Development of functional antibodies against influenza B virus by activation-induced cytidine deaminase in hybridoma cells.

Dear Editor,

Influenza B viruses (IBVs) have circulated among humans for more than 80 years. Seasonal influenza virus epidemics caused by two IBV lines (Victoria and Yamagata) and influenza A virus have considerable effects on public health globally and result in approximately 290,000–650,000 annual influenza-attributed deaths (Caini et al., 2015; Pan et al., 2015). Currently, the most effective countermeasures against influenza B virus infections are influenza virus vaccines (Subbarao and Matsuoka, 2013). However, these vaccines have limited efficacy because only strain-matched humoral immune responses are induced, while the globular head domain of viral hemagglutinin (HA) continues to evolve rapidly. In addition, few anti-influenza drugs are available, and the treatment window for administering the available drugs is small (Subbarao and Matsuoka, 2013; van de Sandt et al., 2015). A universal drug that targets both influenza B virus lineages remains a public health priority. Conferring passive immune protection using broadly neutralizing antibodies (bnAbs) may be a promising alternative viral infection treatment (Wilson and Andrews, 2012; Corti and Lanzavecchia, 2013; Walker and Burton, 2018).

We previously generated a panel of mouse anti-IBV mouse myeloma Sp2/0 hybridoma cells, among which several IgM antibodies were produced to target the receptor-binding site of influenza B and block viral infection with great breadth and potency. One of these antibodies, 7G6-IgM, was described in our previous study (Shen et al., 2019), and the other is 7G1-IgM, which is described in the present study. On the basis of a “cloning free” approach, previous studies have elegantly demonstrated that forced expression of activation-induced cytidine deaminase (AID) in hybridoma cells can induce somatic hypermutation (SHM) and heavy-chain class-switch recombination (CSR) of antibodies (Iglesias-Ussel et al., 2006; Stavnezer et al., 2008; Su et al., 2014). We used this strategy to generate different subtypes of anti-IBV IgG antibodies from 7G1 hybridoma cells and to obtain effective functional IgG antibodies against IBVs in mice. A schematic illustration of this strategy is shown in Fig. 1A. To control the expression of AID, a tetracycline-inducible AID (Tet-On) expression cassette (pLenti-CMVTRE3G-hAID-F2A-eGFP-Puro), which contains a tetracycline response element promoter (CMV-TRE3G), was constructed in house, and 7G1 hybridoma cells were forced to express rtTA after lentiviral transduction with the pLenti-CMV-rtTA-Blast vector (a gift from Eric Campeau, Addgene plasmid # 26429) (Supplementary Fig. S1). The cells were then selected in complete medium containing blasticidin (10 μg/mL) and puromycin (5 μg/mL) for 10 days to generate stable 7G1-TetOn-hAID cells, which were then incubated with 500 ng/mL doxycycline to induce the expression of enhanced green fluorescent protein (eGFP), corresponding to the inducible expression of human AID. Photographs were taken 2 and 4 days after incubation with doxycycline, and the eGFP fluorescence was observed (Supplementary Fig. S2). After continuous induction of AID for an additional two weeks, the cells were incubated with goat anti-mouse IgG antibody Fluor 568, and a significant IgG-positive signal was observed. Next, eGFP+ and membrane IgG+ cells were enriched via flow cytometry (BD Aria III) (Supplementary Fig. S3). After single-cell sorting by flow cytometry, the cells were then seeded into twenty 96-well plates at one cell/well without doxycycline in the medium. After two weeks of culture, anti-B/Florida/4/2006 hemagglutinin (HA) IgG-positive wells were selected by an ELISA assay.

Fig. 1

Class-switch recombination was mimicked in an IgM hybridoma cell line to develop effective functional IgG antibodies against influenza B virus. A Schematic illustration of the strategy used in this study. A Tetracycline-triggered activation-induced cytidine deaminase (AID) was constructed and transducted into 7G1-IgM hybridoma cells. Anti-FluB hemagglutinin (HA; B/Florida/4/2006)-positive IgG clones were isolated after “on-demand” heavy-chain class-switch recombination. B The number of class-switch recombination IgG isotypes obtained from IgM were isolated and determined in this study. Mouse hybridoma cell culture supernatants were collected and detected via the mouse immunoglobulin isotyping ELISA kit. C Hemagglutination inhibition (HI) and microneutralization (MN) of mouse 7G1-IgM antibody and three or four representative monoclonal antibodies (mAbs) of each IgG isotype. 1 × 105 of each hybridoma cells were cultured in the cell culture plate for 72 hours, and the cell culture supernatant was collected, serial 2-fold dilutions of the supernatant were mixed with an equal volume of virus for the assay. The neutralization titer was the highest cell culture supernatant dilution that was negative for hemagglutination. The HI titer was the highest cell culture supernatant dilution that was negative for hemagglutination. The red color labeled clones IgG2a-294, IgG3-126 and IgG3-44 were able to neutralize more influenza virus variants compared to the parental IgM antibody. The values are color-filled: red, strong reactivity; yellow, moderate reactivity; green, weak reactivity. D Hemagglutination inhibition (HI) and microneutralization (MN) activities of mouse 7G1-IgM antibodies and representative 7G1-IgG antibodies. The values are the averages of three independent experiments. Values below 50 μg/mL are color-filled: red-orange shades, strong reactivity; yellow shades, moderate reactivity; beige-light blue shades, weak reactivity; values >50 show no reactivity. E Binding of the indicated antibodies to nine strains of the three influenza B lineages. EC50 (ELISA-based half-maximal effective binding concentration) greater than 104 ng/mL were considered to be negative. F Antibody-dependent cell-mediated cytotoxicity (ADCC) activity of 7G1-IgM and representative 7G1-IgG antibodies against B/Florida/4/2006 virus stain-infected MDCK cells. MDCK cells infected by influenza B viruses at an MOI of 10 were used as target cells and mouse NK cells were used as effector cells in the ADCC assay. The indicated antibodies were tested at 20 μg/mL, 2 μg/mL and 0.5 μg/mL. The bars represent the mean ± SEM. *P < 0.05, compared to the control IgM group. The experiment was performed 3 times with mouse NK cells from 3 different mice. One representative dataset is presented in this study. G-I Comparison of the therapeutic efficacies of 7G1-IgM and representative 7G1-IgG antibodies in mice. Body weight changes (G and H) of BALB/c mice (n = 6 per group) treated intraperitoneally with the indicated antibodies (1 mg/kg or 0.01 mg/kg) 24 h after lethal challenge with MA-B/Florida/4/2006 or B/Brisbane/60/2008 at a 50% mouse lethal dose (MLD50) of 25. This experiment was repeated three times. The body weight curves represent the mean and 95% confidence intervals. Statistical analysis was performed with log-rank test (*P < 0.05, compared to the control IgG-treated group). Prophylactic efficacy of the dose of 10 mg/kg IgM and IgG antibodies against lethal challenge with 25 MLD50 of MA-B/Florida/4/2006 or MA-B/Brisbane/60/2008 by intravenous antibody administration (I). The numbers in the diagonal brackets indicate the numbers of mice infected comparing to mice survived. The survival curves of BALB/c mice (n = 4 per group) treated with antibodies (10 mg/kg) 1 day before lethal challenge are shown. This experiment was repeated three times; Log-rank test was used to assess the significance (*P < 0.05) of survival outcome. The control IgG is a mouse IgG1 monoclonal antibody against the HA of the influenza A virus generated by our laboratory. The ancestral virus strains represent the strains that were isolated before 1980 when the influenza B virus did not evolve into two lineages Yamagata and Victoria.

We successfully generated 380 ​cell clones, including 249 IgG3 subtype cell clones, 90 IgG1 subtype cell clones, 38 IgG2b subtype cell clones and 3 IgG2a-subtype cell clones (Fig. 1B). We next examined whether the increases in CSR were associated with SHM in the V region of 7G1; however by sequencing, no mutations were found in the representative clones. We compared the in vitro microneutralization (MN) and hemagglutination inhibition (HI) activities and the breadth of reactivity of the representative antibodies in the cell culture supernatant obtained from different antibody subtype cultures using a panel of nine representative influenza B virus strains derived from three distinct lineages. The IgM subtype antibody did not show MN or HI activity against the two ancestral ​influenza B virus strains; however, several IgG subtype antibodies exhibited MN and HI activity against all the influenza B viruses tested. The red color labeled clone IgG2a-249, IgG2b-294, IgG3-126 and IgG3-44 were able to neutralize more influenza virus variants compared to the parental IgM antibody (Fig. 1C). Next, four representative antibodies from each subtype were selected and purified from mouse ascites for further evaluation. The purified IgG1, IgG2b and IgG3 subtype antibodies showed HI and MN activity against all the viruses tested, the IgG2a subtype antibody showed HI and MN activity against most of the virus strain, and IgM did not show reactive activity against one ancestral and one Yamagata virus strains (Fig. 1D). In the binding test, all antibodies reacted specifically with all nine tested viruses derived from the three influenza B lineages. The four IgG subtype antibodies revealed significantly lower half-maximal effective concentration (EC50) values than the IgM antibody against the tested influenza virus strains (Fig. 1E). We also evaluated the antibody-dependent cell-mediated cytotoxicity (ADCC) activity of each antibody against B/Florida/4/2006. IgG2a and IgG1 subtype antibodies demonstrated high ADCC activity; in contrast, all the other antibodies demonstrated no ADCC activity against B/Florida/4/2006 (Fig. 1F). We finally evaluated the in vivo antiviral activity of each antibody using two mouse-adapted influenza B virus strains with a mouse model. A dose-ranging study was carried out by intraperitoneal or intravenous administration after intranasal virus challenge with MA-B/Florida/4/2006 or B/Brisbane/60/2008 at a 50% mouse lethal dose (MLD50) of 25. Ten milligrams per kilogram of each antibody administered intraperitoneally completely protected the mice against infection with viruses from the two lineages, and increased body weight was observed at the end of the study (Fig. 1G). In contrast, all the mice intraperitoneally administered 0.01 mg/kg IgM 7G1 died within 8 days of infection, and at this dose, each IgG 7G1 subtype showed a complete or partial therapeutic effect against the two influenza B viruses in mice (Fig. 1H). A single dose of 0.01 mg/kg IgG2a 7G1 protected 75% mice against the Yamagata influenza B virus B/Florida/04/2006 infection, in contrast, all the other classes of 7G1 and control IgG failed to protect any mice infected with this virus. A single dose of 0.01 mg/kg IgG2a and IgG1 7G1 protected 100% mice against the Victoria influenza B virus B/Brisbane/60/2008 infection, and the same dose of IgG3 7G1 protected 25% of mice and IgG2b protected 50% of mice against the infection. All the mice treated with control IgG died within eight days of infection. IgG2a 7G1 showed the best in vivo antiviral activity (Fig. 1H). When the infected mice were administered antibodies intravenously with 10 mg/kg, all the IgG subtypes showed protective potency against influenza B virus (Fig. 1I).

In this study, additional mutations in V genes were not found in either the heavy or light chains. This outcome may be a result of the frequency of CSR is higher than that of SHM (Iglesias-Ussel et al., 2006), and V genes may be insensitive to AID. For obtaining details on frequencies, a large number of cells should be sequenced. Long-term AID expression may lead to higher frequencies of CSR and SHM (Iglesias-Ussel et al., 2006). However, an “on-demand” or “Tet-on AID-expression” system is better for obtaining individuals or families of class-switched monoclonal antibodies from hybridomas. Our strategy can also be used to generate IgM hybridomas from IgG hybridomas with the same or mutated V genes. It is relatively difficult to develop bnAbs against influenza virus due to the high variability of HA; fortunately, the strategy used in this study shows the potential to further broaden the reactivity spectrum of neutralizing antibodies against influenza virus. Additionally, this strategy can be extended to the development of antibodies against other human pathogens. In a recent anti-COVID-19 antibody study, researchers found that engineering an IgM-neutralizing antibody from its IgG version can improve efficacy, reduce resistance and simplify the prophylactic and therapeutic treatment of COVID-19 (Ku et al., 2021; Yerabham and Ho, 2021). Forced expression of AID can facilitate the isolation of different class-switch variants from anti-SARS-CoV-2 hybridoma cells without antibody engineering or expression. Interestingly, in the present study, 65% of the IgG-switching clones were IgG3, and 24% were IgG1. The frequencies of IgG2b (10%) and IgG2a (1%) were much lower than the frequency of IgG1/3. Isotype convergence seems unlikely as we do not have enough clones. One possible reason is that the IgG3-constant gene is closer to the IgM-constant gene on the chromosome of the heavy chain, and they may be followed by IgG1, IgG2b, and IgG2a (Stavnezer et al., 2008). However, it is interesting to consider the reason that the frequencies of IgG3 and IgG1 isotypes were increased. This study observed that 7G1 of different classes displayed slightly different reactive spectrums against IBV and provided different protection efficiency in the mice model, which may be the reason that different antibody classes have different Fc structures and thus have different Fc-based functions. For example, IgG1 and IgG2a have strong Fc-mediated ADCC activity, while other types of antibodies have no such activity. We also observed differences in antiviral activity from different sub-clones with the same IgG isotypes in vitro. However, additional mutations in V genes were not found in all these antibodies. Forced expression of AID in the hybridomas may lead to additional mutations in other genes and affect the binding activity of the antibodies. To our knowledge, a variety of immune cells and cytokines are involved during the natural SHM in the germinal centers. This in vitro model does not fully emulate the complexity of the human immune system or the human in vivo microenvironment. Future studies are needed to screen more clones, harvest some of the hybridoma cells after a prolonged incubation and develop some in vivo model to figure out the mechanism.

In summary, this methodology may represent a general strategy to enhance antibody properties with “on-demand” CSR, and even SHM, without the need for antibody engineering or recombinant protein expression. Moreover, our strategy is convenient for investigating the functions of different isotypes of a therapeutic antibody during early drug discovery and development. We hope to generate an AID “on-demand” myeloma fusion partner cell line in the future study.

Footnotes

This work is funded by the (31870925), Shenzhen Science and Technology Innovation Commission for Research and Development Project (JCYJ20190809183205622) and Guangdong Basic and Applied Basic Research Foundation (2020A1515010368).The authors declare that they have no conflicts of interest. All in vivo studies were performed in accordance with Institutional Animal Care and Use Committee guidelines and were approved by the Ethics Committee of Xiamen University Laboratory Animal Center.

Supplementary data

The following is the Supplementary data to this article:

Supplementary data to this article can be found online at https://doi.org/10.1016/j.virs.2022.03.009.