Scientists from the US have recently characterized a panel of anti-severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibodies to identify molecular features of public antibody responses to SARS-CoV-2 spike protein.
The study is currently available on the nioRxiv* preprint server while the article goes under peer review.
Since the beginning of the coronavirus disease 2019 (COVID-19) pandemic, many studies have been conducted to characterize the humoral immune responses elicited by natural SARS-CoV-2 infection and vaccination. Many monoclonal antibodies targeting the spike protein of SARS-CoV-2 have been isolated and characterized as an attempt to identify effective therapeutic interventions.
The SARS-CoV-2 spike protein has three main domains including the receptor-binding domain (RBD), the N-terminal domain (NTD), and the S2 domain. Of these domains, the RBD is highly immunogenic and has been considered as the primary target for developing neutralizing antibodies.
A public antibody response defines a group of antigen-specific antibodies isolated from different individuals that share genetic elements and modes of antigen recognition. The most common strategy of studying public antibody response is to identify antibodies from different individuals that share the same immunoglobulin heavy variable (IGHV) gene and complementarity-determining region (CDR) H3 sequences.
In the current study, the scientists have conducted a systematic review of the literature and prepared a large dataset of anti-SARS-CoV-2 monoclonal antibodies with donor information. Using the dataset, they have studied the public antibody response to SARS-CoV-2 spike protein.
The scientists analyzed a total of 88 published articles and 13 patents and prepared a dataset of more than 8,000 anti-SARS-CoV-2 spike monoclonal antibodies isolated from more than 200 donors.
They analyzed immunoglobulin variable (V) gene usage and observed that antibodies targeting the RBD, NTD, and S2 have distinct patterns of V gene usage. Given the significance of CDR H3 in determining antigen-antibody binding, they determined the convergence of CDR H3 sequences among anti-spike antibodies. The findings reveal that public antibody responses to the RBD and S2 largely depend on CDR H3 sequence, whereas, the antigen-binding sites (paratope) of most of the anti-NTD antibodies are not dominated by CDR H3.
They identified a set of antibodies with paratopes primarily composed of CDR H3 and light chain. Moreover, they observed high enrichment of an immunoglobulin heavy constant delta (IGHD) gene (IGHD1-26) among anti-S2 antibodies, which were predominantly encoded by the IGHV3-30 gene. About 70% of these antibodies have a CDR H3 of 14 amino acids. With further analysis, they noticed that IGHD-dependent public antibody response to S2 is mainly driven by the heavy chain and that IGHV3-30/IGHD1-26 represents a public antibody response to a highly conserved epitope in S2.
By analyzing somatic hypermutation among anti-SARS-CoV-2 antibodies, they identified multiple recurring somatic hypermutations, including VH F27V, T28I, and Y58F, in IGHV3-53/3-66-encoded public clonotypes. In addition, they identified some novel recurring heavy/light chain somatic hypermutations, including VL S29R in an IGHV1-58/IGKV3-20 public clonotype.
With further analysis, they observed that antibodies belonging to the IGHV1-58/IGKV3-20 public clonotype bind to spike RBD. According to available literature, these antibodies could be induced by both vaccination and infection by distinct SARS-CoV-2 variants. Moreover, these antibodies are highly efficient in neutralizing different SARS-CoV-2 variants of concern (VOCs).
By analyzing the structure of the antigen-antibody complex, they identified that VL S29R forms a salt bridge with another somatic hypermutation to stabilize the antigen-antibody interaction.
Analysis of antigen specificity
The scientists used a deep learning model to differentiate between anti-spike and anti-influenza hemagglutinin (HA) antibodies. They trained the model with six CDR (H1, H2, H3, L1, L2, and L3) sequences and used a total of 4,736 anti-spike antibodies and 2,204 anti-influenza HA antibodies for the analysis.
The model showed the highest performance in distinguishing antigen-specific antibodies when trained by all six CDRs. A similar performance was also achieved when trained by three heavy-chain CDRs (H1, H2, and H3). When trained by three light-chain CDRs, the model performed reasonably. This indicates that the molecular information stored in the heavy-chain sequence is most valuable in determining antigen specificity.
The study identifies diverse molecular features of public antibody responses to SARS-CoV-2 spike protein. As mentioned by the scientists, the study findings can be used as a valuable resource to understand the molecular aspects driving antigen specificity of an antibody.
bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.