Fig 1.
Antibodies isolated from H2L2 transgenic mice by single-cell sequencing.
(A) Schematic diagram of experimental design of antibody discovery. Harbour H2L2 transgenic mice were immunized with Respiratory syncytial virus (RSV) pre-fusion (pre-F) protein, and the serum titers of antigen-specific antibodies in H2L2 mice were determined using enzyme-linked immunosorbent assay (ELISA). Spleen and bone marrow cells were harvested from these mice to isolate plasma B cells. Beads conjugated with RSV pre-F protein were employed to enrich B cells that specifically bind to RSV pre-F. Single B cell sequencing was conducted to recover the paired heavy- and light-chain sequences from each cell. Based on the sequence information, 78 sequences were selected for rapid construction and expression. The resulting antibodies were screened using neutralization assays of cell supernatants to identify candidate antibodies. (B) The chart shows the heavy- and light-chain germline gene pairing of 331 antibodies based on IMGT. The colors in the heatmap represent the number of antibodies identified with various coding genes, with the germline genes for the heavy and light chains displayed along the vertical and horizontal axes, respectively. The right of the graph indicates the abundance of the genes, with darker colors representing the higher abundance of the genes and lighter colors representing the lower abundance of the genes. All the data of this figure can be found in the S1 Table. (C, D) Phylogenetic analysis of the relationship between antibody variable gene segments and germline. The relationships between the heavy and light chain variable regions and the germlines from H2L2 transgenic mouse antibodies are shown. The analysis of the germline nucleic acid sequence of the antibodies was performed with IMGT/V-QUEST.
Fig 2.
In vitro characterization of selected antibodies.
Twelve monoclonal antibody strains were evaluated simultaneously by the cytopathic effect (CPE) assay (A) and focus-forming assay (FFA) (B) methods to neutralize live RSV A2 virus, and six of them showed antiviral activity, with PR306007 being more effective than palivizumab and 14N4 (positive controls). Data represents one of three independent experiments, shown as mean values ±standard deviation (S.D.). The data for Fig 2B can be found in S2 Table. (C, D) The neutralizing activity of PR306007 against the RSV B9320 virus was evaluated using the CPE method. Palivizumab (site II binder) was used as a positive control. Data were represented as the mean ± S.D. from n = 3 biologically independent experiments. The data for Fig 2D can be found in S3 Table. (E, F) The binding activity of PR306007 to the purified recombinant pre-F (DS-Cav1) and post-F from RSV A2 was measured using an enzyme-linked immunosorbent assay (ELISA). Data represents one of three independent experiments, displayed as the mean±S.D. of three technical replicates. Palivizumab and MEDI-8897 served as controls. (G, H, I) These antibodies (PR306007, MEDI-8897, and palivizumab) were tested for binding capability to the RSV pre-F (stain A2) by Biacore sensorgrams from triplicate experiments and representative results from one of these experiments are presented here. The original curves of three experiments are presented in S6 Fig. Palivizumab and MEDI-8897 served as controls. (J) These antibodies (PR306007, MEDI-8897, and palivizumab) were tested for their capability to bind the RSV pre-F by Biacore sensorgrams in triplicate experiments. The calculated KD values for the binding of each antibody to pre-F are summarized as the mean values of three experiments with standard deviation. The original curves of three experiments are presented in S6 Fig. Palivizumab and MEDI-8897 served as controls.
Fig 3.
Structure of the RSV pre-F in complex with the neutralizing antibody (nAb) of PR306007.
(A) Cryo-EM structure of the RSV pre-F ectodomain in complex with the nAb PR306007 fragment antigen-binding (Fab), as represented by the Cryo-EM map. Three RSV pre-F protomers are depicted in different colors, colored grey, dark grey, and purple. The variable regions of heavy chain (VH) is shown in blue, while the variable regions of light chain (VL) is shown in orange. In focus, only the variable regions of the Fab fragments are shown in the structure, while the constant regions are not modeled. The binding interface between RSV pre-F and PR306007 is framed and shown as a close-up view in the right-hand panel. This close-up view illustrates the epitope of PR306007 on pre-F, which interacts with residues at sites II and V of pre-F. (B) Interaction analysis between the RSV pre-F and PR306007. RSV pre-F are shown as molecular surfaces. PR306007-VH (blue): CDR1, CDR2, CDR3; PR306007-VL (orange): CDR1, CDR3, and FR-L3. (C) The binding region of PR306007 on the RSV pre-F. The binding region is colored magenta, and the buried surface areas (BSAs) of the light and heavy chains are 287 Å2(orange) and 298 Å2(blue). (D) Close-up view of RSV pre-F and PR306007 interaction interface. The variable domain of PR306007 and one pre-F protomer are shown as cartoon, colored blue (heavy chain), orange (light chain), and purple (pre-F), respectively. For clarity, the key critical contact residues are labeled and shown as sticks with oxygen atoms colored red and nitrogen atoms colored blue. Hydrogen bonds are depicted as black dotted lines.
Fig 4.
Comparison of PR306007 with previously published antibodies against RSV.
(A) The pre-F structure comprises three protomers displayed as surfaces, colored grey, dark grey, and purple, respectively. The epitope comparison of PR306007 with six other previously published antibodies on the A2 pre-F trimer: site Ø is depicted as a light blue line, site II as a red line, site III as an orange line, site IV as a dark purple line, site V as a dark blue line, and site I as a green line. The PR306007 epitope is highlighted in magenta. (B) The sequence conservation of 2383 full-length RSV F genes in the last decade was downloaded from NCBI. The percent conservation of each amino acid within the binding region was calculated and represented as color-coded bars: hydrophilic amino acid in red (acidic amino acid) and blue (basic amino acid), neutral amino acid in green and purple, and hydrophobic amino acid in black. F sequences of RSV A2 with a highlighted footprint of PR306007. (C) Details of PR306007 and previously published human antibodies of RSV, including the germline information for the heavy and light chains, as well as their binding sites on the F protein. Antibodies closely related to PR306007 in terms of germline are highlighted in green.
Fig 5.
Comparison of known structures of RSV F and the antibodies of site II.
(A-C) Structural comparison of the site II antibodies bound to RSV-pre-F: PR306007 (this study), 14N4 (PDB: 5J3D), and motavizumab (PDB: 3IXT). (A) The overall structure is displayed in surface form and rotated 90° in surface form. RSV F is shown as surface presentation colored in grey, PR306007 is shown as cartoon presentation colored in blue (heavy chain, HC) and orange (light chain, LC), and 14N4 in cartoons colored in green (HC) and brick red (LC), and motavizumab in cartoon colored in light purple. (B) Local magnified details of PR306007 and 14N4 binding to the RSV F site II. Despite both having similar regions, these antibodies adopt completely opposite orientations of the HC and LC. PR306007, 14N4, and RSV F site II are shown as cartoons. (C) Local magnified details of PR306007 and motavizumab binding to the RSV F site II, reveal a 38° difference in their binding angles.
Fig 6.
Prophylactic efficacy of PR306007 in RSV–infected mice model.
(A) Experimental design for PR306007 neutralization activity testing in Balb/C mice. (B) Body weight changes in mice. For prophylactic efficacy testing, mice were intraperitoneally injected with PBS, 2.5 mg/kg, or 0.625 mg/kg of PR306007 or palivizumab (n = 5 per group) at 24 hours before RSV infection. All groups except the PBS group were infected to RSV A2 stain. Weight changes were monitored daily until 4 days post-infection (d.p.i.). (C, D) Viral load in the respiratory tissues (including nose and lung lobes) collected at necropsy on 4 days post-infection was tested by qRT-PCR. Dashed lines indicate the limit of detection. Data were represented as the mean±S.D. from three biologically independent experiments (n = 3). Significance is determined by one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. All the data for this figure can be found in S5 and S6 Tables. (E) Histopathological analyses of mice. Representative images of lung sections stained by Hematoxylin–eosin(HE) infected with RSV at 4 d p.i.. Data are mean±SEM. Black, yellow, and red indicate pathological changes in the alveoli, bronchi/bronchioles, and inflammatory cell infiltration, respectively. The images on the top (bars = 1000 μm) are enlarged regions show a partially enlarged area depicted in the bottom images (bars = 100 μm).