of 5-bromo-4-chloro-3-indolyl-phosphate/NBT (BCIP/NBT) one-step solution (Thermo Fisher Scientific) and incubation at 37°C for ∼15 min. Development of the substrate was terminated by decanting solution and gently washing plates with dH₂O. Plates were air dried at room temperature before automated counting using the S6 macro ELISpot reader (analyzers; Cellular Technology Ltd., Shaker Heights, OH).

For the generation of B cell hybridomas, splenocytes from mice immunized as described above were fused with SP2/0 myeloma cells using polyethylene glycol 1450 (American Type Culture Collection, Manassas, VA) similarly to previously reported protocols (14). After fusion, hybridomas were selected by addition of hypoxanthine (Thermo Fisher Scientific) and azaserine (Sigma-Aldrich) at a final concentration of 200 and 11.5 μM, respectively, in BCM. Ten to fifteen days after the respective fusions, culture supernatants were screened by ELISA for reactivity against the homologous rHA Ag (P1 or CA/09 rHA). Positive wells were further expanded under drug selection. Hybridoma cell lines of interest were then single-cell cloned by FACS into 96-well plates (Greiner Bio-One) using a BD FACSAria Fusion (BD Biosciences, San Jose, CA) and by gating on the highest BCR-expressing cells to sort those secreting the highest amount of Ab (15). Single-cell cloned hybridomas were then rescreened by ELISA to confirm Ag specificity (Fig. 1B). In total, four fusions were performed to generate 28 and 11 B cell hybridoma lines from COBRA P1 or CA/09 immunized mice, respectively. Each hybridoma cell line was assigned a novel identifier to trace back the mouse of origin.

For mAb production and purification, mAb were purified by affinity chromatography using HiTrap Protein G HP Columns (GE Healthcare). Briefly, conditioned hybridoma media were applied to protein G columns and purified through the ÄKTA Pure system. The mAbs were eluted by using 0.1 M glycine (pH 2.5) (Amresco). Eluted protein was immediately neutralized with 200 μl of 1.5 M Tris (pH 8.8) (Amresco), and protein-containing elution fractions were pooled, buffer exchanged into PBS containing 0.05% sodium azide (Sigma-Aldrich), and concentrated using a Spin-X UF filter (Corning). The concentrations of purified mAbs were determined using a Micro BCA Assay kit (Thermo Fisher Scientific) using a mouse IgG standard (Sigma-Aldrich), and purity was assessed by SDS-PAGE.

Additional anti-HA mAbs were used for binding and functional studies: the HA stem-directed mouse mAb C179 was purchased from Takara Bio (Shiga, Japan), the HA stem-directed human mAbs CR6261, FI6, F10 were purchased from Creative Biolabs (Shirley, NY), and SC/18-, PR/34-, SI/06-, Brisb/07- and CA/09-specific mouse mAbs were provided by BEI Resources or the International Reagent Resource (IRR). Mouse mAbs to recombinant H1 HA were from the following: influenza A/Solomon Islands/3/2006 (H1N1), clone AT170.558.146 (FR-499), clone AT170.119.(FR-503) and influenza A/Brisbane/59/2007 (H1N1), clone AT163.272.54 (FR-494), clone AT163.210.182 (FR-495), clone AT163.333.93 (FR-496), clone AT163.104.93 (FR-497), clone AT163.329.189 (FR-498) were obtained through the IRR, Influenza Division, World Health Organization Collaborating Center for Surveillance, Epidemiology and Control of Influenza, Centers for Disease Control and Prevention (Atlanta, GA). mAbs provided by BEI Resources as ascitic fluids were purified and quantified as described above. The C05 mAb-derived peptide (16) was kindly provided by Eva-Maria Strauch (College of Pharmacy, University of Georgia, Athens, GA).

ELISA was used to assess mAb reactivity against different H1N1 HA strains, including monomer or truncated HA1 rHA proteins. ELISA were performed as previously described (11). In brief, Immulon 4HBX plates (Thermo Fisher Scientific) were coated overnight at 4°C with 50 μl per well of a carbonate buffer solution (pH 9.4) containing 1 μg/ml COBRA P1, H1N1 (PR/34, Chile/83, Sing/86, TX/91, NC/99, SI/06, Brisb/07, CA/09, and Mich/15), or cH6/1 purified rHA and 5 μg/ml of (BSA) in a humidified chamber. Alternatively, plates were coated in a similar fashion with monomeric or truncated HA1 rHA versions of PR/34, Brisb/07, or CA/09 proteins. Plates were blocked with 200 μl per well of blocking buffer for at least 1 h at 37°C. The mAbs were 3-fold serially diluted in blocking buffer starting from 20 μg/ml, and plates were incubated for 1 h at 37°C. Plates were washed five times with PBS, 100 μl per well of HRP-conjugated goat anti-mouse IgG (Southern Biotech) diluted 1:4000 in blocking buffer was added, and plates were incubated at 37°C for 1 h. Finally, plates were washed five times with PBS and ABTS substrate (VWR International, Radnor, PA) was added, and plates were incubated at 37°C for 15–20 min. Colorimetric conversion was terminated by addition of 1% SDS (50 μl per well), and OD was measured at 414 nm (OD414) using a spectrophotometer (PowerWave XS; BioTek).

To assess the conformational or linear nature of the epitopes recognized by the mAb panel, the P1, CA/09, PR/34, SI/06, and Brisb/07 rHA were denatured following a 10 min incubation at 100°C. Subsequently, Immulon 4HBX plates were coated and ELISA was performed as described above.

A competitive ELISA was performed using unlabeled and biotinylated mAbs to identify overlapping epitope recognition. ELISA plates were coated with COBRA P1 or CA/09 rHA proteins as detailed above and then blocked with ELISA blocking buffer for 90 min at 37°C. Unlabeled mAbs or the C05 mAb-derived peptide (16) was serially diluted in blocking buffer, followed by addition of biotinylated mAbs, and plates were incubated overnight at 4°C. Plates were washed five times with PBS, HRP-conjugated streptavidin (Southern Biotech) diluted in blocking buffer was added, and the plates were incubated for 60 min at 37°C. Following extensive washing with PBS, ABTS substrate was added and plates were incubated at 37°C for development. Colorimetric conversion was terminated by addition of 1% SDS solution, and OD414 measured. The percent of maximal signal was determined using the equation 100 × [(OD₄₁₄ experimental sample − OD₄₁₄ blank)/(OD₄₁₄ maximal signal − OD₄₁₄ blank)].

Area under the curve (AUC) for ELISA and competition assays were calculated using GraphPad Prism V.8.01 software (San Diego, CA).

Binding kinetics of mAbs for prefusion (HA0) or postfusion rHA proteins was determined by biolayer interferometry using a FortéBio Octet RED384 system and anti–penta-HIS (HIS1K) or Protein G biosensors, respectively (Molecular Devices, San Jose, CA). The affinity (K_D) values of P1, PR/34-, SI/06, Brisb/07-, and CA/09-specific mAbs was determined against the homologous rHA. The K_D values for the control (C179, CR6261, FI6, and F10) and SC/18-specific mAbs were determined against the CA/09 rHA. All the rHA were used at 15 μg/ml. In brief, biosensors were first hydrated for 10 min in kinetic buffer (PBS [pH 7.4] supplemented with 0.5% BSA). Kinetic buffer was also used for mAb and rHA dilutions and for baseline and dissociation steps. For mAb affinity measurements, an automated five-step procedure was used and composed by an initial 60 s reference baseline followed by a 300-s loading phase followed by a 60-s baseline and a 300-s association (k_on) using four mAb dilutions (3.33, 1.67, 8.33, and 4.17 μM). Dissociation (k_off) was then performed for 600 s. All steps were performed at 25°C with 450 rpm agitation. The mAb K_D, defined as the ratio of k_off to k_on, was determined on the pre- and postfusion state rHA. Acquired data were analyzed using the FortéBio Data Analysis software (version 10.0) using a 1:2 model interaction.

Western blot was performed to assess the binding of mAbs to reduced and nonreduced rHA proteins. In brief, 5 μg of reduced or nonreduced rHA (COBRA P1, PR/34, SI/06, Brisb/07, and CA/09) were run on Bolt 10% Bis-Tris Plus SDS-PAGE (Thermo Fisher Scientific) according to the manufacturer’s instructions. Protein reduction was performed by resuspending the protein in SDS-PAGE sample buffer containing 5% 2-ME. Transfer was then performed using a polyvinylidene difluoride (PVDF) Trans-Blot Turbo RTA Transfer kit coupled with a Trans-Blot Turbo device (Bio-Rad, Hercules, CA) according to the manufacturer’s instructions. PVDF membranes were then loaded on an iBind Western device (Thermo Fisher Scientific) according to the manufacturer’s instructions using mAbs at a final concentration of 5 μg/ml and HRP-conjugated goat anti-mouse IgG (Southern Biotech) diluted 1:4000 in iBind solution (Thermo Fisher Scientific). PVDF membranes were then developed using a Clarity Western ECL substrate (Bio-Rad) and visualized using a ChemiDoc MP Imaging System (Bio-Rad).

The HAI assay was performed as previously described (4). In brief, mAbs were diluted in a series of 2-fold serial dilutions in v-bottom microtiter plates (Greiner Bio-One) starting from 20 μg/ml. An equal volume of each H1N1 virus, adjusted to ∼8 hemagglutination units per 50 μl, was added to each well. The plates were covered and incubated at room temperature for 20 min, and then 0.8% of turkey RBCs (Lampire Biologicals, Pipersville, PA) in PBS was added. RBCs were stored at 4°C and used within 72 h of preparation. The plates were mixed by agitation and covered, and the RBCs were settled for 30 min at room temperature. The HAI titer was determined by the reciprocal dilution of the last well that contained nonagglutinated RBCs. Positive and negative controls were included for each plate.

The FRA was performed similarly to previously described protocols (11). In brief, MDCK-SIAT1 cells were seeded at a density of 2.5–3 × 10⁵ cells/ml in a 96-well plate (Greiner Bio-One) the day before the assay was run. The following day, the cell monolayers were rinsed with 0.01 M PBS (pH 7.2) (Thermo Fisher Scientific), followed by the addition of 2-fold serially diluted mAbs at 50 μl per well starting with 20 μg/ml dilution in virus growth medium containing 1 μg/ml of L-(tosylamido-2-phenyl) ethyl chloromethyl ketone (TPCK)–treated trypsin. Afterwards, 50 μl of virus (COBRA P1 or CA/09 viruses) standardized to 1.2 × 10⁴ focus forming units per milliliter, and corresponding to 600 focus forming units per 50 μl, was added to each well, including control wells. Following a 2 h incubation period at 37°C with 5% CO₂, the cells in each well were then overlaid with 100 μl of equal volumes of 1.2% Avicel RC/CL (Type RC581 NF; FMC Health and Nutrition, Philadelphia, PA) in 2× MEM (Thermo Fisher Scientific) containing 1 μg/ml TPCK-treated trypsin, 0.1% BSA, and antibiotics. Plates were incubated for 18–22 h at 37°C, 5% CO₂. The overlays were then removed from each well and the monolayer was washed once with PBS to remove any residual Avicel. The plates were fixed with ice-cold 4% formalin in PBS for 30 min at 4°C, followed by a PBS wash and permeabilization using 0.5% Triton X-100 in PBS/glycine at room temperature for 20 min. Plates were washed three times with PBS supplemented with 0.1% Tween 20 (PBST) and incubated for 1 h with a mAb against influenza A nucleoprotein (IRR) in ELISA buffer (PBS containing 10% horse serum and 0.1% Tween 80 [Thermo Fisher Scientific]). Following washing three time with PBST, the cells were incubated with goat anti-mouse peroxidase-labeled IgG (SeraCare, Milford, MA) in ELISA buffer for 1 h at room temperature. Plates were washed three times with PBST, and infectious foci (spots) were visualized using TrueBlue substrate (SeraCare) containing 0.03% H₂O₂ incubated at room temperature for 10–15 min. The reaction was stopped by washing five times with distilled water. Plates were dried and foci were enumerated using an S6 macro ELISpot reader with ImmunoCapture 6.4.87 software (Cellular Technology). The FRA titer was reported as the reciprocal of the highest dilution of serum corresponding to 50% foci reduction compared with the virus control minus the cell control.

The Ab-dependent cell-mediated cytotoxicity (ADCC) assay was performed using the mFcγRIV ADCC Reporter Bioassay kit (Promega, Madison, WI) according to the manufacturer’s instructions. In brief, Immulon 4HBX plates were coated overnight at 4°C with 50 μl per well of a solution of carbonate buffer (pH 9.4) containing 5 μg/ml COBRA P1 or CA/09 rHA. The mAbs or an irrelevant isotype control (IgG2a,κ, clone MG2a-53; BioLegend, San Diego, CA) were 5-fold serially diluted starting from a final concentration of 20 μg/ml in assay buffer, consisting of RPMI 1640 medium (Thermo Fisher Scientific) supplemented with 4% Super Low IgG FBS (HyClone). The mAb dilutions were then added to the plates together with 5 × 10⁴ mFcγRIV effector cells per well in a final volume of 75 μl and incubated for 6 h at 37°C, 5% CO₂.

Plate content was then transferred in 96-well white polystyrene plates (Corning), and 75 μl of Bio-Glo Luciferase reagent was added and incubated for 5 min at room temperature. Plates were read using a GloMax-96 Microplate Luminometer (Promega), and ADCC activity was expressed as relative luminescence units. The AUC of relative luminescence units obtained at different mAb dilutions was then calculated using the GraphPad Prism V.8.01 software.

To analyze mAb sequences and confirm clonal uniqueness, the H chain variable (V_H) and L chain variable (V_L) fragments of each mAb were cloned. In brief, total RNA was extracted from each hybridoma cell line using the RNeasy MINI kit (Qiagen, Germantown, MD) according to the manufacturer’s instructions. RNA was reverse transcribed using the SuperScript III One-Step RT-PCR System (Thermo Fisher Scientific) and a pool of previously described primers (17) according to the manufacturer’s instructions. Sequences of amplified V_H and V_L fragments were analyzed using the IMGT database (18).

For comparison of Ab-secreting cells (ASC) in ELISpot experiments and mAb ADCC activities, data were analyzed using one-way ANOVA corrected for multiple comparisons (using the Tukey test). Unpaired Student t tests were used for comparison of V_H and V_L germline identity or CDR3 lengths of COBRA P1 and CA/09 mAbs. All statistical analysis was performed using GraphPad Prism V.8.01 software, and a p value <0.05 was considered statistically significant.

To focus on the memory recall response, total, as well as rHA-specific, IgG1 ASC present in the spleen were enumerated by ELISpot analysis 4 d after an i.p. booster injection. Splenocytes from mice vaccinated with COBRA P1 or any of the wild-type HA-immunized mice had similar frequencies (p > 0.05) of IgG1-positive ASC (Figs. 1, 2A). The total number of IgG1-positive ASC that bound the homologous rHA Ag was similar for each of the immunized groups of mice (Fig. 2A, Supplemental Fig. 1, Table I). Splenocytes from mice vaccinated with COBRA P1 possessed IgG1-positive ASC with reactivity against each of the tested H1N1 rHA proteins in the panel (Fig. 2). Specifically, mice vaccinated with COBRA P1 HA had IgG1-positive ASC against both seasonal rHA (NC/99) and pandemic CA/09 rHA proteins (Fig. 2A, 2B). In contrast, mice vaccinated with Brisb/07 HA had IgG1-positive ASC that only reacted with seasonal rHA Ags (i.e., Brisb/07 and NC/99). To this end, mice vaccinated with the COBRA P1 HA had a significantly higher frequency of IgG1-positive ASC demonstrating cross-reactivity with the CA/09 rHA compared with splenocytes from mice vaccinated with any of the wild-type seasonal H1N1 strains (Fig. 2B). All mice had similar numbers of IgG1-positive ASC that recognized the chimeric cH6/1 rHA, which was used to monitor reactivity against the more conserved H1 stem region (Fig. 2A). Collectively, immunization of mice with COBRA P1 HA-expressing influenza virus elicited ASC with increased breadth of H1N1 rHA binding compared with the vaccination with historical seasonal and pandemic H1N1 vaccines strains.

To determine the extent of Ab binding breadth against various H1 rHA proteins, mAbs were isolated following the generation of B cell hybridomas from mice vaccinated with COBRA P1 or CA/09 HA. Additionally, previously generated mAbs from seasonal or pandemic H1 HA-immunized mice were included for comparison. Purified mAbs assayed for anti-HA Ab binding to a panel of H1 rHA Ags (Fig. 3, Supplemental Fig. 1A, 1B). Approximately 20% of the COBRA P1 rHA-specific mAbs (6/28 mAbs) bound all tested rHA proteins in the panel, including the chimeric cH6/1 rHA (Fig. 3A). The remaining mAbs (22/28 mAbs) could be grouped into three additional categories based on binding reactivity against the rHA panel: 1) mAbs that bound only to the COBRA P1 rHA protein (2/22 mAbs) (Fig. 3D); 2) mAbs that bound strongly to P1 rHA and one to two additional H1N1 strains, including the CA/09 rHA (13/22 mAbs) (Fig. 3C); and 3) mAbs that exhibited broad reactivity against the majority of wild-type H1 rHA but did not bind the chimeric cH6/1 rHA (7/22 mAbs) (Fig. 3B). Similar to stem control mAbs, a set of wild-type H1N1 elicited mAbs bound all rHA in the panel including the cH6/1 rHA. Alternatively, a separate set of mAbs only bound the homologous or closely related rHA (Supplemental Fig. 1B).

To better characterize the binding profiles in the mAb panel, and the dependency on disulfide bridges and conformation, mAbs were tested against nonreduced or reduced homologous rHA following 2-ME treatment of rHA. COBRA P1 mAbs bound similarly to both reduced and nonreduced rHA molecules (Supplemental Fig. 1C). There was decreased binding against reduced rHA for some of the SC/18 (5D3 and 39E4), SI/06 (AT170.558.146), and CA/09 (2A12, 2B11, 3B8, and M2-2B9) elicited mAbs. Importantly, no binding was detected to the reduced rHA using the stem control mAbs (C179, CR6261, FI6, and F10) (Supplemental Fig. 1D).

To determine the nature of the epitope recognized by the mAb panel, mAb binding was tested against denatured rHA. Twenty-five percent (7/28 mAbs) of COBRA P1 mAbs (4E6, 3A1, 2C5, 1A12, 1E2, 2H6, and 4B3) recognized the denatured rHA (Supplemental Fig. 1C). Similarly, 1 out of 2 PR/34 mAbs (RA5-22), 1 out of 5 Brisb/07 mAbs (AT163.272.54), and 1 out of 17 CA/09 mAbs (3B6) recognized the denatured rHA (Supplemental Fig. 1D). All the other COBRA P1 mAbs, seasonal- and pandemic-specific mAbs, and the stem-directed control mAbs, lacked binding to denatured rHA protein (Supplemental Fig. 1C).

Finally, the affinity of each mAb was evaluated (Supplemental Fig. 1C, 1D). The K_D values of each mAb were assessed for the prefusion (HA0) and postfusion rHA. Affinity of mAb binding for the HA0 protein ranged from nanomolar to subnanomolar values for the COBRA P1-specific, seasonal, and pandemic as well as the control stem-directed mAbs (Supplemental Fig. 1C, 1D). Generally, increased mAb K_D values were observed for the postfusion state rHA, with some mAbs exhibiting similar affinity for the HA0 and postfusion rHA (Supplemental Fig. 1C, 1D).

Nine of the COBRA P1 HA-elicited mAbs possessed HAI activity against both the COBRA P1 and pandemic viruses (Fig. 4). Moreover, four of these mAbs (3H6, 1F8, 4H3, and 1B7) also had HAI activity against Phil/13 and/or Mich/15, which are antigenic drift variants of the swine-origin, pandemic CA/09 strain. Furthermore, one of these mAbs (1F8) exhibited broad HAI activity against the Chile/83, Sing/86, TX/91, NC/99, and SI/06 seasonal H1N1 strains, as well as HAI activity against distantly related H1N1 strains (SC/18 and PR/34). Three mAbs (2H6, 4E7, and 4C5) isolated from P1-immunized mice only possessed HAI activity against the homologous P1 virus. The remaining 16 mAbs from P1 immunized mice lacked detectable HAI activity against any of the H1N1 viruses in the panel (Fig. 4).

HAI-positive mAbs isolated from mice immunized with wild-type seasonal or pandemic H1N1 strains had a narrow pattern of HAI activity against strictly related influenza strains (Supplemental Fig. 2A). In particular, 11 (∼65%) CA/09 mAbs (5C12, CA09-02, CA09-09, CA/09-11, CA09-15, 1E6, 2A12, 2B11, 1G12, 3G6, 5B-2A12) had HAI activity against the homologous CA/09 virus. Of these CA/09 HAI-positive mAbs, seven also had HAI activity against the antigenically similar Phil/13, Mich/15, and SC/18 strains. The remaining CA/09 HA-specific mAbs lacked detectable HAI activity against any of the H1N1 viruses in the panel (Supplemental Fig. 2A). Similarly, HAI-positive mAbs from mice immunized with other wild-type H1N1 strains had HAI activity against strictly related seasonal strains and none of them exhibited a broad HAI activity against unrelated H1N1 strains tested (Supplemental Fig. 2A).

To determine whether P1 or CA/09 HA-elicited mAbs lacking HAI activity were capable of neutralizing virus infection, these mAbs were also evaluated in a FRA using homologous P1 or CA/09 viruses. As anticipated, the HAI-positive mAbs neutralized the homologous or both COBRA P1 and CA/09 viruses. In contrast, all of the HAI-negative mAbs, with the exception of 4E6 (P1 HA-elicited mAb) (Fig. 4) and M2-2B9 (CA/09 HA-elicited mAb) (Supplemental Fig. 2A), lacked neutralizing activity.

To evaluate the ADCC activity of mAbs generated following COBRA P1 and CA/09 immunization, representative mAbs from each panel with different binding specificities and belonging to different IgG subclasses were selected. Specifically, the following mAb categories were chosen from the COBRA P1, Brisb/07, and CA/09 mAb panel: stem-directed mAbs (IgG1, IgG2a, IgG2b), broadly HAI+ mAbs (IgG2a), and narrowly HAI+ mAbs (IgG1 and IgG2a). An irrelevant mouse IgG2a,κ isotype control mAb was also evaluated as a negative control.

Stem-directed COBRA P1 (2C5, 3A1) and CA/09 (3B6) mAbs using the IgG2a or IgG2b subclass constant regions possessed statistically significant ADCC activity (p < 0.05) (Fig. 5, Supplemental Fig. 2B). Additionally, the HAI-negative, COBRA P1 mAb (4B5), and CA/09 HAI-positive mAbs (CA09-15, 1E6 and CA09-09) also had statistically significant ADCC activity (p < 0.05) (Fig. 5, Supplemental Fig. 2B). Conversely, the COBRA P1 mAbs 1F8, 4C5 4E6, 4G10, 3H6, 3D3, 4B3, and 4A3, the Brisb/07 mAb AT163.272.54, and the CA/09 mAbs 2B11, 2G7, 5B-2A12, 2A12, CA09-02, M2-2B9, and 3B8 had similarly low and non–statistically significant ADCC activities (p > 0.05) (Fig. 5, Supplemental Fig. 2B).

To better define the specificities of the P1 and H1N1 wild-type HA-specific mAbs, competition assays were performed to cluster the mAbs into groups with similar recognition profiles (Fig. 6A, Supplemental Fig. 3A). Using a diverse panel of mAbs as probes, including those with different binding and HAI profiles (broadly, pandemic and postpandemic, pandemic only, and HAI negative) and stem-directed mAbs, the P1 and CA/09 mAbs could be segregated into different categories (Fig. 6A, Supplemental Fig. 3A). There was a set of 10 P1- and 6 CA/09-elicited mAbs that were not competitive with any of the HAI-positive mAbs. Some of these mAbs mildly/strongly competed with HAI-negative or stem-directed mAbs (2A10, 4E6, and C179) (Fig. 6A). However, none of these mAbs were competitive with broadly neutralizing, stem-specific mAbs CR6261, FI6, or F10.

The mAbs that were HAI positive against the COBRA P1 virus exhibited competition patterns that overlapped with CA/09 mAbs. Interestingly, one mAb (1F8) displaying broad HAI activity against H1 pandemic-like and H1 seasonal-like viruses was found to strongly compete with many other mAbs featuring a P1- and/or CA/09-specific HAI activity pattern (Fig. 6A, Supplemental Fig. 3A). Similar to the P1 HA induced mAbs, the CA/09 virus induced mAbs were segregated into two distinct clustering patterns on the basis of mAb competition assays. One group consisted of mAbs that strongly competed with probes that overlapped with P1 HA induced mAbs and the second group consisted of mAbs associated with the HAI-negative profile. In addition, representative mAbs were tested for competition with a peptide mimicking the binding of a broadly HAI-positive mAb, C05; however, no competition with this peptide was observed (Fig. 6B, Supplemental Fig. 3B).

To assess the diversity of P1 COBRA and CA/09 HA-elicited mAb panels, and confirm uniqueness of the individual mAbs, the Ig H and L chains were cloned and sequenced (Supplemental Table I, Table II). Among the single-cell cloned P1 COBRA HA-specific B cell hybridomas, 28 distinct sequences were identified and encompassed a diverse set of V_H and V_L gene subfamilies. In particular, the usage of the VH1 gene was the most frequent (46.4%) in the P1 mAb panel, with the most prevalent segment being the VH1-S22 (18%) (Fig. 7A).

Consistent with the selection of IgG-positive mAbs during the hybridoma screening pipeline, all P1 mAbs were encoded by class-switched H chains.

For the L chains, the VK4 gene was the most frequent (28.6%), whereas the most prevalent segment was the VK6-15 (17.9%) (Fig. 7B).

Furthermore, the H chains and L chains of all COBRA P1 HA-specific mAbs exhibited evidence of somatic hypermutation.

Similarly, the H chains from CA09 HA-specific B cell hybridomas were also cloned and sequenced. Among the CA09 mAb panel, 36.4% derived from the VH14 gene, with the most prevalent segment being the VH14-3 (27.3%) (Supplemental Table I). For the L chains, the VK14 gene was the most frequent (27.3%), with the two most prevalent segments being the VK14-111 and the VK1-110, with an equal frequency of 36.4% (Supplemental Table I). There was no statistical difference in the percentages of identity between V_L and the V_H germline sequences between the two groups. Also, the CDRH3 and the CDRL3 lengths were not statistically significant between the two groups, with the average CDRH3 aa length for COBRA P1 and CA/09 HA-elicited mAbs, ranging between a 12.6 and 14.6 aa (p > 0.05).

The development of a universal or broadly protective influenza vaccine is a high priority for researchers and public health authorities (19–21). Our research group has previously described a strategy, termed COBRA, that generates synthetic HA Ags that elicit broadly cross-protective Abs against multiple influenza viruses within a subtype compared with Abs elicited by a wild-type, strain-based vaccine (4–6, 9, 22). In accordance with serum Ab binding profiles from COBRA P1 HA-vaccinated animals (11), we have confirmed, for the first time, to our knowledge, that the COBRA P1 HA elicits ASC with reactivity against seasonal and/or pandemic HA Ags, including strong responses against the NC/99 and CA/09 rHA. Although both P1 and wild-type H1N1 virus immune splenocytes mounted detectable responses against the homologous rHA Ags, a minor component of these responses recognized the cH6/1 probe. Instead, the responses elicited by wild-type H1N1 viruses were relatively narrow in breadth (Fig. 2). Collectively, these observations support the immunodominance of the HA globular head in our immunization scheme and suggest that broadly reactive ASC elicited by COBRA P1 are mostly head directed. Subsequent B cell hybridoma generation, and detailed characterization of mAbs, was performed to characterize the COBRA P1 elicited B cell response at single-cell resolution.

Among the mAbs isolated from COBRA P1 HA-immunized mice, we identified only two that were exclusively specific for the P1 rHA Ag (Fig. 3D, Supplemental Fig. 1A). The remainder of the isolated mAbs (n = 26) were cross-reactive with different rHA Ags in the panel to varying extents. Interestingly, several mAbs isolated from COBRA P1 mice displayed broad recognition of multiple rHA Ags, including the cH6/1 probe. We also identified several P1 mAbs exhibiting potent HAI activity against the homologous P1 virus (Fig. 4). Additionally, there was another group of P1 HA-elicited mAbs with a more restricted rHA recognition pattern, similar to mAbs isolated following immunization with wild-type H1N1 viruses.

An important class of mAbs isolated from COBRA P1 immunized mice (n = 12) had HAI activity against the homologous P1 virus. The functional activity of these mAbs was also supported in a microneutralization assay (FRA). These mAbs segregated into four distinct categories: 1) broadly HAI-positive, 2) pandemic/postpandemic HAI-positive, 3) pandemic-only HAI-positive, and 4) COBRA P1-only HAI positive. Collectively, this implies that the COBRA P1 HA-elicited a B cell response targeting at least four distinct epitopes on the HA globular head region. Furthermore, these findings also demonstrate the ability of COBRA P1 HA to elicit B cell responses targeting conserved HA epitopes that were maintained on future H1N1 drift variants. Moreover, among the isolated HAI-positive mAbs from COBRA P1 HA-immunized mice, mAb 1F8 demonstrated broad and potent HAI activity against seasonal and pandemic H1N1 virus strains. Identification of mAb 1F8 in the context of our studies confirms the long-standing hypothesis that COBRA P1 is capable of eliciting B cells targeting conserved epitopes proximal to the receptor binding site (RBS), and which are capable of conferring protection against multiple H1N1 strains. Similarly, human mAbs targeting conserved epitopes encompassing or close to the HA RBS have been described (23, 24). As an example, the 5J8 (23) and the C05 (24) mAbs target regions adjacent or within the RBS and possess cross-neutralizing homosubtypic and heterosubtypic activity, respectively.

The remaining HAI-positive, COBRA P1 HA-elicited mAbs failed to demonstrate HAI activity against any of the seasonal H1N1 viruses included in the panel. This is noteworthy because HAI activity against several of these seasonal H1N1 virus strains was present in the serum of mice used for B cell fusions. Therefore, our studies indicate that COBRA P1-elicited HAI activity against seasonal H1N1 viruses is unlikely to be the result of narrowly reactive B cells. Instead, our data suggest that broadly reactive Ab specificities are at least partially responsible for conferring broad HAI activity in serum of COBRA P1 immunized mice. Future passive transfer studies will aim to verify this hypothesis.

Evaluation of prototype, broadly reactive stem mAbs in our binding studies confirmed their specificity for multiple rHA Ags in the assay panel, including the chimeric cH6/1 probe. These observations support structural integrity of rHA reagents used for binding studies but also offer an excellent dichotomy with mAbs isolated from COBRA P1 and wild-type H1N1 virus immunized mice. Whereas the prototype stem-reactive mAbs recognize a conformational epitope, composed of both HA1 and HA2 sequences (25), these mAbs do not bind the truncated CA/09 HA1 monomeric Ag. By contrast, many of the COBRA P1 and wild-type HA-elicited mAbs recognized the HA1 Ag, despite also binding the cH6/1 probe and monomeric rHA. This suggests that cross-reactive epitopes exist on the globular head (HA1) of H6 (A/mallard/Sweden/81/2002) used for probe construction (26) and multiple wild-type H1 strains, including the eliciting COBRA P1 HA Ag. Alternatively, these mAbs may recognize nonneutralizing conserved epitopes on the monomeric HA (27) or in close proximity to the C terminus of the HA1 domain, which are the targeted by HAI-negative Abs (28). Moreover, these P1 HA-specific mAbs also bind reduced rHA and do not compete with prototype anti-stem mAbs. Collectively, these data suggests that broadly reactive COBRA P1 HA-elicited mAbs in our panel are likely to recognize different stem epitopes compared with previously described prototype anti-stem mAbs.

The observation that several of the broadly reactive, COBRA P1 HA-elicited mAbs recognized the cH6/1 probe, along with the CA/09 HA1 monomer, also highlights two important points. First, this implies that recognition of the cH6/1 probe can occur through Ab reactivity directed against the HA1 region. Second, this demonstrates that the COBRA P1 HA was capable of selecting such broadly reactive B cell specificities from the repertoire and recruiting their participation in the elicited immune response. Collectively, this class of broadly reactive P1 HA-elicited mAbs serves to strengthen the hypothesis that the COBRA P1 HA immunogen recruits a diverse B cell repertoire recognizing multiple conserved epitopes. In this regard, future studies will aim to identify the specific epitopes targeted by these mAbs.

IgH sequence analysis of the COBRA P1 cell hybridoma panel was performed to assess the diversity and clonal relationship between isolated mAbs. Of note, COBRA P1 mAb sequences demonstrate evidence for somatic hypermutation and affinity maturation. Moreover, P1 isolated mAbs were clonally distinct and encompassed a diverse Ig repertoire. Interestingly, the IgH and IgL sequence analysis confirmed also the similarity in the Ig Ag-driven somatic hypermutation among clonally different mAbs endowed with a similar binding and functional profile. In particular, a set of two broadly reactive HAI-negative COBRA P1 mAbs (4E6 and 4H11), two narrowly reactive HAI-positive mAbs against pandemic and postpandemic strains and two narrowly reactive HAI-positive mAbs against pandemic strains are encoded by the same V_H gene segment and paired with the same V_L-encoded L chain (Table II).

Despite the overall sequence variability of the HA globular head region, the structure of the RBS is relatively conserved to preserve its receptor-binding function. Broadly reactive mAbs can mimic the conserved HA RBS residues and bind to sialic acid receptors (23) or, like the human C05 mAb, insert, through a long CDRH3, a single-loop into the receptor binding pocket of HA, thereby blocking receptor binding (24). The broadly cross-neutralizing mAb 1F8 may bind the H1 HA head region using similar mechanisms (23, 24, 29). However, 1F8 possesses a shorter (12 aa) CDRH3 compared with C05 (24 aa), and its epitope may partially overlap a region targeted by narrowly HAI-positive mAbs. This phenomenon has been described for other broadly cross-reactive head-directed mAbs (29). Additionally, no competition for binding to rHA was observed between 1F8 and a C05-derived peptide, thus suggesting a different binding mechanism (16).

This study identified mAbs encoded by the IgG1, IgG2a, or IgG2b subclasses, and which possess distinct binding specificities. In accordance with previously published studies (30), Abs using the IgG2a or IgG2b subclass constant regions were endowed with ADCC activity. In particular, cross-reactive mAbs, such as 3A1 and 2C5, possessed the highest ADCC activity. However, most of the mAbs elicited in this study were of the IgG1 subclass. The elicitation of predominately IgG1 Abs in our mAb panel is linked to the alum-adjuvanted vaccine administered i.p. during the immunization (31, 32). Based on prior experience, administration of an alum-adjuvanted virus booster immunization following initial intranasal priming was the most effective regimen for eliciting high titer Ab responses in influenza naive mice (11). Nevertheless, our data formally demonstrate the elicitation of both functionally neutralizing Ab and Fc-mediated ADCC effector activity by the COBRA P1 Ag. Collectively, these findings indicate that COBRA P1 HA-elicited humoral immunity encompasses multiple mechanisms of protection.

Despite characterization of 28 unique mAbs isolated following COBRA P1 HA virus vaccination, the identification of broadly cross-reactive and cross-neutralizing Abs was rare. This finding is consistent with previous studies showing that the breadth of HAI activity of COBRA P1 HA-specific mAbs against seasonal and pandemic strains as a whole is proportional to the HAI activity of polyclonal sera from COBRA P1 immunized mice (4, 11). Moreover, we cannot exclude that combinations of Abs targeting different HA epitopes can have a synergistic effect and thus extending the overall breadth of HAI and neutralizing activity as observed with polyclonal sera. Future investigations will address this hypothesis. In the context of a preimmunity setting, vaccination with COBRA P1 elicits broader neutralizing immune responses compared with other vaccine strains, suggesting an improved ability to recall broadly neutralizing Ab-secreting B cells (9). Further studies will be necessary to interrogate the prevalence of broadly cross-neutralizing Abs in naive and preimmune animals immunized with COBRA P1. The elicitation of broadly reactive and cross-neutralizing Abs suggests that a COBRA-based antigenic platform could represent a novel approach to manage antigenic drift. In the context of a preimmune settings, as is the case for the human population, COBRA-based HA could be superior at recalling cross-reactive Ab specificities and consequently may confer broader Ab-mediated protection.

We thank Emily Francis Clutter for technical assistance and Anne-Gaelle Bebin-Blackwell for helpful discussions.

H.K. is employed by Sanofi Pasteur. The other authors have no financial conflicts of interest.