VIC researchers analyzed features for each of 171 monoclonal antibodies in the VIC panel to define correlations between antibody characteristics and ability to offer protection (Graphical abstract from Saphire et al. doi: 10.1016/j.cell.2018.07.033)
In their paper published online on August 9, 2018 in Cell the VIC describes findings from a multi-year study conducted on a pool of antibodies donated by filovirus researchers from around the world. This study represents the first ever parallel comparison of antibodies against ebolavirus surface glycoprotein (GP) that were collected from a variety of sources, including those isolated from survivors of the 2013-2016 Ebola virus epidemic. By testing this large pool of antibodies in parallel, statistically significant relationships between antibody features and the likelihood of in vivo protection could be determined.
As expected, in vitro neutralization activity in cell culture had the strongest correlation with in vivo in a mouse model. In particular, those mAbs that targeted epitopes that are preserved following intracellular cleavage of the glycoprotein by host cell cathepsins had a higher likelihood of having both neutralization and protection.
However, some findings were not anticipated. This study is the first to compare the performance of three different neutralization assays that are standard in the field: i) neutralization of vesicular stomatitis virus bearing Ebola virus GP (rVSV-EBOV); ii) neutralization of ∆VP30 EBOV, which lacks a protein required for replication that is supplied in trans; and iii) neutralization of authentic EBOV. Results for these three assays did not always agree. In particular, the assay measuring neutralization of authentic EBOV captured a broader range of neutralization, especially for those mAb that target the GP glycan cap.
In addition, a subset of mAbs (around 5%) protected in the absence of significant neutralization activity. Rather than mechanically neutralization infections, these mAbs appear to mobilize an immune response, as shown by high throughput assays indicating that these mAbs evoke both antibody-mediated phagocytosis and to activate natural killer cells. Indeed, the ability to invoke natural killer cell activation seems to tip the balance in favor of protection.
Together these findings of discordancy between neutralization assay results and immune effector function suggest that those antibodies that do not have strong neutralization could nonetheless offer protection and thus should be be immediately excluded from further analysis.
Using statistical analyses and machine learning approaches, the VIC team compiled a correlation network that shows relationships between antibody features and likelihood of protection. Access this correlation network and view the features of each antibody in the VIC panel using the data explorer at the Apps link.