Machine-learning-based Structural Analysis of Interactions between Antibodies and Antigens

Computational analysis of paratope-epitope interactions between antibodies and their corresponding antigens can facilitate our understanding of the molecular mechanism underlying humoral immunity and boost the design of new therapeutics for many diseases. The recent breakthrough in artificial intelligence has made it possible to predict protein-protein interactions and model their structures. Unfortunately, detecting antigen-binding sites associated with a specific antibody is still a challenging problem. To tackle this challenge, we implemented a deep learning model to characterize interaction patterns between antibodies and their corresponding antigens. With high accuracy, our model can distinguish between antibody-antigen complexes and other types of protein-protein complexes. More intriguingly, we can identify antigens from other common protein binding regions with an accuracy of higher than 70% even if we only have the epitope information. This indicates that antigens have distinct features on their surface that antibodies can recognize. Additionally, our model was unable to predict the partnerships between antibodies and their particular antigens. This result suggests that one antigen may be targeted by more than one antibody and that antibodies may bind to previously unidentified proteins. Taken together, our results support the precision of antibody-antigen interactions while also suggesting positive future progress in the prediction of specific pairing.

Structures and therapeutic potential of anti-RBD human monoclonal antibodies against SARS-CoV-2

Background: Administration of potent anti-receptor-binding domain (RBD) monoclonal antibodies has been shown to curtail viral shedding and reduce hospitalization in patients with SARS-CoV-2 infection. However, the structure-function analysis of potent human anti-RBD monoclonal antibodies and its links to the formulation of antibody cocktails remains largely elusive. Methods: Previously, we isolated a panel of neutralizing anti-RBD monoclonal antibodies from convalescent patients and showed their neutralization efficacy in vitro. Here, we elucidate the mechanism of action of antibodies and dissect antibodies at the epitope level, which leads to a formation of a potent antibody cocktail. Results: We found that representative antibodies which target non-overlapping epitopes are effective against wild type virus and recently emerging variants of concern, whilst being encoded by antibody genes with few somatic mutations. Neutralization is associated with the inhibition of binding of viral RBD to ACE2 and possibly of the subsequent fusion process. Structural analysis of representative antibodies, by cryo-electron microscopy and crystallography, reveals that they have some unique aspects that are of potential value while sharing some features in common with previously reported neutralizing monoclonal antibodies. For instance, one has a common VH 3-53 public variable region yet is unusually resilient to mutation at residue 501 of the RBD. We evaluate the in vivo efficacy of an antibody cocktail consisting of two potent non-competing anti-RBD antibodies in a Syrian hamster model. We demonstrate that the cocktail prevents weight loss, reduces lung viral load and attenuates pulmonary inflammation in hamsters in both prophylactic and therapeutic settings. Although neutralization of one of these antibodies is abrogated by the mutations of variant B.1.351, it is also possible to produce a bi-valent cocktail of antibodies both of which are resilient to variants B.1.1.7, B.1.351 and B.1.617.2. Conclusions: These findings support the up-to-date and rational design of an anti-RBD antibody cocktail as a therapeutic candidate against COVID-19.

Advances and insights in the diagnosis of viral infections

Viral infections are the most common among diseases that globally require around 60 percent of medical care. However, in the heat of the pandemic, there was a lack of medical equipment and inpatient facilities to provide all patients with viral infections. The detection of viral infections is possible in three general ways such as (i) direct virus detection, which is performed immediately 1-3 days after the infection, (ii) determination of antibodies against some virus proteins mainly observed during/after virus incubation period, (iii) detection of virus-induced disease when specific tissue changes in the organism. This review surveys some global pandemics from 1889 to 2020, virus types, which induced these pandemics, and symptoms of some viral diseases. Non-analytical methods such as radiology and microscopy also are overviewed. This review overlooks molecular analysis methods such as nucleic acid amplification, antibody-antigen complex determination, CRISPR-Cas system-based viral genome determination methods. Methods widely used in the certificated diagnostic laboratory for SARS-CoV-2, Influenza A, B, C, HIV, and other viruses during a viral pandemic are outlined. A comprehensive overview of molecular analytical methods has shown that the assay's sensitivity, accuracy, and suitability for virus detection depends on the choice of the number of regions in the viral open reading frame (ORF) genome sequence and the validity of the selected analytical method.