Background
Influenza A virus still is a major cause of disease in humans, accounting for three to five million cases of severe illness and 250,000 - 500,000 deaths each year [
1]. Efficient influenza A vaccines are available, which induce antibodies predominantly against the two major components of the virus membrane, hemagglutinin (HA) and neuramidase (NA). Protection is mediated primarily by neutralizing antibodies against HA [
2,
3]. Since HA undergoes continuous change due to mutations (antigenic drift), new antigenic variants of influenza A arise every year requiring constant update of the vaccines. Effective vaccination is further complicated by the occasional reassortment of the segmented viral genome leading to the replacement of HA or NA from one subtype by another subtype, a processs called antigenic shift [
4]. Passive immunization with monoclonal antibodies (mAbs) targeting HA is very efficient [
5‐
7], however, suffers the same disadvantages as the current vaccines due to antigenic shift and drift.
An ideal target for active and passive immunization strategies would therefore be a conserved viral protein. The matrix protein 2 (M2) fits the bill and has received considerable attention as a potential target against influenza infection over the past decades [
8‐
23]. M2 is a tetrameric ion channel [
24‐
26] which is involved in virus uncoating in the endosome and in virus maturation in the trans-Golgi network [
27‐
29]. Its 23 amino acid extracellular domain has remained remarkably conserved in human influenza A virus isolates over the last hundred years [
30], at least in part due to the fact that the M2 protein is co-transcribed with the matrix protein 1 (M1) [
31,
32]. Whereas M2 is abundantly expressed on infected cells, only very few M2 molecules are present in Influenza A virus membranes [
23,
26]. In accordance with this, current seasonal influenza vaccines do not induce a significant humoral resonse against M2, and M2 specific antibodies (administered intravenously or induced by active immunization) mediate protection not by neutralizing virions, but by eliminating infected cells by ADCC [
15,
22].
Passive immunization with monoclonal antibodies has several advantages over vaccination. In particular, it allows treating people which poorly respond to vaccines, such as the elderly, young children or immune compromised individuals. In addition, passive immunisation is the treatment option of choice in situations where rapid protection is crucial, such as for post-exposure treatment or prophylaxis for the acutely exposed. A number of M2 ectodomain (M2e)-specific mAbs have been reported to protect mice from a lethal challenge in a prophylactic setting [
12,
17,
21‐
23]. While these mAbs include fully human antibodies derived from transchromosomic mice [
22], no natural human M2e-specific antibodies have been reported to date. However, for application in human subjects, natural human antibodies are the preferred choice. In contrast to humanized and fully human antibodies derived from phage display or transchromosomic mice, natural human antibodies combine the advantage of minimal immunogenicity with the smallest possible off-target reactivity and toxicity. Furthermore, human derived antibodies have the advantage of having gone through the affinity maturation process, resulting in high affinity antibodies.
We recently described a novel method for the efficient isolation of antibodies from humans by mammalian cell display [
33]. Here, we used this method for the isolation of natural human antibodies directed against M2e. We demonstrate that the antibodies bind M2 with high affinity and efficiently recognize M2 from a recently isolated H5N1 influenza A strain. The antibodies not only have potent prophylactic activities in a mouse model of Influenzy A infection, but also show efficacy in a therapeutic setting. Thus, the natural human antibodies described here have potential as immunotherapeutics against influenza infection.
Methods
Identification of M2-specific antibodies by mammalian cell display
Peripheral blood mononuclear cells (PBMC) were isolated from 10 ml of heparinized blood by Ficoll gradient. PBMC were pre-incubated with Alexa 647 nm-labeled Qβ and human gamma globulin (Jackson ImmunoResearch) and then stained with: (1) M2e coupled to Qβ in combination with a Alexa 488 nm-labeled Qβ-specific mouse mAb, as well as the M2-specific mouse mAb 14C2 (Abcam) in combination with FITC-labeled donkey anti-mouse IgG antibody (Jackson ImmunoResearch); (2) PE-labeled mouse anti-human IgM, mouse anti-human IgD, mouse anti-human CD14, and mouse anti-human CD3 antibodies (all BD Biosciences/Pharmingen); and (3) PE-TexasRed-labeled mouse anti-human CD19 antibody (Caltag Laboratories). After staining, cells were washed and filtered, and 334 M2-specific B cells (FL1-positive, FL2-negative, FL3-positive, FL4-negative) were sorted on a FACSVantage® SE flow cytometer (Becton Dickinson).
A Sindbis virus-based scFv cell surface display library was produced from antigen-specific B cells as described [
33]. BHK cells were infected with the Sindbis library and cells displaying M2-specific scFv antibodies were isolated using M2e coupled to RNase A in combination with an RNase-specific rabbit polyclonal antibody (Abcam) and a FITC-labeled donkey anti-rabbit IgG antibody (Jackson ImmunoResearch). Alternatively, cells displaying M2-specific scFv antibodies were isolated using Qβ-M2e in combination with M2-specific mouse mAb 14C2 (Abcam) and FITC-labeled donkey anti-mouse IgG antibody (Jackson ImmunoResearch). Each cell was sorted into a well of a 24-well plate containing 50% confluent BHK feeder cells. Upon virus spread (2 days post sorting), the infected cells were tested by FACS analysis for M2-binding to identify virus clones encoding M2-specific scFv antibodies.
Expression and purification of scFv antibodies
Fusion proteins were generated carrying an N-terminal human scFv fused to a C-terminal mouse Fc-γ2c domain. Thus, scFv coding regions were PCR amplified from Sindbis virus-containing supernatants by RT-PCR, digested with the restriction endonuclease Sfi1 and cloned into the expression vector pCEP-SP-Sfi-msFc-γ2c. This vector is a derivative of the episomal mammalian expression vector pCEP4 (Invitrogen), carrying the Epstein-Barr Virus replication origin (oriP) and nuclear antigen (encoded by the EBNA-1 gene) to permit extrachromosomal replication, and contains a puromycin selection marker in place of the original hygromycin B resistance gene. The resulting plasmids drive expression of scFv-msFc-γ2c fusion proteins under the control of a CMV promoter.
Fully human γ1 heavy chain and κ light chain coding regions were generated by total gene synthesis (GeneArt AG, Regensburg, Germany) and combined into the EBNA-based expression vector pCB15 essentially as described [
33].
Expression of the scFv-msFc-γ2c fusion proteins, as well as fully human IgG1κ antibody was done by transfecting the expression vectors into HEK-293T cells, using Lipofectamin Plus (Invitrogen). For large scale production and purification, stable protein-expressing cells were enriched by selection in the presence of 1 μg/ml puromycin (Sigma). Pools of resistant cells were maintained in serum-free medium on Poly-L-Lysine coated dishes or in roller bottles for up to 3 weeks. Supernatants containing the respective antibodies were collected twice a week and filtered through a 0.22 μM Millex GV sterile filter (Millipore). Both types of antibodies were purified by affinity chromatography over a protein A-Sepharose column (GE healthcare).
ELISA analysis
ELISA plates (96 well MAXIsorb, NUNC) were coated with RNAse-M2e-cons or RNAse-M2e-VN at a concentration of 4 μg/ml in coating buffer (0.1 M NaHCO3, pH 9.6) for one hour at 37°C. The plates were then washed with wash buffer (PBS/0.05% Tween) and blocked for 2 h at 37°C with 3% BSA in wash buffer. The plates were then washed again and incubated with 3-fold serial dilutions of scFv-msFc-γ2c in wash buffer containing 1% BSA. Plates were incubated for 2 h at room temperature and then extensively washed with wash buffer. Specifically bound antibodies were then detected with HRPO-labeled, Fcγ-specific, goat anti-mouse IgG antibody (Jackson ImmunoResearch Laboratories). After extensive washing with wash buffer, plates were developed with a 0.4 mg/ml solution of 1, 2-ortho-phenylenediamine dihydrochloride (OPD) in citric acid buffer (35 mM citric acid, 66 mM Na2HPO4, pH 5.0) containing 0.01% H2O2. After 10 min the reaction was stopped with a 5% solution of H2SO4 in H2O, and plates were read at 450 nm on an ELISA reader (Biorad Benchmark).
Affinity measurement by Friguet ELISA
A 10 ng/ml solution of, respectively, scFv-D005-msFc-γ2c, scFv-E040-msFc-γ2c or scFv-F052-msFc-γ2c, was incubated in the presence of different concentrations of M2e-cons peptide (3-fold serial dilutions corresponding to 10 nM to 0.17 pM) in PBS/1% BSA. After 2 h at room temperature, free antibody was detected by a classical ELISA similar to the one described above. For this, ELISA plates that had been coated with RNAse-M2e-cons conjugate at a concentration of 20 ng/ml at 4°C overnight were washed with wash buffer (PBS/0.05% Tween) and blocked for 2 h at 37°C with 3% BSA in wash buffer. The plates were then washed again and incubated with the solution binding reactions for 30 min at room temperature. After extensive washing with wash buffer, bound scFv-Fcγ2c fusion proteins were detected by a 1 h incubation at room temperature with a HRPO-labeled, Fcγ-specific, goat anti-mouse IgG antibody (Jackson ImmunoResearch Laboratories). After extensive washing with wash buffer, plates were developed as described above. The Kd values were determined as the EC50 of the ELISA signal as a function of the M2e-cons peptide concentration present in the solution binding reaction.
Mouse model of Influenza A infection
Prophylactic and therapeutic activity of antibodies was tested in a mouse model of Influenza A infection. Thus, six weeks old female C57BL/6 mice (6 per group) were infected intranasally with a lethal dose of mouse-adapted (m.a.) influenza A virus PR8 (4 × LD50), followed by close monitoring of weight-loss and fever (twice daily at peak of infection). Antibodies were injected intraperitoneally either 2 days before (prophylactic setting) or 1 to 3 days after infection (therapeutic setting). One day later, mice were bled in order to verify the presence of the antibodies in the blood by ELISA (not shown). Antibodies were readily detectable in the sera of all mice, except for one mouse receiving 500 μg scFv-D005-msFc-γ2c (Figure
3) and one mouse receiving 60 μg scFv-D005-msFc-γ2c (Figure
4A); these mice were subsequently removed from the analysis. All animal experiments were carried out in accordance with protocols approved by the Swiss Federal Veterinary Office. Animals whose body temperature reached 30°C were considered moribund and had to be euthanized immediately due to abortion criteria defined by the Veterinary Office.
Competing interests
All authors are present or former employees of Cytos Biotechnology AG and hold stocks or stock options in the company. The authors have no additional financial interests.
Authors' contributions
RRB, MB, NS, WAR, PS, and MFB designed research; RRB, MB, RBB, MG, and SM performed research; RRB, MB, NS, WAR, PS, and MFB analyzed data; RRB wrote the paper. All authors read and approved the final manuscript.